You are on page 1of 82

A GUIDE TO THE PANASONIC DVX CAMERAS

How to Get the Most From the Revolutionary DVXl 00, DVXIOOA and DVXIOOB Cameras

Written by Barry W. Green, Jan Crittenden Livingston and Harry W. Foulds

Published by Fiercely Independent Films Inc., P.O. Box 371282, Las Vegas, Nevada 89137, USA. (702) 312-FlLM.

© 2004, 2005 Fiercely Independent Films, Inc. All rights reserved. No part of this book may be reprinted or reproduced in any form or by any means without the express written consent of Fiercely Independent Films, Inc.

First printing October 2004 Revised September 2005

TABLE OF CONTENTS

Foreword 1

Introduction to the DVX Cameras 4

Improvements in the AG-DVX100A 10

How These 26 New Improvements Benefit the Cinematographer 12

Improvements in the AG-DVX1 OOB 17

DVX Articles, Frequently Asked Questions, and Commentary

What's That Click? 21

Strobin 21

Audio Delay in 24P ModelEcho in DVX100A Headphones_27

Recommended F-Stop Range 28

Battery Life 30

The DVX ASA/ISO Rating 30

24P vs. 24PA 32

Frame Rates and Slow Motion 37

Anamorphic vs. Letterbox vs. Squeeze 40

Interlace Lines on Progressive Footage 46

Depth ofField 49

Matching DVX100 to DVX100A or DVX100B 55

Synchronizing Multiple Cameras in a Multi-Camera Shoot __ 55

MENU SETTINGS

SCENE FILE SETTINGS

Detail Level 58

V. Detail Level 58

Detail Corin 59

Chroma Level 59

Chroma Phase 59

Color Temp 60

Master Pedestal 60

Auto Iris 61

Gamma 61

~ee 64

Skin Tone DTL 64

Matrix 65

V Detail Fre 65

Progressive 68

Name Edit 69

Savellnit 69

CAMERA SETUP SETTINGS

Syncro Scan~ 70

Aspect Conv 70

Color Bar 72

Setup 73

1

SW MODE SETTINGS
Mid Gain 75
High Gain 76
ATW 76
Handle Zoom 77
Iris Dial 77
Userl,2,3 78
AUTO SW SETTINGS
A. Iris 80
AGC 80
ATW 80
AF 81
RECORDING SETUP SETTINGS
Rec Speed 82
Audio Rec 82
MicALC 83
Mic Gain 83
1394 TC Regen 84
TCMode 84
TCG 85
FirstRec TC 86
TC Preset 86
1394 DB Regen 87
DB Mode 87
DB Preset 88
Interval Rec 89
RecTime 90
Interval Time 90

A V IN/OUT SETUP SETTINGS
A. Dub Input 91
DVOut 91
DISPLAY SETUP SETTINGS
Zebra Detect 92
Marker 93
Video Out OSD 94
DatelTime 94
Level Meter 94
Zoom-Focus 95
Tape - Battery 97
Other Display 97
Camera Data 98
EVF Color 98
Display Aspect 99
11 OTHER FUNCTIONS SETTINGS

Remote 100

DV Control 100

DV CMD Sel 101

RecLamp 102

Beep Sound 102

Tape ProtectIPower Save 103

H.P. Mode 103

User File 103

File TranslFile Receive 104

PHYSICAL CAMERA SWITCHES AND JACKS, 106

ANAMORPHIC ADAPTER USERS GUIDE 128

ABOUT THE AUTHORS 157

• "LEICA" is the registered trademark of Leica Microsystems.

• "DICOMAR" is the registered trademark of Leica Camera AG

"Panasonic", "DVX", "DVXIOO", "AG-DVXIOO", "AG-DVXIOOA" and "AGDVXIOOB" are registered trademarks ofPanasonic Broadcast and Television Systems

Company.

111

Foreword

At the National Association of Broadcasters convention in 2002 Panasonic made two startling announcements. First, they announced that they would be developing a new video camera that shoots progressive-scan video at 24' and 30 frames per second, as well as regular interlaced video. Second, they announced that this amazing new camera would be available and on the market by October of that same year!

Many~eople scoffed. 24P, up until that time, had existed only in expensrve high-definition cameras like the $100,000 Sony CineAlta and Panasonic's own $65,000 VariCam. 24P delivers a filmlike look, which is one reason why it was chosen to shoot the new Star Wars films. Could Panasonic be serious? Would they really bring out a 24P-capable video camera, at such a low price ($3795 MSRP for the original DVX100) and in such an abbreviated time frame?

They were serious. They did produce the camera. And they actually beat their "street date", with some units being delivered to dealers before the October 10 deadline! Not only did Panasonic deliver the 24P feature as promised, but they surrounded it with the most fullfeatured, powerful, and innovative video camera available at the un~er-$4,000 price point. The DVXI00 delivered powerful imaging optt~ns, to be sure, but Panasonic set new standards for audio quality and Image controls and durability and lens quality.

The DVXI.OO was truly a watershed moment in camcorder history. Not only did the camera deliver the most sought-after features, but the company representatives went "to the people" to find out exactly what the customer wanted, and then built it. Jan Crittenden Livingston, DVXI00 Product Line Business Manager for Panasonic US, personally spent countless hours on various web discussion groups, internet mailing lists, and at conventions and trade shows interacting with video professionals and indie filmmakers, learning what features would be most important to these customers. It was on one o~these discussion lists where I first met Jan (on-line), and I was truly Impressed that a manufacturer such as Panasonic would devote

I

so much time from such high-level employees to meet and discuss with end-users, and take our suggestions so seriously as to develop a product that nearly perfectly met our ''wish list."

And it certainly did meet our "wish list", and in many ways exceeded it. While many are focused on the 24P features, it is my assertion that the DVX would have been a great success even without that trademark feature: it is truly an outstanding interlaced video camera, with unparalleled image quality, audio quality, and image control features. The addition of24P is a bonus that easily put the DVX at the top of the "prosumer" camera category. Sales reflected its wide acceptance. The camera was, and is, a runaway success, and one of the most talked-about products on video forums and in video magazines.

How could Panasonic top that? In just over one year from the DVXI00's date of release, and without any competitor's challenge, Panasonic revised and introduced a new model of the camera, the DVX100A. They incorporated 26 changes to the original, adding some very nice upgrades, and kept the price almost exactly the same (MSRP $3995, vs. $3795 for the original). And inSeptember 2005, Panasonic introduced the DVX100B, adding several enhancements that make the DVX an even more useful professional production camera.

Does the DVX100 live up to the expectations? Most definitely. As a professional film/video producer, producing corporate films and television commercials, I can say that fully 90% of my work in 2002 was shot on film, usually 16mm, with about 10% shot on video. After acquiring a DVXIOO in March of2003, that ratio changed significantly: in 2003, approximately 75% of my work was shot on the DVX, with film dropping to about 15% and other high-end video formats accounting for the remaining 10%. In 2004, the numbers grew: approximately 85% of my company's work that year was shot on the DVXI00. It really is that good. The DVXIOO has nearly eliminated the need to shoot 16mm film, and with the P+S Technik mini35™ adapter, we've even been able to shoot some work on the DVXI00 that would otherwise have needed to use 35mm film.

2

The DVX cameras have found a rabid following among video shooters and independent filmmakers. One of the most popular sites is www.DVXUser.com. where Jan and I and several others frequently participate. Over the last couple of years of using the camera, and participating in the forums, it's become clear that there are many, many users who wish that there was more information available about how to use the camera and how to get the best performance from it. To that end, I have collaborated with Jan Crittenden Livingston and also with Harry Foulds ofPanasonic to present this guidebook and DVD. It is our sincerest wish that you will be able to use this material to gain a greater understanding of your camera, and that this better understanding will allow you to take your cinematography to higher heights.

This guidebook is not intended as a replacement for the user's manual. Rather, it is a collection of essays on the camera, as well as a thorough examination of its features and settings, accompanied by suggestions for how to best use those settings to get the results you're after. The DVXlOO owner's manual is an important tool in understanding how to use the camera; this guidebook will refer back to the manual (and specifically to the numbering system for switches and buttons on the camera). Additionally, some subjects covered adequately in the manual will not be discussed here; rather this guidebook will expand upon subjects that I feel may not have been explained thoroughly enough in the manual. You may find some technical jargon in these pages, but we've tried to provide plainEnglish explanations so that users of all experience levels will

benefit.

3

Introduction To The DVX Cameras

The AG-DVXI00 has captured the imagination of creative filmmakers, broadcasters, and video professionals around the world, with its 24P (progressive) capabilities. With the introduction of the DVXlOOA and new DVXI00B, Panasonic has further raised the bar.

Panasonic has won a total of 15 Emmys, 12 of them for advancements in the area of digital video technology. A Technical Emmy is awarded for a technology that has materially changed the way the industry works. Panasonic's engineers are continuously pushing the limits of product design, and several of these awardwinning technologies come into use within the body of the DVX cameras, namely, Digital Signal Processing and CCD Design. There are a couple of others that come into play in the design of the DVX which are directly derived from the technology that brought about the HDC-27V VariCam. Suffice to say that the DVXlOO and its successors (DVXI00A and DVXI00B) are breakthrough products.

Main Features oftbe AG-DVXIOO Cameras

3 CCD Imaging System

Exceptional picture quality, due to a new 113 inch, three CCD progressive imaging system that has 410,000 pixels per chip. This new system was developed specifically for broadcast and professional applications. These new CCD's are able to record light as low as three lux, enabling the shooter to record nighttime events in good detail. The picture quality is outstanding, with a high signal-to-noise ratio and low smear in bright areas.

High Image Quality with 12-Bit AID Conversion

The AG-DVXI00A and DVXI00B feature an AID converter that uses the same 12-bit processing as a broadcast camera-recorder. Precisely digitizing the gradation and color captured by Progressive CCD's, this AID converter supports gamma switching and other fine downstream image adjustments - one of the keys to achieving rich image expression. The DVXlOO uses a lO-bit AID converter.

4

World's First DV 24p Mode and Cine-Like Gamma

The DVXlOO series of handheld cameras was designed to produce video images with strikingly similar tones to film images. You have a choice of three shooting modes: 24p (24 fps progressive) for images with a movie-like look and motion; 30p (30 fps progressive); or standard 60i (60 fps interlaced). On the PAL version two shooting modes are available: 25p and standard 50i interlaced. Images that are captured in 24p mode are recorded onto the videotape in NSTC standard 60i video format, which allows for playback on any ordinary DV VCR and TV monitor, and edited on any DV system. Most nonlinear editing programs can recognize the 24P sequence from the camera and allow you to edit the original 24P footage as 24P footage, even through the final finishing stages of mastering your video to a 24P DVD, just like Hollywood does!

Zoom Lens

The DVX100 series of cameras use a Leica Dicomar™ lens which incorporates the optical technology ofLeica Camera AG, one of the world's premium lens design companies. This lens system employs 15 lens elements in 11 groups, including two asphericallenses, to render a shaper image. A Leica multicoatingprocess is applied to the lenses to reduce flare and ghosting. The lOX zoom has been designed to increase the lens' wide-angle (f=4.5 to 45mm) capabilities. Wider-angle lenses are very helpful in many shooting situations, and with the DVX, Panasonic has delivered a camera with the widest stock lens available. The lens utilizes a cam-driven manual zoom ring that offers true manual zoom - a first for a lowcost camera. Whereas other cameras expect you to use servo-driven focus and zoom controls, the DVX provides pure mechanical zoom, and also provides precise manual focus, including readouts in the display that let you know the precise zoom and focus position of the lens! The power zoom of the DVXIOOA and DVX100B offers a slow 30-second zoom and a multi-speed servo-driven zoom rocker switch. Additionally there is a three-speed zoom rocker on the handle.

Optical Image Stabilizer Compensates for Hand Held Production Incorporated in the DVX series is an Optical Image Stabilizer (OIS) that compensates for slight hand shaking that may occur when

5

shooting handheld. A gyro driven sensor detects handshake and sends signals to a linear motor, which then adjusts the lens to compensate. Because this image stabilization system is optical, rather t~an. electronic, this system won't compromise the picture qualIty.In any way. Some other cameras use digital image stabilization which c~n lower picture quality, but with the DVX's OIS you're always getting pure untouched images.

Gamma Processor Provides Rich, Cine-Like Tones

Panasonic has greatly expanded the expressive capability of the AGDVX100 cameras by creating a unique gamma function, such as Cine-Like gamma curves, which produce images strikingly similar in tone to film images. For each of the Y/CRICB signals, the ga~a curve settings are processed immediately upstream from the digital signal processing circuit. In the DVX100A and B, three new gamma curves have been added: Cine-Like-D, Cine-Like-V, and B.Press.

The Gamma curve settings menu allows you to select from seven gamma curves·

Gamma Image
CINE-LIKE Film-like images
. CINE-LIKE-D The Cine-Like mode shifted to prioritize
dynamic range
CINE-LIKE- V The Cine-Like mode shifted to prioritize
contrast
LOW Images with strona black contrast
B. PRESS Images with even stronger black contrast
NORM Standard video 2amma
HIGH Bright images with enhanced gradation
in dark areas and soft contrast Scene File Dial for Quick, Easy Camera Setup

The scene file dial (which is located on the back of the camcorder) is preset for a variety of shooting conditions. In use, you can instantly retrieve the settings by simply selecting the desired position. Six suggested preset files are provided (F1 to F6) by the factory. The scene file dial lets you keep a variety of "looks" on hand, ready to switch to at a moment's notice. You can change a wide variety of

6

settings, including color response, detail level, noise suppression, auto-iris compensation, and many more.

You can freely change any of the six flle names and their settings to develop your own customized camera.

Scene File Name Function
Fl -- Standard settings
F2 FLUO Indoor shooting under fluorescent
lights
F3 SPARK Highlighting subjects at
receptions, dinners, and other
gatherings
F4 B-STR Enhanced gradations of
luminance in low li~ht scenes
F5 24P 24p mode + Cine-Like-V ~amma
]f6 ADVANC Advanced 24p mode + Cine- Like-
D~amma IEEE 1394 terminal for non-linear production with Synchro . Lock Function

T?e DVX ca~eras come equipped with an IEEE 1394-compliant 4- pm DV termmal that makes it easy to upload data to a non-linear editor (NLE) or dub onto a DV recorder. This terminal also features a new synchro lock function that allows the DVX to remotely start and stop an external DV device connected to it via a DV cable. Three recording modes help protect against mistakes: record only onto the external recorder, record onto both the DVX and external recorder or begin external recording when the DVX's tape ends. Using the ' synchro-lock function in "CHAIN" mode, it's possible to record programs that are longer than one DV tape, because the DVX can automatically trigger an external recorder to start recording when the DVX's tape is nearly full.

Userl, User2 and User3 Buttons for Customized Operation

The DVXlOO provides two user buttons, and the DVXlOOA and B provide three us~r buttons~ each of which can be assigned anyone of the eleven functions descnbed below. These assigned functions can then be accessed at the touch of a button.

7

Items Assignable Functions
COLOR BAR Display / hide the SMPTE color bars
SPOTLIGHT Turns auto iris spotlight on / off
BACKLIGHT Turns auto iris backlight compensation
on/off
BLACKFADE Fade out to black (and fade audio)
WHITEFADE Fade out to white (and fade audio)
MODECHECK Display camera settings in viewfinder
ATW Turns auto tracking white balance
function on/off
ATWLOCK Lock/unlock white balance in ATW
operation
GAIN 18 dB Switches the image gain to + 18dB
INDEX Writes an index signal to the tape
(DVXIOOAIB only)
SLOW SHUT Turns the Slow Shutter mode ON or OFF
(DVXIOOAIB only) 3.5" Color LCD Monitor

This large 3.5" LCD monitor rotates 270 degrees. This improves your shooting flexibility by making it easier to monitor high-angle shots or self recordings. On the DVXlOOA and B, a new detail (PEAKING) function has been added which helps focusing; also you can set the viewfinder monitor to color or black and white. Peaking is normally found only on expensive broadcast cameras.

Built-In SMPTE Time Code Generator and Reader

The DVX cameras record SMPTE- compliant Vertical Interval Time Code (VITC) into the Sub-Code area of the videotape. You can select from Drop Frame or Non Drop Frame recording modes, and Free RunlRec Run modes. User bits (VB) are also provided, letting you record your choice of date, time, Time Code value, frame rate or user data.

New functions in the 24P/30P Progressive Mode

The DVXlOOA has been updated to allow some features to be used in progressive-scan mode which were previously disabled on the

8

original DVXlOO. Focus assist, Gain (+12dB max) and SMPTE Color bar display and output have been added. Focus Assist has been added to the Progressive modes (24P, 24PA, 30P, and 25P PAL). It

is slower than Auto Focus so it is called Focus Assist. Gain can be set up to 12db in the Progressive modes. SMPTE Color Bar Signal can now be used in the Progressive modes. The DVXIOOB retains all these features.

XLR Audio Input with +48 Volts Phantom Power Supply

In addition to a built-in stereo microphone, the DVX cameras are equipped with two XLR audio input connectors with switchable +48 volts phantom power. Both Input 1 and Input 2 can be switched between line and mic level by using the hard switches at the front of the camcorder. The audio limiter can be turned on and off, and the input mic level (-50dB / -60 dB) can be selected from the menu to ensure proper matching to an external MIC output or the environment. According to tests run in a leading magazine, the DVXIOO has a noticeably better audio recording system than any of the competing prosumer cameras on the market when the review came out (DV Magazine, November 2002 and December 2002 reviews).

In order to properly control this camera for generating the most powerful video images, one should understand camera process terminology. Experienced camera operators know there are always tradeoffs in every shooting situation. You'll trade shutter speeds for depth of focus, resolution for sensitivity, detail for noise, and the list goes on. The DVX's extensive menu system will provide the camera operator with additional tools for more creative solutions for image capture. It is our hope that by providing this book, you will gain confidence in addressing the menus and operation of the camera. Once you finish reading this guidebook, and the camera's manual, and watch the DVD material that accompanies this guide, you should have a thorough understanding of what the camera can do, what it's capable of, and how to extract the most from it under any given situation.

9

Improvements in the AG-DVXIOOA

In iust over one years' time from the AG-DVXIOO hitting store sh~lves, Panasonic turned around and introduced the AG-DVXlOOA with 26 changes and improvements.

The following chart highlights the changes; they will be explained in the following pages.

Interlace and Progressive modes

Interlace and Progressive modes

Color Bar

16 x 9 Aspect Ratio

Letterbox and Digital modes

USER Buttons (9 to 11 .

MOD (Minimum Object Distance) for close focus

10

Perimeter Protector for Switch

30 seconds

Every 15 seconds

Lens Cap

EVF's eyepiece lens

Body Paint Color

11

> Bow These 26 New Improvements To the DVXIOOA Benefit The Cinematographer

1. Slow Shutter: This new feature allows the user to create motion blur effects (such as swish pans and dream sequences), and is also quite effective for shooting low-lit scenes. By slowing the shutter speed from 1148 down to 116 (for example), more light is captured by the CCD chipset and more motion blur is registered. A shutter speed of 116 lets in eight times as much light as 1148 does!

2. Electronic Viewfinder: Now you can change the viewfinder from color to black and white. It may help in easier focusing, especially when combined with the PeakinglEVF DTL function.

3. EVF DTL (also known as "Peaking"): This feature adds an edge enhancement around objects that are in focus in the Electronic ViewFinder (EVF), to assist in easier focusing. When the camera registers something as being in-focus, the viewfinder will draw a tiny white outline around those objects that are in focus.

4. Gamma Settings (4 to 7 choices): Additional Gamma Settings have been added, which are listed in the scene file menu. Low, Normal, High & CINE-LIKE are used to complement the Low, Normal, High & CINE-LIKE positions on the older DVXIOO so the user can match both old & new cameras together on a production. CINE-LIKE D is for a wider dynamic range that is more like film. CINE-LIKE V is good for video programming like sitcoms and reality shows. B.PRESS helps bring some contrast and "punch" back to low-contrast scenes.

5. Knee: The Knee settings Auto, Low, Mid, and High allow you to set the compression point of the highlights. This allows the user some control over the dynamic range, without changing the Gamma. When the whites are too high, many of the details will normally be washed out. By setting the knee value you can maintain some of the tonal value in the highlights, and control where the knee begins its work.

12

6. Color Matrix settings: An Enriched mode has been added, which increases the color level for more vibrant color saturation, which may come in handy for a drab cloudy day when colors tend to look washed out. Cine Matrix also adds color, although in a different way, designed to achieve a more filmic look. Fluorescent adds red to the flesh tone vector for office shots that can't be lit to offset the overly green light coming off of fluorescent lighting. Normal is the proper amount of color for "regular" video, for accurate color rendition.

7. Vertical Detail Level in Progressive: THIN detail level renders the highest vertical resolution, rendering thin lines with I line of detail, and is usually best used only for progressive monitor playback or going out to film; the newly added MID position yields a compromise between best vertical resolution while minimizing line twitter on an interlace monitor. THICK line detail uses thicker lines for compatibility with interlaced monitors and televisions; thicker lines reduce line twitter, at the expense of being lower in resolution. The new MID setting retains much of the resolution of THIN while giving some of the protection against interline flicker that THICK gives.

8. Camera Gain in Progressive: Gain in the progressive modes allows you to be able to capture scenes in lower light conditions. Gain isn't possible on the DVXIOO in progressive mode, but the greater processing power of the DVX100A gives it the ability to offer up to + 12dB of gain even in the progressive-scan modes.

9. Auto Focus Assist: In interlace (60i/50i) the auto focus system gets updated every 60th of a second (1I50th in PAL), and the Auto Focus function is fast and accurate. However in Progressive modes, the CCD runs slower (in 24P mode, the CCD is operating at 40% of the speed it does in 60i mode). Because the CCD is running slower, the auto focus system gets updates less frequently. Less information means slower response. Even though the focusing works

beautifully .. .it works more slowly. (It can be useful for a dramatic effect for some scenes because of the slow focus.) However because it is slower, Panasonic doesn't call it true Auto Focus; instead it is referred to as Focus Assist in the progressive modes. The DVXI00

13

didn't have any sort of autofocus in progressive modes; the addition of Focus Assist in the DVXI00A is a bonus.

10. Built-in SMPTE Color Bar: Color bars are now available in Progressive modes (24P & 30P, and 25P in PAL). On the DVXI00, no color bars could be generated in progressive mode, but in the DVXI00A they're accessible in all modes, either from enabling them in the menu, or by assigning them to one of the USER buttons.

11. 16: 9 Anamorphic: The Squeeze Mode is new on the DVXI00A. It basically uses the same amount of the image as the Letterbox mode, but it changes the aspect ratio ofthe video, making 16:9-shaped images that will display properly on a widescreen television, and will be recognized as 16:9 widescreen images by non-linear editing programs. (The EVF and LCD will show a squeezed, "skinny" display, although the full actual image content is represented.)

12. USER Buttons (from 9 to 11 choices): There are now 3 USER buttons and the functions that can be assigned to them have increased from 9 to 11 by adding INDEX and SLOW SHUTTER settings. This can be very powerful for customizing the DVX100A to the user's needs. White or Black fade [linked with audio] can be handy for event videography, and the Mode Check setting can replace the Mode Check button when the LCD panel is closed.

13. New Prism: A new prism is responsible for better color reproduction and cleaner edges.

14. Minimum Object Distance (MOD): The minimum close-focus point has been reduced from one meter in the DVXI00 to.6 meters in the DVXI00A. This is a great feature. In wide angle, Macro focusing allows for focusing right up to the front of the lens, but in telephoto mode on the DVXI00 it was limited to 1 Meter. Now you can focus at full telephoto all the way down to .6 meters (in inches and feet terms, the new MOD is approximately 20 inches instead of the old MOD of approximately 39 inches.) This lets you get the camera closer, and zoom in more, giving a better macro ability as well as the ability to get shallower depth of field.

14

15. Zebra Settings (2): The ZEBRA settings help monitor exposure, and the DVX100 measured in levels from 80 to 100 IRE. The DVX100A adds a new 105% setting, making it easier to know with more assurance when highlights are approaching the clipping level.

16. Scene File Dial (Cine~witchTM): A perimeter wall has been added to the Scene File Dial to help prevent unintentional changes while shooting. Another improvement is that there's no longer a processing delay when switching the scene files from interlaced to progressive modes (as there was in the DVXI00.)

17. Focus/Zoom, Improved Tactile Feel: Panasonic has increased the dampening of the focus and zoom to give the cinematographer better control on the lens focus and manual zoom. Smoother focus and smoother manual zooms are now possible.

18. Power Zoom Speed (Slower): The slowest zoom speed on the power zoom system has been slowed down even more. It now takes 30 seconds to travel the full range of the zoom system, whereas on the DVXI00 the slowest speed took just 20 seconds to go from full wide-angle to full telephoto. This slower zoom speed can give the cinematographer a more dramatic look, a slower "creep zoom", more like that found on an expensive professional lens found on a shoulder mount camera.

19. Interval Recording Time Minimum: The shortest interval time (for intermittent recording/time lapse) has been shortened to 1I15th of a second.

20. Color Level, Brightness & Contrast Adjustment: You can now adjust the Color Level, Brightness and Contrast levels independently for the viewfmder and flip-out LCD panel.

21. Headphone Output Level: The audio level has been raised 6 dB to provide more volume to the headphones (that means the headphone output is twice as loud on a DVX100A as compared to a DVX100). Additionally the volume +/- controls are now active in Camera mode as well as VTR playback mode. The "glow in the dark" Audio recording volume controls still govern the strength of the audio

15

signals being recorded on the tape; the headphone volume controls are just monitoring controls.

-,

22. Marker Zone: The marker zone now has a smaller spot zone and a larger marker range. This gives you more accurate f~edback over your scene's exposure. MARKER now goes to a maximum of99 (rather than 90+ on the DVXI00) and the sensor box is the smaller box (more accurate) on the left within the larger box.

23. Lens Cap: The lens cap has been redesigned so that the storage clip doesn't protrude, so it will not touch filters mounte~ on th~ lens. Finger grab tabs have been enlarged (deepened) to provide easier removal and mounting.

24. Electronic Viewfinder (EVF)'s Eyepiece Lens: A specially designed magnifier lens has been added, ~o give the viewfinder be~er accuracy, planar view and it acts almost like a sports finder. Also It has a more scratch resistant coating.

25. Body Paint Color: The camcorder's body is finished i~ Diamond Graphite and black, a more professional color, which also produces less reflection.

26. DV Tape Included: Now you have a 63 min. blank vide~tape to get you started making images and learning instead of a cleanmg tape. (Of course, you should also buy a cleaning tape as we1l!)

Other Differences Not in the Chart

The AID Conversion circuit has been changed from 10 bit to 12 bit. This converter is the same converter used in some broadcast camcorders. This 12 Bit Digital Signal Processing allows for more precisely defined colors and tonal ranges.

16

Improvements in the AG-DVXIOOB

Panasonic has again updated their popular AG-DVX100A camera. There are over 22 changes in the AG-DVX100B that improve upon the AG-DVX100A.

Focus is now remotecontrollable

Iris is now remotecontrollable

Remote iris control

Images are cleaner with less video "noise"

Twice as sensitive

Can now sync timecode the 1394

17

How These 22 New Improvements To The DVXIOOB Benefit The Cinematographer

/

1. Viewfinder Aspect Ratio: The camera can now display 16:9 footage in a proper 16:9 aspect ratio. This helps users to properly frame their Squeeze mode or Anamorphic Adapter shots.

2. Viewfinder and LCD Display Area: The Viewfinder and LCD now show the full frame. All prior under-$5,OOO cameras show only the ''underscan'' frame, which doesn't display the full frame. This camera shows the entire frame from edge to edge.

3 and 4. EVF and LCD Resolution: The Electronic Viewfinder (EVF) and the LCD panel have both been substantially boosted in resolution. This helps them resolve the full underscan frame, as well as helping the operator achieve more accurate focus.

5. Remote Focus Control: The camera can now use an optional remote focus controller. This is exceptionally useful for circumstances where the camera is beyond the camera operator's reach (such as on a jib arm, or mounted on a car mount, etc).

6. Remote Iris Control: In another "first" for an under-$5,OOO camera, the DVX100B allows the use of an optional remote iris controL This allows the camera operator to control exposure remotely. The remote focus and remote iris control share the same connector jack; accessory manufacturers will develop combined iris/focus controllers for use with the DVX100B.

7. Headphone Monitoring: The DVX100B offers the ability to select "live" or "tape" mode for the headphone jack. On a DVXlOOA (which only allowed "tape" mode), when monitoring 24P or 25P footage, the audio would exhibit some "echo" unless the operator used noise-isolating headphones. On the DVXlOOB, monitoring audio in "live" mode totally eliminates that echo.

8. LCD Panel Rotation: The DVX100A LCD panel opens up to a 90-degree angle. The DVX100B LCD panel can open even further;

18

to 120 degrees. This makes the LCD panel even easier to position, giving the operator more flexibility, as well as making it easier for other people on the set to see. Additionally, it helps protect against accidental damage if someone were to bump into the LCD panel.

9. Improved Video Sig~al-To-Noise Ratio: The DVXI00B image processing has been updated to provide a "cleaner" image, with less video noise than either the DVXI00A or DVX100.

10. Built-in Microphone Sensitivity: the onboard stereo microphone on the DVXI00B is now twice as sensitive as the microphone on the DVX100A and DVXI00. This makes getting strong audio recording levels easier on the new DVX100B camera.

11. Jam-Sync Time Code Input: The DVX100B becomes the first under-$5,000 camera to allow direct synchronization of Time Code to another DV camera. This allows shooters in multi-camera shoots to precisely synchronize the time code between two or more cameras.

12. Transfer Scene Files: The DVX100B allows users to transfer scene files and user files between cameras, by using the 1394 cable. This makes matching multiple cameras instantaneous and foolproof.

13. Power Zoom Speed: The DVX series of cameras offer three zoom speeds (low, middle, and high). On the DVXI00B, the middle zoom position has been slowed down somewhat; from 8 seconds to 10. This new slower zoom speed is a better "middle ground" between the 30-second slowest zoom speed, and the 3-second fastest speed.

14. Disp Off Button: The new Display Offbutton allows you to quickly hide all on-screen display info, letting you see the full frame free of any readouts or displays. Pressing the button again brings all the displays back. This lets you quickly check the full frame and composition.

15. End Search in Camera Mode: you can now search to the end of the tape when in camera mode (DVX100 and DVXI00A could only perform an End Search when in VCR mode). This means quicker access to end-searching without having to swap camera modes.

19

16. Speaker Placement: The speaker on the DVX100A is located inside the LCD panel; on the DVXI00B it's been relocated outside to the top of the camera. This new speaker placement makes monitoring playback much easier, especially for groups of viewers.

17. Body Paint Color: The camcorder's body is finished in Black Sapphire, lending an even more professional look. The color is also less reflective than the prior cameras, and hides fingerprints.

18. Tripod Socket: The DVXI00B features a newly-engineered extra-heavy-duty tripod socket plate. This tripod socket is more resistant to damage, and can be easily replaced.

19. Tape Door: The latches to the tape compartment door have been significantly reinforced and "ruggedized." Since the tape door is attached to the handgrip, some users found that using the camera handheld (and supported only in the right hand) could occasionally interfere with the tape recording. The heavy-duty door latches in the DVXI00B prevent that from ever happening.

20. Menu System: The menu system has been given a design makeover. Menu options are now grouped in colored boxes that clearly show the available options and make menu selections easier to navigate. Some menu items have been moved to present a more logical grouping of menu functions, as well as to rename some items to make them easier to understand (such as "POWER OFF" instead of "TAPE PROTECT").

21. Included Battery: The DVX100B now includes the CGR-D54 battery, a much longer-lasting battery than the CGR-DI6 that was included with the DVX100 and DVX100A. The CGR-D54 has 3.5 times as much capacity as the original battery, and will allow for over five hours of continuous operation.

22. Included Tape: The DVXI00B includes a head-cleaning tape instead of a blank tape. A head-cleaning tape is a valuable thing to keep in your equipment bag - if you ever see the red "flashing X" warning sign, that means you need to run the head cleaning tape.

20

DVXIOO, DVXIOOA and DVXIOOB Articles

,

Frequently Asked Questions, and Commentary

WHAT'S THAT CLICK?

The first question most new DVX owners have is "What's that cl~ck?" When a new owner picks up the DVX for the first time, they will usually hear a click or "rattle" when tipping the camera back and forth, like there is something loose inside the camera. To reassure the new .use~, there's nothing wrong with the camera. What you're heanng IS an element of the optical image stabilizing system. When the camera is powered on, the OIS system is energized, and the rattle dis~ppears. This is one of the most frequently asked DVX questions, so If you were con~emed.b~ the "rattle", have no fear, you're in very good company! Since this IS one of the first questions DVX users ask, we thought it only appropriate that it be the first answer in the guide!

STROBING

When new users pick up the DVX for the first time, they switch it o~er to 24P/25P mode, wave the camera around, then say "hey, wait a mmute, what's all this strobing?" The motion in their shots appears very "jerky," and not smooth. For those who are unaccustomed to sh~oting film, the look is unfamiliar and unsettling; many times they think that there may be something wrong with their camera.

There's not~ing wrong with the camera! What's happening is a perceptual difference between what the shooter is familiar with (interlaced video) and what they're now using (progressive-scan vid~o!. The~e's a very different look between 24P (or 25P) and traditional VIdeo, also known as "60i" (or "50i" in PAL). In 24P, the car_ne~a shoots twenty-four frames every second. In regular interlaced 601 VIdeo, the camera shoots sixty "half-frames" per second. The difference is dramatic: the 24p footage looks more like a movie and

60i video looks more like a soap opera. '

21

But shooting only 24 frames per second has a side effect called "strobing". Pan the camera, or move it side-to-side too fast, and you might see choppy, stuttery movement. Strobing can be unpleasant to watch. Film runs at 24 frames per second, and film exhibits the same strobing issues that the DVX does. Fortunately for us, film cinematographers have been fighting this issue for almost 100 years, and over the years they've developed some ways to combat the strobing effect.

First, pan the camera slowly.

The chart below is an adaptation from the American Cinematographer's Manual, a publication of the American Society of Cinematographers. It's been adapted to reflect the focal lengths of the DVX lens, and shows just how slowly you need to move the camera to execute a 90-degree pan, depending on what focal length you have the lens set at. The general rule of thumb is to make sure that an object takes at least seven seconds to cross the screen. That's really slow, but it's what you need to do if you want smooth motion.

ZOOM

24P PAN SPEED 30P PAN SPEED

zoo 15 seconds 11 seconds
Z17 18 seconds 14 seconds
Z25 21 seconds 16 seconds
Z35 23 seconds 20 seconds
Z55 36 seconds 27 seconds
Z65 50 seconds 38 seconds
Z72 60 seconds 45 seconds
Z83 80 seconds 60 seconds
Z88 90 seconds 75 seconds
Z99 120 seconds 120 seconds To read that chart properly, what it's saying is, if you're at Z25, and you want to pan 90 degrees, with no juddering or stutter, and you're shooting 24P, you have to take 21 seconds to execute the pan. If you take the full 21 seconds, the pan will be glass smooth. Any faster than that, and smoothness will suffer and strobing will begin. That's

for shooting 24P (or 25P). You can go faster if you're shooting 30P;...,. /

22

at Z25, you can pan the 90 degrees in 16 seconds at 30P and it'll still be smooth. When shooting interlaced, you can pan at any speed with no restrictions, but when shooting progressive you have to carefully monitor how fast you pan.

If you're at maximum telephoto, things change dramatically: you have to go much, much slower. If you want a glass-smooth 90-degree pan, it will take two minutes to execute that pan at 24P. Any faster than that and the pan will stutter.

Right now you're probably thinking "that's outrageous - nobody pans that slowly!" Those are the panning rates, as established by the combined experience of the American Society of Cinematographers. Having shot film for the last 100 years they've figured out the speeds you need to stick to if you want smooth pans. When people say "pan slower", they're not kidding!

Obviously those speeds are very restrictive. If you choose to go faster, you just need to determine what level of stutterljudder is acceptable to you. Fortunately with the DVX, you can hook up' an external monitor and see exactly what it looks like (with film, they had to wait until it was developed and projected, which is why they took the effort to devise the chart in the first place).

Keep in mind that the LCD panel appears to show the strobing a bit more exaggerated than it looks on an NTSC monitor. To truly judge the speed of your pans and the amount of strobing in your scene, you should view the scene on an interlaced television: that will show you what your video "really" looks like, and then you can make the most informed decision as to how quickly or slowly you need to pan. Also, the quality of the television isn't really much of a determining factor: as long as it's an NTSC (or PAL) compatible set, it'll render the motion properly, so you could use a cheap portable television if that's all you can afford, or for convenience for use in the field.

Second, follow a screen-stationary object.

The other rule to minimize strobing is that you can move the camera much quicker if you put something stationary (relative to the screen) in the foreground for the eye to follow. If you're following a person

23

or a car or some other foreground object, as long as y?U keep that object relatively stationary on t~e screen, ~ou can basically pc:m as

fast as you want. Having a stationary subject on the scr~en grves the viewer'S eye something to focus on other than the strobing background. The background's strobi~gjust as m_uch, but the viewer doesn't notice it because they're watching the subject. To see an example of this, be sure to watch the strobing segment on the

included DVD. Once you see how this technique works, you'll start noticing it all the time in movies and film-shot television show.

Why does the motion strobe? To understan~ this, you ~ave to . understand the difference between 24p and interlaced VIdeo. WIth 24P/25P, the camera's shutter is only open for half the time, but in 50il60i the shutter is always open. In 60i the camera shoots 60 fields per second, and each field is exposed for 1I~0 of a ~econd, so ~e camera's eye is always open, always recording motion. There s

never a time when the camera is "blind" to the action. But with 24P, the camera captures 24 frames per second, and each frame is exposed for half that time, or approximately 1/48 of a second. This means that the camera's eye is 'shut' half the time! In 60i it's always open, but in 24P it's open/closed, open/closed. So the motion "strob~s" ~ust like a strobelight. If you've had the chance to observe motion m a room that's lit only by a strobelight, you'll recognize the choppy, stuttery movement we're talking about (except obvious~y i~'s quite exaggerated in the strobelight example; however the principle remains the same.)

Film works the same way. Film cameras have a rotating disc "shutter". The shutter is an opaque disc with a wedge of it "cut out". This wedge is where the film gets expose~; on film cameras.the shutter opening is usually about half the dISC, or (expressed m terms of a circle) 180 degrees. As film is shot, the shutter rotates, alternately covering-up or revealing the film frame. When the shutter

". is closed the film advances to the next frame, and then the film gets held stationary while the shutter is open. So film works on th~s same principle of "open/closed". Fully half the time a film camera I~ running, it's "blind" to the action! That's why motion strobes in film, and why it also strobes in the DVX.

24

So, is this strobing a good thing or a bad thing? Obviously it makes motion more choppy, but if you want your footage to look like film, it's a necessary effect. Although some people say that film doesn't strobe like the DVX does, we decided to find out. On the DVD that accompanies this guide, you'll fmd a comparison where we strapped a DVX to a 16mm camera and mounted them both on a tripod, and shot the exact same scene side-by-side. We were able to demonstrate that the DVX renders motion exactly the same way film does (when that film has been transferred to video). What this means for you is, if you find the strobing in your footage objectionable, the camera's not at fault, it's doing exactly what it's supposed to do. You need to adapt your technique to follow the American Society of Cinematographer rules for shooting film: slow down your pans, and follow a subject. Disobey the rules, and your footage WILL strobe, exactly like film does.

Lessening The Strobing Effect

There are two ways you can lessen the strobing effect. First, using a longer shutter speed will give you a little more motion blur. Film cameras typically use shutter speeds anywhere between 1140 and t"/60 of a second. Faster shutter speeds mean sharper footage and more strobing. Slower shutter speeds mean more motion blur and slightly less strobing. On the DVXIOO, you could select a shutter speed as slow as 1124 in 24P mode, or you could go into the SYNCRO SCAN menu and select any shutter speed between 1124.1 and 1/250.0. On the DVXIOOA and B, provision is made for even slower shutter speeds, including 1112 and 116. Shutter speeds that slow will result in extreme motion blur and a loss offilmlike motion rendition. To keep a realistic simulation of what film looks like, while getting a little better low-light performance and a little more motion blur, stick to a speed no slower than about 1130.0 or 1136.0. The more motion blur you have, the less noticeable the strobing will be. Strobing

practically disappears at 1124, because then the camera is operating more like in interlaced mode: shooting 24 frames for 1124 of a

second, so the camera's eye is always "open". The motion's still not as smooth as interlaced mode because of the presence of2:3 pulldown, but shooting at 1124 would eliminate almost all strobing. If you have a certain shot in mind, which demands a certain panning speed, and you find the strobing unacceptable at that speed, you can

25

.e.onsI,oer using a slower shutter speed to get the shot while not }.,i,.,vp.TwhelIlm·mg the viewer with strobing; the tradeoff will be that

're getting blurrier video, so be sure to test the technique on a television to make sure that the blur/strobe compromise will deliver the look you're after.

The second thing you can do to minimize strobing is increase your frame rate. Shooting at 30P will get 25% more frames per second, which will mean less strobing. The faster the frame rate, the less

, strobing. 30P footage will look a little more like video, but still ~~ok sort offilmlike. 30P can't be transferred to film or to PAL television, but if you're bothered by the strobing in 24P, you may find 30P more to your liking. The PAL version of the camera ~oes~'t have ~ equivalent middle speed setting, however, so this middle settmg wouldn't be an option for PAL shooters.

Or, you can always shoot in interlaced mode at 50i~60i and completely eliminate all strobing, but your shots will look completely like video instead of film. But the high speed of the interlaced frame rate will mean that there's no strobing whatsoever. The choice is yours, but unfortunately you can't have it all: if you want the filI? look; you have to work to control the strobing, and if you shoot m interlace mode you don't have to even worry about strobing at all, but the footage will look like video instead of film. Some people have tried to shoot in interlaced mode and then use aftermarket film-look plug-ins that convert interlaced footage into a simulation of 24P, and find to their dismay that yes, their footage starts strobing again. Strobing is inherent in the slower capture rate. You can't avoid it; all you can do is intelligently manage your circumstances (and heed the wisdom of 100 years of cinematographer experience!) to control the strobing.

26

AUDIO DELAY IN 24P MODE, and ECHO IN DVXIOOA HEADPHONES

On the original DVXIOO there is a slight audio sync offset that can be observed when shooting 24P footage. The audio gets recorded on the tape a full two frames ahead of the video. The audio stays in perfect sync with the video (there's no drift, for example) but it is recorded two frames earlier than the video. This sync issue is not very noticeable; in fact the camera was on the market for six full months before anyone even mentioned it on Internet discussion groups. The PAL version of the camera has a smaller sync offset, but it does still exist. If the sync delay bothers you, or if you absolutely need to have perfect sync, the delay can easily be addressed in post by disconnecting the audio from the video and sliding the audio two frames to the right on the timeline. Some programs automatically resync the audio for you; Apple's Final Cut ProTM has an option for correcting the audio sync, as does Avid Express™, and DVFilm.com's Maker= software can also re-sync the audio.

The reason the audio is not lined up with the video is because of how 24P mode works: the camera buffers frames, waiting to have eriough information so it can write the frames to tape inserting 2:3 pulldown. But the audio gets written to tape immediately. So the audio precedes the video by two full frames (the duration of the frame buffer for building the 2:3 pulldown).

In the DVXI00A, Panasonic implemented an audio delay circuit in 24P mode. The audio gets buffered and delayed for 2 frames so it gets written to the tape exactly in sync with the video. The result is perfect sync on tape (or through the 1394 port). An unfortunate side effect of this is that the audio delay can be heard through the headphone jack: if you don't have perfect isolation with your headphones, (which means that you can hear real-world sound as well as the sound coming through the headphone jack), you'll hear what sounds like an "echo". This is caused by the sound in the headphone jack not playing back in "real time;" instead it's buffered and delayed by two frames to match the video being recorded on the tape. You're hearing the live audio "live," and the camcorder audio delayed, and the combination makes an echo. There's nothing wrong with the camera when you hear this, and the echo is not being recorded on the

27

L

tape; the echo effect comes from the audio buffer, combined with having headphones that allow "real world" sound to leak in. You won't hear this delay when shooting in interlaced mode, since in interlaced mode neither the audio nor the video gets buffered before being written to tape. But in 24P mode (and, to a lesser extent, in 25P PAL and 30P mode) you'll probably hear some echo.

Unfortunately, on the DVXlOOA there's not a lot that can be done about it. The only way to eliminate the echo would be to use completely noise-isolating headphones, which would pre~ent you from hearing any "real world" sound, only the sound coming from the camcorder through the headphones jack. If you had completely noise-isolating headphones, you wouldn't hear any echo whatsoever.

On the DVXlOOB, Panasonic has implemented a new menu item that allows the user to select whether the headphone jack outputs the audio "live", or from tape. Choosing "tape" will result in the same echo as the DVXIOOA exhibits. Choosing "live" completely eliminates the echo. It's important to note that both options result in perfect sync on the tape (i.e., choosing "live" doesn't return you ~o the same sync offset situation as the original DVXlOO; your audio win still be perfectly in sync on tape). I heartily recommend setting the H.P. MODE menu setting to "LIVE".

RECOMMENDED F-STOP RANGE

A well-known factor in lens design is that lenses typically perform at their sharpest in a certain f-stop range, usually a range from about two stops down from wide-open to two stops up from fully-closed (on many lenses this would mean a range of around fl4 to fill). Lenses typically begin to lose some sharpness at their widest-open and mostclosed apertures. A further issue that confounds small-sensor cameras is the issue of diffraction: the propensity of light to scatter when forced through a too-small aperture, which makes it appear like the image is out of focus. The Canon XLI had serious soft-f~c~s issues at fl32, which were traced to diffraction. Is the DVX SImIlarly affected?

28

We tested the DVX lens for diffraction and to find the f-stop range where the lens is sharpest. We used a high-resolution monitor and a resolution chart, lighting the chart at different light levels so we could test it at proper exposure using different apertures on the camera.

Keeping in conjunction with prevailing wisdom about lens design, fl5.6 was definitely the best, delivering clear sharp lines up to the 540 mark, which is the theoretical limit ofDV (the DV format can record an absolute maximum of540 TV Lines of resolution). F/5.6 represents two stops down from wide-open, and the DVX lens performs sharpest in a band from about fl4.8 to flll.O.

fl16 is just a tiny bit lower resolution, resolving cleanly up to about 520 lines. There is no substantial effect on resolution due to diffraction between fl4.8 to "true" fl16.

At wide-open apertures, resolution loss does become somewhat more of an issue. At fully open, the resolution takes a hit, going to about 470 lines. That's a significant drop (about 15% less resolution).

The most diffraction comes at the smallest apertures (as expected). The DVX iris reads out to fl16, but the lens can be closed down by two increments past fl16, just before the iris closes completely. At one tick above CLOSE, resolution drops off to about 270 lines. That's a dramatic drop-off, only about half the potential resolution of the camera, and primarily attributable to diffraction. For this reason, we recommend that you try to avoid fl16, since it's very difficult to see the "ticks" of gradation, so you may not know if you're at "true" fl16, or fl16 plus one "tick", or fl16 plus two "ticks". For maximum image sharpness, it's best to stay at fl14 for your smallest aperture.

For optimal image sharpness and resolution, stick to apertures between fl2.8 and fl14. At iris levels much wider than fl2.8 you may lose a little resolution, maybe 10 to 15%. But avoid really closing the lens down -- pure fl16 is okay, but you can get some smaller apertures that still say fl16 that suffer from extreme resolution loss and extreme image softness. Stay between fl2.8 and fl14 and diffraction will not be an issue.

29

CQ]nITlOn question is "why doesn't my 5-hour battery last five hours "" "~ .. my DVX?" Panasonic made the batte?es cross-compatibl~ ""l..,,,,hup.p.n most of their consumer camera line and the DVX senes. \You can buy batteries that are marketed for the consum~r cameras,

." d use them in the DVX and they'll work just fine. ThIS also means an ·1 d "bku " that if you buy a small Panasomc pa mcor er as a ac p camera ,

you could share batteries (and battery chargers and remote controls) between it and your DVX. (Note: the new "GS" series ofPanasomc camcorders use a different battery design.)

But the DVX cameras are rather power-thirsty. The DVX draws a lot more wattage than a small palmcorder. So when you see a battery that advertises "5-hour" or "9-hour" life, what they mean is that the battery would probably last that long in a tiny palmcorder, but it won't last nearly that long in the big thirsty DVX. As a rule of. thumb the battery capacity is a guide to how long the battery will last. For example, Panasonic batteries are named according to th~ir milli-amp-hour (mah) capacity, A CGR-D28 battery has a capacity of 2800 mah. A CGR-D54 has a capacity of 5400 mah. Those numbers correspond (roughly) to how long the battery will last in a DVX camera: a 2800 mah battery lasts for approximately 2.8 hours, and a 5400 mah battery will last for about 5.4 hours. So ignore the "5-hour" or "9-hour" ratings, and instead look to the battery's actual capacity to get an idea of how long the battery will last in your DVX.

THE DVX ASA/ISO RATING

When using an external light meter to meter your scenes, it's . necessary to know what the equivalent "film speed" the camera IS operating at. For instance, when using a light meter with 500 ASA film, you set the meter to 500 ASA and the appropriate shutter speed, meter the scene, and the light meter tells you exactly what f-stop to set on the camera for proper exposure. If you knew the "ASA" of the camera, you could use an external light meter with the DVX and get similar results.

30

That's the theory, anyway. In practice it's quite difficult to judge the actual "speed" of the camera, because it depends on several factors. We have tested the DVX thoroughly under all light conditions, with various settings, and the general response is that the DVX is roughly 640 ASA in progressive mode, and roughly 1000 ASA in interlaced mode. However, it's not really as simple as that, because the speed rating changes on several factors.

All testing was done on a DVXI00 at 1160 and 111000 shutter speeds, with A. IRIS compensation set to zero, and all other menu settings set to zero. Testing was performed using a gray card and also using reallife scenes, using the DVX's auto-iris to determine exposure for the camera as tested against a Sekonic L-508C lightmeter in spotmeter mode, and using a waveform monitor to gauge the camera's response. Testing was accurate to within 116 of an f-stop on the DVX, and 1110 of an f-stop on the Sekonic meter. Testing was cross-referenced against a Minolta Spotmeter and the results were confirmed.

First, if the camera is set in interlaced mode, it's approximately one stop more sensitive. For shooting in low light conditions, that's an important consideration - the DVX will deliver a picture twice as bright in interlaced mode than it would in progressive mode at an equivalent shutter speed (or, put another way, the DVX only needs half as much light when in interlaced mode to produce the same picture brightness as it does when in progressive mode.)

Second, the exact ASAJISO response depends on what you're shooting and how brightly it is lit. Testing against a general mixedimage scene delivers a slightly different response than testing against a fixed gray card. On a gray card the results varied depending on the overall light level in the scene: at low light levels, low enough for the camera to recommend an fl2.8, the light meter showed an equivalency to 400 ASA in progressive mode, and 800 ASA in interlaced mode. In a very bright scene, bright enough for the camera's auto-iris to recommend an fl16, the spotmeter reported that in progressive mode the DVX was operating at 800 ASA, and in interlaced mode it was an amazing 1280 ASA. In real-world conditions, shooting a variety of subjects, the results were more consistent at varying light levels, usually indicating 640 ASA in

31

... ·ft1~ogreSSIV'e and 1280 ASA in interlaced. The lesson to learn here is the DVX's exposure curve is not linear, but the camera is more

, •. responsive at brighter light levels,

Finally, the gamma setting can change the exposure response ', In testing against a gray card in progressive mode, ~der very bnght (:t716) conditions, under NORMAL gamma we calibrated the camera at 800 ASA. In CINE-GAMMA it corresponded to 1000 ASA.

. In general, if you were to "rate" the camera at 640 A~A .in

progressive, and 1000 ASA in interlaced, you'd be within about a half-stop of being accurate at either end of the exposure spectrum

(i.e., the difference between 400 ASA and 640 ASA is about 112 of a stop, and from 640 ASA to 800 ASA is ~bout 114 of a st~p.)

However, our recommendation is that USIng an external light m~ter as your sole judge of exposure is not necessarily an acc~ate technique. Using an external light meter can give you a general Idea of what the exposure should be, but because the .camera responds d~ffere~t!y at different levels of light, and under different gamma settings, It s . much more accurate (and reliable) to use the camera's zebra settmgs to determine overexposure, a professional (and properly calibrated) CRT monitor to judge overall exposure, and the MARKE_R to finetune exposure. If you wanted to invest in an external device to help you judge exposure, a waveform monitor would be a .far more accurate tool for use with the DVX than an external light meter would be. One of my personal favorite software programs is ~erious Magic's DV Rack=, which gives you a waveform monitor, . vectorscope, production monitor, and digital disk recorder all In one; DV Rack can be very helpful in determining proper exposure.

24PVS24PA

(This section is not applicable to PAL cameras; only NTSC cameras

have 24P and 24PA recording systems and pulldown.)

Panasonic introduced 24P shooting to the DV camera world. By definition DV as a video format doesn't make any provision for 24P. Panasonic engineers cleverly found a way to shoot 24 frames per

32

second, but record on tape in DV -standard interlaced video mode, thus ensuring that DVX recordings (even 24P recordings) can be played back on any DV deck or camera. Even though 24P is not part of the DV format specifications, Panasonic was able to bring fully DV-compliant 24P recording into being.

Panasonicemploys the same 2:3 pulldown sequence as that which is used when film is transferred to video. Film runs at 24 frames per second, just like 24P video does. And video runs at 60 fields per second. In order to transfer film to video, engineers long ago came up with the "2:3 pulldown sequence", which prints one frame of film onto two fields of video, and the next frame of film onto three fields of video, The pattern then repeats. Using this pattern, every two frames of film get printed onto five fields of video, and this pattern happens 12 times per second, so 24 frames of film (two frames x 12) get printed to 60 fields of video (five fields x 12). Here's a pictorial example that shows how the frames get recorded to video, and how they get interlaced. First, assume that we have four frames of video, which look like this:

When those frames get recorded on videotape, the DVX will split each frame into "fields" and interleave them sequentially on the tape, like this:

In our 2:3 example, one frame will get written to tape as two video fields, and the next frame will get written out to three video fields.

33

So every other frame is written out as three video fields. In our example, frames "A" and "C" will be written for two fields, and frames "B" and "D" will take up three fields. How this looks on tape is, frame "A" gets written out to two video fields, and since each pair offields is treated as a "frame", frame A gets written out to tape intact. Frame "B" gets printed to three video fields. The first two fields can be thought of as a distinct frame, and the third field gets combined with one of the fields from frame "C", resulting in what's called a "split frame". Frame "C" will be written out to two fields, so the first field is combined into a split frame with the third field from frame "B", and the second field will be combined into a split frame with frame "D". Frame "D" gets written out as three fields, so one field is used as a combined split frame with a field from frame "C", and then the frame gets written out to two more fields, which result in a complete, intact frame.

In 24P mode, the DVX CCD actually runs at 24Hz, capturing a new and distinct frame 24 times per second (some other cameras, such as certain Sony models, actually run at 60Hz and try to "simulate" the 24P look in-camera, but the DVX actually records at a genuine 24fps frame rate). However, the DV recording standard requires 60 interlaced fields per second to be recorded to tape, so the DVX slices each frame into fields, and performs in-camera pull down to transform 24 frames into 60 fields. The result is video that looks like it was

shot on film. The frames look "right" when displayed on an interlaced television. However, in order to edit them in a progressivescan timeline, the "split frames" will need to be un-split, and properly recombined with their proper field-mate, in order to reconstitute the original 24-frame sequence of "A BCD". This re-combining

process requires an uncompress/recompress cycle in the editing program and can result in a small loss of quality. However, if editing in a 60i timeline, the video needs no processing and the 2:3 pulldown system will result in the smoothest 24P motion on an NTSC TV.

For higher quality, Panasonic developed a new pulldown method, which is used in the "24P Advanced" recording mode. This new method is called 2:3:3:2 pulldown. In 2:3:3:2 pulldown, one .frame gets printed to two video fields, and the second frame gets pnnted to

34

three video fields. The third frame also gets printed to three video fields, and the fourth frame gets printed to two video fields.

The end result is th~t for every group of four 24P frames, they get recorded to four "video frames", with one additional "split frame" crea~ed to round out the sequence. Four 24P frames get printed onto 10 VIdeo fields, and this happens six times per second so 24 frames (four x 6) get printed onto 60 video fields (10 x 6). The difference is the cadence is not quite as smooth as in 2:3 pulldown, so 24PA ' footage may look just slightly more "stuttery" or 'jerky" than 24P Normal footage on NTSC TV. However, the 24PA footage can be reconstituted to its original 24 distinct frames easier and with no ~compressionl recompression, something that's not as easy to do WI.~ 24P Normal ~ootage. As you can see in the example, all of the original frames .e~ust in their original un-blended, un-compromised fo~; all th~ editing program needs to do is discard the "split frame" and It can directly edit on the original, pure progressive frames.

The DV compression algorithm includes two compression t~chniques: when it detects significant amounts of temporal d~fference between the fields of a video "frame" (a grouping of two VIdeo fields is called a "frame", even though it bears little resemblance to a film frame or a 24P frame) it will employ fieldb~sed com~res~io~, but when it doesn't detect much temporal difference, It WIll instead employ frame-based compression. When 24P Normal footage is written to tape, the 24P frames get split into fields and blended together: the odd field from one frame may be combined with the even field from the next frame and then written to tape, resulting in a "split frame". This means that the field-based

35

compression algorithm is more likely to be used on a split frame. When 24P Advanced footage is written to tape, all the original 24P frames are kept intact: for every group of four 24P frames written out they are written out one-for-one onto video "frames", and then a fifth' "split frame" is also created. The original24P frames, having been kept in their original state, can be compressed using the frame-based compression algorithm, resulting in more efficient compression. The fifth frame, the "split frame", will probably use field-based

_ compression, but it's really irrelevant, because when the editing program imports this footage it can take the four original frames and simply discard the split frame. This is not possible with 24P Normal footage, where the footage needs to be decompressed and have the fields re-combined into their proper order. 24P Normal is therefore a little less efficient for editing in a 24P timeline than 24P Advanced, and 24P Advanced may benefit from better compression efficiency.

In the end, there's not really much of a difference between the two. 24P Normal will look a little tiny bit smoother on playback on a 60i timeline than 24P Advanced would. 24P Advanced will probably have a tiny bit better compression, maybe a tiny bit fewer artifacts. The main difference as to when to choose which mode is determined by how you intend to edit the footage: if you are going to edit it on a 24P timeline, it's better to shoot in 24P Advanced. If you're going to edit it on a 60i timeline, it's better to shoot 24P Normal. However, you can easily edit 24P A footage on a 60i timeline, and 24PN footage on a 24P timeline. The differences are minor, so if you choose one or the other you're not necessarily going to lock yourself out of other options, but if you know in advance how you intend to edit the program, you can get a little better quality or a little smoother footage by choosing the appropriate pulldown sequence.

As for 30P, there is no pulldown involved. Each frame is written directly to two video fields, and frame-based compression is employed. In PAL mode, there is also no pulldown: PAL 25P mode works exactly like NTSC 30P mode, where each PAL frame is written to two video fields.

36

FRAME RATES AND SLOW MOTION

The DVX is the first prosumer camera to offer a shooting speed of 24P, but it also can shoot at 30P as well as 60i and, with a little manipulation in post, you can get 12P and 15P too. These frame rates give you flexibility that's previously only been available to film shooters. For PAL users, you can get speeds of 12.5P, 25P and 50i.

In film, slow motion is shot by running the camera at a faster frame rate. Film normally runs at 24 frames per second (fps), but for slow motion the cameraman might shoot it at something like 48 fps. When those 48 frames are played back at the 24 fps speed, it'll take twice as long to play back, so everything will be moving at half speed, giving that superb film-style slow-motion look. Shooting at a faster frame rate is called "overcranking", because in the early days cinematographers would actually crank the film faster.

The NTSC DVX is the first prosumer camera to offer the ability to actually overcrank, the 'same way a film camera does. If you're shooting at 24 frames per second, you can shoot your slow-motion shots in 30P, at 30 fp. In post, you then play that 30P footage at 80% speed, and you get frame-for-frame accurate slow motion, just like overcranking a film camera. It's very mild slow motion, about 25% slower,but it's the cleanest slow motion possible from a video camera. If this rate of slow motion is appropriate for your shot, shooting 30P and running it frame-for-frame on a 24P timeline will give you footage that's just as clean and free of motion artifacts as if you'd shot film at 30 frames per second and played it back at a 24- frame-per-second speed.

For slower slow motion, you should shoot interlaced video at 60il50i. In 60i mode the DVX shoots 60 frames per second, but they're not full frames, they're fields, which are only half-resolution. But what's important about slow motion is that it's smooth, and to get the smoothest slow motion, you need to capture as many motion samples as possible. Interlaced mode captures 60 motion samples per second. When you import that footage into a 24P timeline, you slow it down to 40% of its speed, and your editing program will make a full frame out of each field. Even though the frames are lower resolution, the result is glass-smooth slow, slow motion. The picture will be softer,

37

but the motion will be smoother than if you just used 30P footage and played it back at 50% speed. PAL users can import 50i footage onto a 25P timeline and play it back at 50% speed to get the same glasssmooth slow motion effect.

When shooting 60i for slow motion, you'll want to use a higher shutter speed. Set the shutter speed to 1I120th of a second to simulate the look a film camera would give if it was shooting at 60 frames per second. For PAL 50i, use a shutter speed of 11100.

Quality slow motion can only really be created at certain speeds. Slowing footage down will require that either new frames be created (which will make the motion stutter) or, if you slow it down the right

percentage, you can get the editing program to match the motion onefor-one without needing to generate artificial frames. An example is using 30P footage in a 24P timeline: the only smooth way to do this is to slow the footage down to 80% of its speed. Any other speed will make the motion jitter and stutter, but at exactly 80% speed, the editing program can make a direct copy of every frame, resulting in perfect slow motion. When using 60i in a 24P timeline, the proper slowdown speed is 40%. Playing back at any other rate will make for stuttery, jittery movement. Only 40% will work to deliver smooth playback. When slowing down 60i footage in a 60i timeline, only 50% speed will deliver smooth playback (because at 50% speed, the computer can create one frame out of each field, again matching motion l-for-l ). Any other speed will make for non-smooth motion. You may be tempted to slow down by another speed (for example, you may want the sequence to run a little quicker, so you may be tempted to slow down to 60% instead of 50%) but doing so will make your footage jitter and stutter. It will not be as smooth. You have to slow down by the proper percentages to make sure the footage will play back smoothly.

Now, what about fast motion? With a little manipulation in post, the NTSC DVX can deliver the look of a film camera shooting at 12 or 15 frames per second. To get 12P, set your shutter speed to 1/24th of a second, and shoot in 24P mode. Then in post, speed up your footage to play back twice as fast. The result will be 12 individual frames, exactly matching the motion of film shot at 12 frames per

38

second. For 15P, set your shutter speed at 1I30th of a second, and shoot in 30P mode. When you bring it into your 24P timeline, play the footage back at 160% speed. That will give you 15 distinct frames, just like a film camera that shot at 15 fps. For pAL users, 12.5P is accomplished by shooting 25P at 1125 shutter speed, and importing into a 25P timeline and playing back at 200% speed.

On the DVXI00A and DVXI00B, even more fast-motion frame rates are possible, by using the "SLOW SHUTTER" function. This lets NTSC users get additional frame rates of2P, 3P, 4P, 6P, and 8P. PAL users get additional frame rates of3P, 6P, and 12P.

Frame NTSC Camera Mode Shutter Speed Playback Rate
Rate
2P 30P on DVXlOOAIB SLOW: 114 1280%
3P 24P on DVXI00AIB SLOW: 116 800%
4P. 30P on DVXI00AIB SLOW: 118 640%
6P 24P on DVXI00AIB SLOW: 1112 400%
8P 30P on DVXI00AIB SLOW: 1115 320%
12P 24P on all models 1124 200%
15P 30P on all models 1130 160% Frame PAL Camera Mode Shutter Speed Playback Rate
Rate
1.5P 25P on DVXI00AIB SLOW: 113 1600%
3P 25P on DVXI00AIB SLOW: 116 800%
6P 25P on DVXI00AIB SLOW: 1112 400%
12.5P 25P on all models 1125 200% Don't forget single-frame animation too: the interval recording mode of the DVXlOO and DVXI00A make it possible to simulate singleframe animation (not applicable to the DVXI00B).

Using these techniques, your DVX can act like a mini-VariCam, shooting at frame rates of 1,2,3,4,6,8, 12, 15,24, and 30 progressive frames per second! (PAL users get 1, 1.5,3,6, 12.5, and 25 progressive frames per second).

39

ANAMORPIDC VS. LETTERBOX VS. SQUEEZE

The DVX has 4:3 CCD's, designed to match the shape of a regular television screen. But many shooters, especially aspiring filmmakers, want the look of widescreen, and some broadcasters prefer to buy wide screen product. People like the look of 16:9 widescreen television, but the vast majority of televisions in the USA are still 4:3, (in PAL territories, there is a higher percentage of 16:9 televisions, but the majority are still 4:3). What's the best way to get the

wide screen look from your camera?

The DVXI00 offers two different ways to get the widescreen look, and the DVXI00A and DVXlOOB offer a third option. How do you know which to choose? To find the answer, you have to answer a different question: what do you intend to display your footage on? If your work is going to be seen on a standard 4:3 TV, the best and easiest way to get the widescreen look is to shoot in Letterbox Mode. Letterbox just adds black bars over your video, but the center portion is still the full resolution the camera can deliver. Letterbox mode delivers the look of a letterboxed widescreen movie on a 4:3 TV.

If you're planning on displaying your work on a widescreen 16:9 TV, you have to convert the shape of the video to match the widescreen display. You have two options to convert your video to widescreen: you can do it optically by using the Panasonic Anamorphic Adapter or, if you have a DVXI00A or DVXI00B, you can get widescreen digitally by using "Squeeze Mode".

The anamorphic adapter is a specially-designed add-on lens that optically squeezes a wide screen picture into your camera, and then when it's displayed on a widescreen TV, the shape of the TV stretches the image back out to be widescreen. This is the highestquality way to get 16:9 footage from a DVX camera: it uses the whole surface of the CCD to record a full frame of information. It basically converts your DVX to be a "native 16:9" camera.

40

The other option is to use "Squeeze Mode". Squeeze Mode works by chopping off the top and bottom of the image, and then digitally stretching the image vertically. Then, when the image gets displayed on a widescreen TV, the shape of the TV causes the image to get stretched out horizontally, giving you a full widescreen picture. Because Squeeze Mode only uses 75% of the potential resolution (the rest gets chopped off the top and bottom), and uses digital resizing to stretch the image, a Squeeze Mode picture will be lower in resolution than the Anamorphic Adapter version. But, it'll still be the highest resolution an interlaced TV can safely display without flicker, and Squeeze Mode is free, whereas the Anamorphic Adapter costs about $850 (street price). And Squeeze Mode, when used in 24p/25p/30p mode, delivers images that are just as high-resolution as a "native 16:9" interlaced camera (such as the Sony FXl in "DV" mode).

The Anamorphic Adapter optically gathers a 33% wider field of view. It gives you a wider perspective, like a wide-angle adapter would (except only on the horizontal axis). Because the anamorphic adapter's field of view is wider, you actually have to zoom in more to get the same field of view that you would get with Squeeze Mode. Because the equivalent anamorphic shot is more zoomed-in than the Squeeze Mode shot, Anamorphic Adapter shots may have a slightly shallower depth of field than equivalent Squeeze Mode shots, when at the same aperture. So when reviewing comparison footage, it can be misleading to gauge sharpness by looking at the background. Look for details in the subject that you've focused on.

In summary, if you're shooting for a 4:3 television, and want a widescreen look, use Letterbox Mode. If you're shooting for a widescreen television, and you have a DVXlOOA or DVXI00B and you like the look of Squeeze Mode, go ahead and use it - it's free, and built into the camera. If you want the highest resolution possible, for a film transfer or for a larger-screen television, the anamorphic adapter produces the highest resolution possible for widescreen footage from the DVX.

If you're using an original DVXI00 and you want widescreen footage (not just letterbox) and you don't have access to the anamorphic adapter, you can simulate "Squeeze Mode" by stretching your footage

41

in post. The DVXlOOA and DVXI00B create "Squeeze Mode" footage by basically shooting Letterbox footage, then digitally stretching the central 75% (the area between the black bars) to fill the full frame. You can do something similar in the post-production edit suite. The trick is to do it in such a way as to get high-quality footage. Non-Linear Editing programs such as Sony's Vegas?', Adobe's Premiere ProTM and Apple's Final Cut™ have the ability to stretch footage, but the quality varies dramatically, and a low-quality stretching algorithm will do more harm than good to your footage.

An alternative that does a great job is PhotoZoom ProTM. PhotoZoom is a software program specifically designed for stretching/resizing images, and on the DVD that accompanies this guide, we used PhotoZoom Pro to conduct the DVX vs. HD up-rezzing test. PhotoZoom Pro was also used by American Cinematographer Magazine to conduct the film-out tests of the DVXI00 (the program used to be known by the name of S-Spline Pro).

PhotoZoom Pro is not technically designed to process motion footage, but it can process batch files. What you can do is exp?rt your footage as a series of still images (up to about 1,000 at a time) and then import them into PhotoZoom Pro as a batch. Tell the program to re-size the footage to 720 x 618 (NTSC). Resizing to 618 pixels will mean that the footage gets stretched far enough that the central 75% (the part between the black bars) will occupy a full 480 lines of height. After resizing, you import all the still frames into your non-linear editor as a still-frame sequence, and tell the program to crop your footage to 720 x 480. It will chop off the black bars from the top and bottom, leaving a properly "squeezed" 720x480 image that will play properly on a widescreen television.

This is a method you can use to generate high-quality widescreen footage from the DVXlOO even if you don't have "Squeeze Mode" or

an anamorphic adapter.

A bigger question to face is, how do you make one video that plays properly on both kinds of televisions? In the USA the vast majority of televisions are 4:3, but a larger and larger percentage of new sets available for sale are 16:9. Walk into any television store and look at

42

the available offerings and you'll see that a lot of the new sets, especially the larger sets, are all 16:9. Whether HDTV comes to fruition or not, the transition to widescreen seems to be underway; however, the vast, vast majority of the "installed base" of televisions are sti114:3. In Europe there's been much more acceptance of widescreen.

How can you shoot one video that plays properly on both types of televisions?

In short, you can't. The aspect ratio is very different, and any particular video is going to look "right" on one shape oftelevision, and "wrong" on the other. It's unavoidable. If you shoot in Squeeze Mode or with the Anamorphic Adapter, and display the footage on a 4:3 television, it wi11100k horizontally squished. If you shoot in

full screen 4:3 mode and display the footage on a widescreen television, it wi11100k horizontally stretched out. If you shoot in Letterbox mode and display the footage on a widescreen television, it wi11100k horizontally stretched out with black bars on the top and 'bottom. The two formats are, unfortunately, basically incompatible.

There are steps you can take to minimize the 'conflict, however. First, consider releasing your video on DVD format. DVD players all have a built-in aspect-ratio conversion system. DVD players will automatically reformat a video signal to properly fit on either type of television. Most Hollywood DVD's are authored in widescreen mode, and most US televisions are 4:3, so the DVD player will automatically un-squeeze and letterbox the movie for display on the 4:3 television. When displayed on a 16:9 television the DVD player will send the full unmodified signal through. This means the best picture quality is delivered on a widescreen TV, but perfectly acceptable video can be shown on 4:3 TV's, so long as the DVD player changes the aspect ratio of the video and adds letterbox bars at the top and bottom.

If you distribute your footage on DVD, you can take advantage of this same feature. Shoot your video in widescreen mode (either using Squeeze mode, the Anamorphic Adapter, or by stretching in post) and author the DVD as a 16:9-enhanced anamorphic DVD. The DVD

43

player will automatically re~c?mpose and letterbox the footage when it's displayed on a 4:3 television,

If a DVD player is not going to be part of the equation, how should you proceed? For example, if the signal is going to be ?r~adcast, . what format should you deliver to the broadcaster? ThIS IS a question with an easy answer, and no easy answer, depending on where in the

world you live.

In the US all standard-definition broadcasts are currently (Sept.

2005) do~e in 4:3 mode, so if you're intending on broadc~sting your video the answer is simple: shoot 4:3. If you want the widescreen look on your broadcast video, shoot in Letterbox mode (or letterbox it in post). For high-definition broadcasts, all hi~-defi~ition . broadcasting is done in wide screen mode. Obviously If you intend to up-rez your footage for a high-definition broadcast: you'll need to shoot in a widescreen aspect ratio. The Anamorphic Adapter would be preferred, as a high-def up-rez needs all the pic~e information it can get. If you want to "future-proof' your production, yo_u c~ consider shooting wide screen mode for future use, but dehvermg a 4:3 master to the broadcaster. Create the 4:3 master by either unsqueezing/letterboxing in post (which will give you the full widescreen image but with letterbox bars) or by "pan 'n' scan'> . where the sides of the image are discarded and the central section IS magnified to fill the screen (similar to how. movies are "f~rm~tted to fit your screen", but when starting with a VIdeo source, this will leave

you with a lower-quality image).

If you live in Europe, where widescreen televisions are ~lUC~ m~re prevalent and widescreen broadcasting does occur, the situation IS more complex - and yet even easier to s.o~ve. European broadcasters broadcast full-size full-height 16:9 on digital channels, and on analog channels (for display on 4:3 TV's) they simulcast a co~promise aspect ratio of 14:9, which is halfway between 4:3 (WhICh could also be expressed as 12:9) and 16:9. In 14:9, small letterbox bars are added to the top and bottom, and small porti~ns of the left and rig~t are chopped off. The result is an imag~ that IS 14:9 sh~ped, and ~Ill display on a 4:3 television. It will retam.much of the picture while adding mild letterbox bars. When shooting for European broadcast,

44

the recommendation is to shoot widescreen, while "protecting" for 14:9 display. The way you protect for 14:9 is to try to make sure that no critical information is in the leftmost or rightmost 10%. Try to keep the action more in the center of the screen so that when the edges are chopped off (in the conversion to 14:9) the critical portion of the image will still be displayed.

Shooting widescreen (either through Squeeze Mode or through using the Anamorphic Adapter) can present a challenge for monitoring the footage on a DVXI00 or DVXI00A. Since both Squeeze Mode and the Anamorphic Adapter change the shape of the video to be widescreen, it no longer displays properly in the DVXI00 or DVXI00A's viewfinder (or LCD panel). This can make framing and composing your shots a little trickier, as you have to "imagine" what the frame would look like in its ''unsqueezed'', widescreen shape. There are some products on the market to help see the properly ''unsqueezed'' image:

1) You can always output the video to a 16:9 monitor. The monitor will unsqueeze the image and display it in its proper aspect ratio. Many production monitors are switchable between 4:3 and 16:9, and there are several inexpensive 16:9 LCD monitors on the market. Most of these inexpensive LCDs will not be suitable for focus or color accuracy, but they will show the image properly unsqueezed.

2) Century Optics markets a Widescreen Eyepiece (part #VSWSEP-DVX) that attaches to the DVX's viewfinder. This attachment optically un-squeezes the image, providing the proper aspect ratio to the viewfinder.

3) Century also markets an option for the LCD panel- the VS-LCWS-DVX LCD Widescreen Magnifier. This product slides over the flip-out LCD panel and unsqueezes and magnifies the LCD image, as well as providing a hood for shading the LCD from sunlight.

4) The DVXI00B has a built-in aspect ratio conversion option for the LCD and Viewfinder. Choose the 16:9 aspect ratio in the menus, and the camera will automatically unsqueeze the image and display it in its proper aspect ratio.

45

INTERLACE LINES ON PROGRESSIVE FOOTAGE

When shooting in the progressive modes of the DVX, many first-time users are disappointed or even alarmed to see interlace-line artifacts on their televisions. They assume there must be something wrong with their camera, or that they did something wrong: after all, they reason, if shooting in progressive, they should never see interlace artifacts, right?

If displaying the footage on an interlaced television, the answer is "no." The interlaced television will display interlace artifacts even on progressive footage. To explain, let's first separate 30P (or PAL's 25P) away from 24P. The way 24P is recorded, there will be frames that are blended together on the tape - that's unavoidable. 24P Normal blends two frames together out of every five, and 24P Advanced blends one frame out of five. However, 30P (and 25P on PAL) record with no blended frames: each progressive frame is written out to two fields of video (called a "frame" of video); there is no blending or field-mixing occurring. When displayed on a progressive device (such as a computer monitor) you will never see any interlace artifacts from 30P/25P footage.

But on an interlaced television, you'll see lots of interlace artifacts. It's unavoidable, based on the way interlaced televisions function.

For the sake of clarity in this discussion, we will define the term "frame" to mean a complete, discrete picture of full resolution (similar to how the term is used to describe an individual "frame" of movie film). In DV video, a frame consists of480 horizontal scan lines, each containing 720 pixels of information (for NTSC; in PAL a frame consists of 576 scan lines of 720 pixels each).

Interlaced televisions do not display "frames", and much confusion comes about when people describe NTSC televisions as running at 29.97 "frames" per second, or PAL televisions at 25 "frames" per second. They do not ever display actual "frames". What interlaced televisions display is a series of "fields", and the term "field" can be

46

described as basically Y2 of a frame, consisting of (altematingly) the even or odd lines. Fields are not divided in half by something as simple as "top" or "bottom", instead each pair of fields consist of one field that contains every even scan line, and one field that contains every odd scan line. Because the fields are divided in this manner, interlaced television can provide the illusion of updating the entire display every time, when in reality it's updating only half the picture (the even half or the odd halt). Because the television displays the even lines, followed by the odd lines, the image is said to be "interlaced". Think of it as clasping your hands together: the result is a unit that consists of I 0 fingers, interlaced together. The "even" fingers may belong to your left hand, and the "odd" fingers to your right hand, but when clasped together they form a 10-fmger unit. That's basically how interlaced televisions work: they have one field that consists of the "even" lines, and one field that contains the "odd" lines, and they get drawn on the television one after the other.

Televisions "paint" the image on the screen, using a raster scan system that starts at the upper-left comer and draws an entire line,

" then it starts over at the left again, skips a line and draws the next full line, and so on until it's drawn an entire field, every other line. After completing an entire field (say, the "even" lines) the raster beam will start back at the upper-left comer, drop down one line, and commence drawing in all the odd lines. When the raster fmishes drawing all the odd lines, it's said to have displayed a "frame", since all the even and all the odd lines are existing on the display at the same time.

Now, the question of how progressive scan can display interlace artifacts can be more easily understood when you understand that the television doesn't draw two fields and then take a break, as it were: instead the television regularly and continuously draws new fields:

Frame 1 Even, Frame 1 Odd, Frame 2 Even, Frame 2 Odd, etc. It draws these fields at regular and continuous intervals in time. But what happens is that at any particular moment in time, two fields will be displayed: there will always be an even field, and always an odd field being displayed by the television. But fully 50% of the time, those fields will not be coming from the same frame!

47

The following illustration will show how this works. Assume that when recorded, our 30P video consisted of four distinct progressive~~.u._u, like so:

If this sequence were to be played back on a progressive-scan device, such as an LCD television, a computer monitor, or transferred to film, you would see exactly that: a progressive-scan sequence of frames, with no interlacing and no artifacting. However, if played back on an

interlaced TV 'Il instead see some.me this:

Because the interlaced television runs at 60 fields per second, it will actually update twice per frame, drawing the odd and then the even lines. So for the first frame, the odd and even lines of frame "A" get drawn. When moving to the next frame, it will draw the odd lines, but the even lines from frame "A" are still displayed! So for that instant that 1/60th of a second, you will see an interlaced frame on the television. The next 1/60th of a second the television will draw the even lines from frame "B", and it will look proper. But the next 1/60th of a second, it will draw the odd lines from frame "C", and again you'll see an interlaced frame on your television.

This is not a problem with the camera, it's an inevitable occurrence that happens from mixing media, trying to display 'progre~sive content on an interlaced display. Watch carefully in movies that have been transferred to video, and you'll see the same effect. Film frames are transferred to video the same way the DVX transfers its progressive frames to video, and the television displays the~ th~ same way, so you will see the same type of effect. But, agam, display

48

your progressive footage on a progressive-scan display and you'll see it in all its pristine progressive nature.

When editing on the computer, if you shoot in 30P (or 25P in PAL) you should never see an interlaced frame. However, when shooting 24P, it's entirely possible that you will see interlaced frames. Because the frames are written to tape in their interlaced format (as shown in the examples above and also in the 24P vs. 24PA article), importing that footage into a 60i time line will bring it in exactly as it's written to tape. And on tape, some frames are blended. On the computer screen, you'll see blended frames. The way around this is to be sure to use an editing program that's capable of extracting the "pulldown" from the footage. Use a program such as Sony's Vegas'", edit in a 24P timeline, instruct the program to remove

pull down from imported footage, and the program will automatically discard split frames out of 24 P A footage, or split up and recombine footage from 24PN footage, to give you access to the original 24P frames with no interlacing lines. Of course, if you intend to output that footage back to DV tape (or VHS tape or any other NTSC video standard) you will have to "re-introduce" pulldown, since 24P . footage cannot be recorded on NTSC unless it has pulldown introduced, which rounds out the 24P sequence into a grouping of 60

NTSC.fields. .

DEPTH OF FIELD

Perhaps no effect is more sought-after than the look of shallow depth of field (DOF) - when the subject is in sharp focus, and the background is out of focus.

Unfortunately, the effect is quite difficult to achieve on a small CCD camcorder - there's a reason that the shallow-depth-of-field look reminds people of the movies! It's because you can usually only ever see that look IN the movies. But there are ways to maximize the shallow depth-of-field look on the DVX.

First, let's explain what depth offield is. A lens can only focus on one plane in space at anyone time - there's really only one infmitelythin plane that's ever truly "in focus". Everything closer to the lens,

49

or further from the lens, will be progressively more and more out of focus; but there's a certain distance from that focal plane where objects are still acceptably sharp. That distance, that area in front of and behind the in-focus point, is called the "depth offield." It's interesting to note that the depth of field is not equally split: it extends for 113 of the distance in front of the subject, and 2/3 behind it. That means it's easier to get objects out of focus in the foreground than it is to get them out of focus in the background.

- Depth of field is controlled by three primary factors: distance to the subject, focal length, and aperture (also called iris, or f-stop). By changing any of these factors, you can deepen or narrow the depth of field. Some are more important than others. And, some cancel each other out. For example, getting closer to the subject will make for shallower depth of field. However, in order to get the same subject size, you have to zoom out, and zooming out to a wide-angle shot makes for deeper depth of field. The two basically cancel each other out, although, for any given subject size, the more telephoto the shot, the shallower the depth of field will appear to be.

The factors work to control depth of field like this:

SHALLOWER DOF DEEPERDOF
Close to subject Far away from subject
Telephoto Lens Wide-Angle Lens
Open Aperture (like fl2.8) Smaller aperture (like fl16) So: to maximize the area that's in focus (to get the deepest Depth Of Field), use a wider-angle on your lens, and stop down the iris as much as you can, and back away from the subject.

To minimize the Depth OfField (to get the shallowest focus possible): zoom in as far as you can, get as close to the subject as you can, and open the iris up as much as possible.

While the optics and laws regarding depth of field are complex, for most circumstances there are really only two factors you need to pay attention to: f-stop and zoom setting. And even then, to get the shallowest depth of field possible, you usually only have to remember

50

two numbers: f12.8 and Z99. Treat your camera as ifit has only one aperture, f12.8, and the zoom is stuck at maximum telephoto, Z99, and you're well on your way to getting the shallowest depth offield the camera can deliver.

First, let's look at zoom setting. One of the key factors in getting the shallow depth of field look is to get a long telephoto lens. Unfortunately, the DVX's maximum telephoto is fairly mild, at only 45mm. Other cameras have longer telephoto, and can get shallower Depth OfField: the PD170 has a 72mm lens, and the XL2 has a very long 11 Omm lens. With the DVX, you only have a maximum telephoto of 45mm. To get shallow DOF shots, you need to use all the telephoto you can get. Lock that lens at Z99, and if you need to adjust the framing of your shot, do it by moving the camera, not by zooming out. Even zooming out a little bit may ruin the shallow Depth ofField look you're trying so hard to create. You can also get a telephoto extender lens from Century Precision Optics that will give you even more telephoto reach, and shallower DOF.

Also, aperture is crucially important. You want the aperture as wideopen as you can get it. On the DVX, when zoomed all the way in, that means f12.8 (or "OPEN"). If your shot is overexposed at f12.8, don't stop down the iris: instead engage the Neutral Density filters to bring the exposure back where it should be. If you need more control over the exposure, get an external l-stop and a 2-stop ND filter (also called an ND.3 and an ND .6). These, combined with the camera's built-in filters, will give you the ability to control exposure to f12.8 under almost any lighting conditions.

Once you have your camera set at f12.8 and Z99, and you've moved the camera so that your framing is appropriate and you've set the neutral density filters so that the exposure is correct, there are a few more tricks you can employ to get the shallowest Depth ofField possible. First, remember that the depth of field extends for 113 in front of the subject and 2/3 behind it. We can take advantage of that to try to slide the whole zone of Depth ofField closer to the camera. Try to focus as close as possible, getting the smallest MF number you can, while maintaining acceptable sharpness on your subject. This is a "cheat" that will help shorten the perceived depth of field.

51

L

Second, a very effective technique is to "cheat" the backgro~d away from your subject. Separate your subject as far away as possible from the background, and you can force the background outsi~e the depth of field and push it further out of focus. Whenever possible, stage your shots so the subject is as far as you can g~t from the background _ even if that means cheating the shot, or staging the shot such that the background is across the street. By using a long telepho~o lens, the lens will optically compress the space between your subject and

- the background, making them look like they're a lot closer than they

really are.

In tight quarters it gets very difficult to achieve the shallow~D~pth Of Field look. At some point you have to acknowledge that this Just isn't a 35mm movie camera, it's a 113" CCD video camera and there's only so far you can push it, but we're not ready to give up yet: there are still a few tricks up our sleeve.

First if the room is so small that you simply cannot use Z99 to frame your'shot, you can try going close instead. Zoo~ out and get as absolutely close to the subject as you can. Getting extremely close will narrow the depth of field and also expand the background, and allow you to open up the iris some, maybe even to f11.6. Th~s technique really only works for extreme close-up shots, but ~t can be very effective. If you have an object in the foreground too, It can really add depth to the shot: when at wide angle focal lengths, the lens will exaggerate the distance between objects, so with a shallowfocus background and a shallow-focus foreground, and exaggerated depth between them, it can make for some compos.ition~lly . interesting shots. If you'd like to be able to zoom in while still

, getting close (to compress the space between objects), you can add a high-quality diopter lens, like the Century Achromats or ~e inexpensive Nikon 5T, which will let you focus closer while zoomed. Opening the iris and getting closer will compensate for some of the loss from not being able to zoom all the way m.

Another way to get the shallow depth of field look is to use a product called the SoftScreen to noticeably soften any background. The SoftScreen is a sheet of transparent acrylic that de-focuses everything

52

behind it. Place the SoftScreen behind your subject, and everything behind it will be significantly more out of focus, regardless of your aperture or zoom setting. The SoftScreen is available exclusively from http://indietoolbox.com.

Finally, you can try using.a greenscreen or blue screen technique to force the background out of focus. This is not really an easy or practical solution, but if you need to get the background out of focus and can't do it any other way, you can consider this method.

Set up a greenscreen behind our subject. Shoot your subject in the sante place and the same light as if the greenscreen wasn't there. Then, remove the greenscreen and the subject, and shoot a background shot as out-of-focus as you want. Then when you composite them in post, you can make for a fairly convincing softbackground look. Greenscreen or bluescreen is quite difficult to pull off effectively with a DV camera, so we're not recommending this technique, but you can consider it an option of last resort if none of the other techniques are working for you.

Finally, instead of shooting at your subject, shooting into a mirror may be a way to get enough distance that you can zoom all the way in. You'll have to "flip" the footage in post (because the mirror will reverse the image) but it may help when in tight quarters.

In summary, for shallow depth of field, remember these four things:

Z99, fl2.8, get as close to the subject as possible, and separate the subject as far from the background as possible.

Depth of Field Myths:

Myth 1: "imaging size affects DOF, larger imagers = shallower DOF". False. The imaging size has nothing to do with the depth of field. The imager size (whether larger like 35mm film, or small like 1/6" CCDs) affects the FIELD OF VIEW, but not the DEPTH OF FIELD. Larger imagers require longer lenses to deliver a usable field of view. Smaller imagers require shorter lenses. Consider a standard 10: 1 zoom lens: on a 35mm camera, it would typically be something like a 25-250mm zoom, on a 16mm camera it would be IOmm- 100mm, on the DVX it's 4.5mm-45mm, and on a 1/6" CCD camera it

53

might use 2.5-25mm. The 35mm camera gets its shallower DOF from its longer lenses, not from its imager size. On the 35mm camera, the very widest it can zoom out with that lens is 25mm; on the 1/6" CCD video camera, the very most telephoto it could get is 25rom! It is the focal length, not the imager size, which dictates the deep depth of field. Sensor size is tangentially related to DOF in that the smaller sensor demands the wider angle lenses in order to get a usable field of view, but there's nothing inherent about a smaller sensor that is causing the depth of field to magically deepen.

Myth 2: "Focal length is irrelevant in DOF. To use telephoto, you have to back up (which increases DOF). To get close, you have to use wide-angle (which increases DOF). They cancel each other out completely." No they don't. Technically they do, but for any given subject size, the telephoto shot will clearly look like its background is more out of focus. Whether it's technically as "in-focus" or not is irrelevant, the telephoto lens delivers the optical illusion that the background is more out of focus, and that's what you're after: the APPEARANCE that the background is out of focus. The reason it delivers the appearance of a softer background is because the telephoto lens has a much narrower field of view, so for any given shot composition, the entire background will be made up of a smaller chunk of background, as compared to a wider-angle shot (see the examples on the accompanying DVD). So the background will be more optically magnified to fill the screen. It is this optical magnification which makes the background look like it's more out of focus. With a wide-angle lens the background is actually shrunk down (because the perspective exaggeration of the wide angle forces the background to appear further away, thus fitting more background into the shot, which forces what background there is to be rendered smaller); with the telephoto lens the background is actually magnified. So even though technically the background may be just as out-of-focus in the wide-angle shot, it doesn't LOOK like it's as blurry, because by shrinking it down smaller it looks sharper. Whereas with the telephoto shot the out-of-focus-ness of the background is magnified by the optical compression that the telephoto lens exhibits, so the background looks more out of focus than it actually is.

54

MATCHING DVXIOO TO DVXIOOA OR DVXIOOB

With all the changes and improvements made in the DVXIOOA and DVXIOOB, is it possible to use a DVXIOO and a DVXIOOA or B in a multi-camera shoot and get images that will be able to be intercut?

The answer is, yes, if you're careful. First, make sure all the cameras are set to identical settings. This means you cannot use any of the image adjustments that are new to the DVXlOOA (no B.Press gamma, no Cine-Like-V, no Enriched matrix, no Detail Coring, etc). Second, and this is vitally important: you must manually white-balance the cameras. Take the cameras to a place where you can white-balance them under the same lighting conditions to the same white object. If you use the 3.2k or 5.6k preset, the cameras will not match well, color rendition will be significantly different. However, if you whitebalance them to the same white object in the same lighting

conditions, they can intermix and match very well.

SYNCHRONIZING MULTIPLE CAMERAS IN A MULTI":

CAMERA SHOOT

The DVXIOOB offers the ability to synchronize its timecode preset to any DV device - you can plug in a DV camera or DV deck into the DVXIOOB (through the 1394 jack) and the DVXlOOB can automatically read the timecode from the other device and synchronize to it. This makes synchronizing timecode between multiple cameras a breeze.

The DVXIOO and DVXIOOA do not have this feature, but they can still be set to synchronize with each other. Because the DVX offers complete user control over timecode, it's possible to synchronize the cameras to have exactly matching timecode for any particular time of shooting. This can be extremely handy when shooting concerts, reality shows, or multi-camera dramatic shoots.

The key to synchronizing the cameras is to use FREE RUN timecode. For most normal recording situations, the standard is to use REC RUN. However, for synchronizing multiple cameras, FREE RUN has some compelling advantages. If all the cameras are initialized

55

with the same timecode preset, and they're all set to FREE RUN at the same time, then the timecode will be identical on all the cameras throughout the day. This can make matching up takes in the edit bay easy and effortless.

For the DVXIOO and DVXIOOA, the process relies on using one remote control to send a signal to multiple cameras simultaneously. For the DVXlOOB, timecode is synchronized by connecting two cameras together using a 4-pin to 4-pin 1394 cable.

All cameras have to be configured to use the same timecode mode in the RECORDING SETUP menu. First, go to the FIRST REC TC menu item and select PRESET mode; then set them all to the same TC MODE setting (either DF or NDF, whichever you prefer, but they must all be identical).

For DVXIOO and DVXIOOA: Go to the TCG menu item and set REC RUN mode. Next, set the TC PRESET menu item and initialize the preset value so that all the cameras have the exact same timecode preset. Make sure all the cameras are set to respond to the same remote control (in the OTHER FUNCTIONS menu, make sure they're all set to either VCRI or VCR2, but that they're all the same as each other). Then get all the cameras together in one place, as close to each other as possible, and aim a single remote control so that it's pointing equally at all the cameras.

Then, using only the remote control (so all the cameras ~re receiving the signal at the exact same time), go to the TCG menu Item and set the cameras to FREE RUN mode. From that moment on, all the cameras should have identical free-running timecode.

For DVXIOOB: You'll designate one camera as the "master timecode" camera, and all other cameras will sync to the master camera. (Note: the master camera does not have to be a DVXlOOB; the DVXIOOB can sync to any camera's DV timecode). Make sure the cameras are set in FREE RUN mode, and all the cameras need to . be set equally to either DROP FRAME or NON DROP FRAME.

56

Make sure the master camera is in camera mode. Set the DVXIOOB into VCR mode. In the RECORDING SETUP menu, make sure the DVXIOOB is configured to FREE RUN and PRESET mode (don't worry about setting the TC PRESET menu item; the DVXIOOB will automatically read that from the master camera through the 1394 cable). Also, make sure 13~4 TC REGEN is set to OFF. Finally, set the DV IN PRESET menu item to ON. Exit the menu system. You'll see "DVTC" displayed on the LCD indicating that the camera's ready to receive external timecode data.

Next, connect the DVXlOOB to the master camera via a 1394 cable. Press the TC SET button and the DVXIOOB will force its TC PRESET menu item to match the timecode it's receiving through the 1394 port. You can disconnect the 1394 cable, as from that point on the cameras should stay in sync. Note: the only data that gets transferred through the 1394 cable is the value that gets assigned to the TC PRESET menu setting; all other settings (such as DFINDF, REC RUNIFREE RUN, etc) are not transferred when pressing the TC SET button. You have to make sure that those settings are compatible between cameras manually; or, if you're configuring - timecode between two DVXlOOB cameras you can ensure that they're set equally by transferring the User File from the master camera to all the other DVXlOOBs before setting the timecode itself Transferring the User File will send the settings for all camera setup information (but not scene files) to the receiving camera; this includes setting all timecode-related settings equally.

Sync Drift: Once you set the timecode preset to be the same, all the cameras will record the exact same timecode for any scene, regardless of when you start or stop shooting, so when shooting with multiple cameras you can easily synchronize the footage by lining it up according to timecode. However, be aware that sync may drift slightly between the cameras throughout the day, and if you tum off one of the cameras, change batteries, or switch a camera to VCR mode, that can briefly interrupt the flow of time code. To maintain perfect sync you will probably need to re-sync the cameras after

doing so. With the DVXlOOB it's a simple matter of reconnecting the 1394 cable and pressing the TC SET button. With the DVXIOO and DVXIOOA, it means going through the whole process again.

57

SCENE FILE SETTINGS

The Scene File Dial is also known in Panasonic literature as the CineSwitch™. The Scene File Dial allows you to program in several image-control settings, and with six positions on the dial, this allows you to have six different "looks" pre-programmed into the camera. The settings that are controllable for the Scene File dial are controlled by the SCENE FILE menu.

The following descriptions of the scene file properties are designed to give you a better understanding of how these settings affect the image. The DVD that accompanies this guide shows examples of the settings being changed, along with interactive commentary. Further commentary about the scene file settings is provided here.

DETAIL LEVEL:

DETAIL LEVEL controls edge enhancement and overall sharpness of the picture. With DETAIL LEVEL set all the way down to -7 the image will be noticeably blurry. With DETAIL LEVEL set up to +7 the image will look much, much sharper: perhaps too sharp. Setting a higher DETAIL LEVEL brings in edge enhancement, an artificial video sharpening that can look unnatural. Excess detail also starts to interact with the DV compression algorithm: with DETAIL LEVEL turned up to a really high level, you may see an increase in "mosquito noise" artifacting in the picture. With DETAIL LEVEL it's probably best to avoid the extremes: -7 is quite blurry and +7 shows some artifacting. The smaller the DETAIL LEVEL number, the softer and more organic the image will look. The larger the DETAIL LEVEL, the sharper and more electronic it'll look. For most purposes, unless you're going for a soft-focus effect, some detail is good. The default is zero, and for good reason, as the negative values can get blurry.

V DETAIL LEVEL (DVXIOOAIDVXIOOB ONLy)

V DETAIL LEVEL affects perceived sharpness, similar to the DETAIL LEVEL control, but in a different way: it primarily enhances the contrast between horizontal lines. Whereas DETAIL

. LEVEL works with edge enhancement, artificially drawing outlines around objects to accentuate their edges, V DETAIL LEVEL works

58

with the existing image, accentuating vertical contrast between horizontal elements. It will slightly sharpen (or slightly unsharpen) between vertical lines in the video image. V DETAIL LEVEL doesn't draw new elements into the picture, but it will enhance the existing contrast on vertical edges (white edges become whiter, dark edges become darker). This enhanced contrast leads to the illusion that the picture is actually sharper. The V DETAIL LEVEL setting has a much milder sharpening effect than DETAIL LEVEL.

DETAIL CORING (DVXIOOAlDVXlOOB ONLy)

Detail Coring applies a slight smoothing effect to noisy areas of the video. The higher the Detail Coring setting, the more it will blend noisy areas together. The effect is most noticeable when DETAIL LEVEL is set to a high value, because high DETAIL LEVEL really brings out the noise in the picture - DETAIL LEVEL actually sharpens the edges of the noise and makes it more visible, which makes it easier to see the effect of DETAIL CORING. At -7, DETAIL CORING has minimal effect on the noise. At +7 it has a significant smoothing effect, and really cleans up the noise in the video signal. The higher the DETAIL CORING setting, the more smoothing you'll see, and at higher levels it can lead to your picture looking a little softer. Also, the lower the DETAIL LEVEL setting, the less effect DETAIL CORING will have on the image.

CHROMA LEVEL

CHROMA LEVEL refers to the amount of color saturation the picture has. The lower the CHROMA LEVEL, the more pale and muted the colors will be. The higher the CHROMA LEVEL, the more saturated the colors become. At -7 the picture will have the least amount of color saturation, but there will still be quite a bit of chroma recorded (you can't get a black and white picture even with the CHROMA LEVEL turned all the way down). At +7 the colors will be very strong and very vibrant.

CHROMA PHASE

CHROMA PHASE lets you adjust the color response of the camera along the yellow-green and purple axis. It works like the "TINT" control on NTSC televisions. At -7 the picture shifts slightly towards yellow/green hues, and at +7 the picture shifts towards

59

l.

purple/magenta hues. The effect is quite mild though - you won't make the picture strongly greenish or strongly purple-ish, but you can mildly balance the picture.

COLOR TEMP

Whereas CHROMA PHASE lets you adjust the color between green and purple, COLOR TEMP lets you adjust the color between red and blue. A -7 setting makes the picture warmer and redder; a +7 setting makes it cooler and bluer. The COLOR TEMP setting gives you a strong degree of influence over your picture, much stronger than CHROMA PHASE does: at -7 the picture will be very orange, and at +7 the picture will be very blue. At smaller increments you can use this setting to add a slight warming or cooling to your picture: a COLOR TEMP of -1 would provide a warming effect similar to using an 812 warming filter. Note: the COLOR TEMP setting has no effect when the white balance switch is set to "preset". It only works if you manually white balance, or if the camera is set to Auto Tracking White balance mode. Using the 3.2K or 5.6K preset white balance will disable the COLOR TEMP function. For a brief defmition, Color Temperature refers to the property of a "black body" (say, a piece of iron) to glow different colors as it is heated. At lower temperatures (say, 3200 degrees Kelvin) the iron will glow reddish, and at higher temperatures (such as 5500 degrees Kelvin) it will glow bluish-white. Color temperature generally expresses chromaticity of a source oflight such as indoor or outdoor light. For example, if the temperature is relatively low, the light appears reddish. As the temperature increase, the light changes from red to orange to yellow to blue to white. This COLOR TEMP setting lets you shift the overall color of your video picture along this scale, from red to blue.

MASTER PEDESTAL

The MASTER PEDESTAL governs the way the camera handles the darker sections of the picture, and also acts similar to an overall "contrast" control. The lower you set the MASTER PEDESTAL, the more dark items will all blend together into black, giving you stronger, harsher contrast and a loss of detail in the shadow areas. The higher you set the MASTER PEDESTAL, the more the contrast washes out, and the overall contrast will look softer and flatter, but you'll also preserve more detail in the shadows. A higher MASTER

60

PEDESTAL value will preserve detail in the darker areas of the picture, but make the blacks "milky." The MASTER PEDESTAL governs the contrast of the darker sections of the picture, but doesn't, affect the brighter sections. A low MASTER PED (such as -15) makes for a dramatic, harsh picture, more like reversal film: limited latitude and strong contrast. A higher MASTER PED (such as +15) results in a pastel, flat, extremely low-contrast look, with pale grays instead of sharp blacks in the picture. For high contrast, moody shots, a lower MASTER PEDESTAL setting might be appropriate, and for the maximum exposure latitude you might think about setting the MASTER PEDESTAL higher. In standard analog video the Pedestal (or "black level") is normally set at 7.5 IRE, and in digital video it is set at 0 IRE. This adjustment allows the user to slightly control the pedestal. This will change the black reference for the video signal and allow for either slightly rising of the black level (making for "milky" blacks) or crushing the shadow detail all into black.

AUTO IRIS

The AUTO IRIS setting lets you instruct the automatic exposure system to bias the exposure to be darker or brighter. The range is' from -4 to +4, and each increment represents roughly 113 of a stop of exposure. When using the camera's automatic iris control, an AUTO IRIS setting of -4 will make the camera select an f-stop about 1.3 to 1.5 stops darker than it otherwise would, and a setting of +4 will make the camera choose about 1.3 to 1.5 stops brighter than it otherwise would. Use care with the AUTO IRIS setting, especially with + values, as you can easily overexpose your video, leading to ugly "blown out" highlights. Take special care when using CINELIKE gamma, as there is no blow-out protection in the CINE-LIKE gamma curve. This is why the camera scene files default to an AUTO IRIS setting of -3 in scene files that specify CINE-LIKE gamma, to provide some protection against blowouts.

GAMMA

The gamma curves in the DVX control how brightness information is distributed in the picture. Gamma correction can correct for the nonlinear light-output characteristics of a standard TV picture tube. Picture-tube gamma (like on your television) stretches the whites and compresses the blacks. Camera gamma compresses the whites

61

and stretches the blacks. For the technically inclined, camera gamma can be properly set by using logarithmic gray scale charts and a waveform monitor. Camera gamma must be the reciprocal of picture gamma which is 2.2, so the camera gamma is usually 0.45.

With NORMAL gamma, standard video-looking pictures are produced, and the camera responds like a typical video camera.

LOW gamma compresses some of the brighter parts of the picture slightly, while preserving the darker parts. This may help you preserve shadow detail, at the expense of some contrast range in darkening the brighter parts of the picture, but also gives you the ability to iris up a little, since the gamma curve will be compressing the highlights. Because of this, LOW gamma lets you get more detail in the darker sections of the picture. This setting allows for a moderate gradient for the low luminance areas. The video image looks a little sharper due to the change in the gamma curve.

HIGH gamma brings up the brightness level of the darker parts of the picture, while maintaining the same level of brightness in the brighter levels. This raises the dark level, making for flatter contrast and an overall brighter-looking picture. It also provides some leeway for you to be able to iris down and still preserve detail in the shadow areas, thus extending the range of highlights that can be captured. This setting allows for a sharp gradient for the low luminance areas, thus the dark areas are extended to make them appear brighter.

CINE GAMMA changes the brightness response substantially. The overall contrast range is compressed and lowered, the brights are brought down, and more headroom is provided at the top of the scale. This actually gives the overall shot more latitude, especially in the brighter parts of the picture. While CINE GAMMA can provide more latitude (providing for more range from the darkest dark to the brightest bright), there's a potential drawback as well: there's no "knee" protection at the top end of the scale. In LOW, NORM and HIGH gamma, there's a "knee" circuit that governs overexposure, ensuring that overexposure is gently rolled off, making for a smooth transition from bright to overexposed. In CINE GAMMA no knee protection is available - brights are rendered smoothly and cleanly

62

right until they smack into overexposure and blowout. This can ruin your shot if you're not careful, but if you carefully control your shot and prevent overexposure, CINE GAMMA can result in the most - latitude the camera can deliver. It has a characteristic soft-contrast look to it because of the compression of the highlights. Because there is no "knee" protection, it is wise to underexpose about Y2 stop to avoid going into overexposure on the highlights.

The DVXlOOA and DVXlOOB add three additional gamma settings. CINE _ LIKE _ D provides the same overall gamma response as CINE GAMMA, but the highs are amplified quite a bit, stretching the apparent dynamic range. Everything gets brighter in CINE _ LIKE _ D, even into overexposure. However, CINE _ LIKE _ D provides knee protection to smooth the transition into overexposure. CINE_LIKE_D is optimized for Dynamic Range (hence the "D" in its name) but it accomplishes this by using electronic gain to boost and stretch the signal. The side effect is an increase in noise in the picture, when compared to CINE GAMMA. The overall appearance of dynamic range of the video is higher than at the Cine-Like setting.

CINE_ LIKE _ V is optimized for contrast. The response is similar to CINE GAMMA, but the camera stretches both ends of the spectrum to exaggerate the distance between the darks and the brights, resulting in sharper, punchier contrast. More of the potential dynamic range is used so the image appears to have sharper contrast, but it's done electronically by stretching the gamma curve. This means you actually may end up with a little less latitude than regular CINE GAMMA. CINE LIKE V also provides knee protection against

_ _

blown-out highlights.

Unlike LOW and HIGH, B.PRESS affects primarily the midtones of the image. It leaves the highs alone, and presses the lows down a little, but really presses down the midtones, providing for more dynamic range in the middle of the tonal scale. This setting produces a sharper contrast compared to the low setting.

63

KNEE (DVXIOOA and DVXIOOB ONLy)

The KNEE helps prevent overexposure by rolling off the intensity of the brightest parts of the picture. With the KN~E circuit. engaged, ~e camera will detect when the highlights are getting too bnght and WIll start attenuating the signal to bring them back into the exposure range that the camera can properly resolve. This menu setting lets you decide at what point the knee begins working. When set to LOW, the knee will start attenuating signals at 80% of brightness. At MID, it'll leave 80% alone and start rolling off the signal at 90%. At HIGH, it doesn't affect the signal until it sees 100% brightness (out of 109% total). In AUTO, the camera decides what level it should set the knee at (and, like all auto settings, it may change in the middle of a shot).

Obviously the sooner the knee begins working (i.e., LOW) the more it will be able to protect your image against overexposure. Ho.wever, too much knee attenuation can look artificial, turning your whites to gray; whereas using a higher KNEE setting lets the image app~ar "naturally" over more of the dynamic range of the camera. Using the knee is a way to extend the camera's dynamic range: the lower you set it the further you'll extend the dynamic range ofthe camera, at the potential expense of the highlights looking a bit gray.

Normally, as the intensity of light increases, so does the signal proportionally until the signal exceeds the a?i1ity ?f the camera to record it; any brighter and it just clips to solid white, !hat's how CINE GAMMA works. The knee extends the dynamic range by compressing high intensity signals, somewhat lik~ an audi~ limi~er compresses audio signals to prevent overmodulation and distortion.

SKIN TONE DTL

SKIN TONE DTL is designed to help smooth wrinkles, blemishes,

and acne on people's skin. When set to ON, it .applies a ve~ mild blur to anything it perceives as "skin tones," "':Ithout a!fecting any other aspect of the picture. When set to OFF, It doesn t try to blu~ skin tones. The overall effect is, however, very subtle. Its effect IS more noticeable the higher you have the DETAIL LEVEL set. For a description of how the SKIN TONE DTL function works, look. at the description for DETAIL CORING. SKIN TONE DTL works like DETAIL CORING, except only on colors and tones that it perceives

64

to be "skin". The higher you have DETAIL LEVEL set, the more noticeable the effect will be. Even at the highest setting, the effect of SKIN TONE DTL is very, very mild and subtle.

MATRIX

The MATRIX setting lets you, choose different patterns of overall color response. The color matrix basically remaps the original colors to a new matrix of colors. Using the MATRIX control you have a limited amount of control over how saturated the colors are, and which colors get enhanced and which do not.

NORM is a mild, normal color response. In NORM the colors are at their least saturated. This matrix is frequently used for shooting outdoors or under halogen lighting in the studio.

In FLUO the colors are boosted on the red, yellow, cyan and magenta, but not so much on the blue and green. This matrix doesn't enhance green because it's designed to be used when shooting under fluorescent lighting conditions. Many fluorescent fixtures (especially older office fixtures) have a strong spike of green in their color spectrum. The FLUO setting accentuates the other colors, but not green, in the assumption that the green spike of the lights themselves will compensate and bring the green up to a level comparable to how the other colors are boosted by the FLUO matrix setting.

In CINE-LIKE all the colors are strongly saturated, there's an especially big boost to green and to magenta. This matrix is used to produce richer, cinema like color.

The DVXIOOA and DVXlOOB add another matrix setting, ENRICHED. ENRICHED boosts the yellow/red/magenta section of the color scale, while having less of an effect on the blue/cyan/green section. ENRICHED can work something like a didymium red enhancing filter, really drawing out the red portion of the spectrum, especially useful in nature photography.

VDETAILFREQ

V DETAIL FREQ setting is only applicable when shooting in progressive-scan mode, either 24P (25P PAL) or 30P. It has no effect

65

on interlaced video (50i PAL or 60i NTSC, when PROGRESSNE is set to OFF). There are two settings for V DETAIL FREQ, THICK and THIN (and on the DVXIOOAIB there's a third setting, MID).

The reason this setting exists is because regular television sets display an interlaced picture. Cathode Ray Tube (CRT) televisions don't display an entire frame all at once; rather they divide the frame up into even lines and odd lines (called "fields"), and display each field sequentially (i.e., the television draws all the odd lines, then starts over at the top and draws all the even lines, then draws the next field of odd lines, etc.) Televisions use glowing phosphors to display information, and once the phosphors are triggered, the brightness starts to decay and darken. If an image consists of thin horizontal lines, lines that only show up on one field and not the other, those lines can flicker very noticeably and very annoyingly.

To compensate for this flickering artifact, all interlaced cameras employ a slight vertical blur, smearing data across fields to ensure that the image will be updated each field, thus minimizing the flicker that can happen when an object is drawn only on one field. The drawback to this method is that it results in a significant lowering of resolution: interlaced cameras are capable of a maximum of about 360 lines of resolution in NTSC, even though the DV frame is capable of holding a full 480 pixels of resolution in NTSC.

Newer display devices are not interlaced: plasma and LCD TVs, for example, display full frames rather than dividing the frame up into alternating fields. So do computer monitors. And when footage is transferred to film, there's no interlacing in a film frame either.

The DVX is designed to work with regular interlaced televisions, and also with progressive-scan displays such as plasma or LCD televisions, computer monitors, and transfers to film. In THIN mode the DVX can actually create video that is higher-resolution than an interlaced television can safely display. THIN mode produces the highest resolution possible, but interlaced televisions aren't designed to display that much resolution, so the image may suffer distracting flickering. Progressive-scan devices like computer monitors or LCD televisions have no problem displaying the high-resolution lines.

66

THICK mode is designed to minimize the flickering on an interlaced television. THICK mode employs the same type of blur filter used in interlaced cameras, lowering the resolution and blurring the lines across fields so there won't be flickering on an interlaced TV.

If a video is intended to be displayed on an interlaced TV, THICK mode is probably the right choice to shoot in. If the video will be transferred to film, or displayed on a computer screen or an LCD or Plasma television, THIN mode will deliver much higher resolution.

The DVXI00A and DVXI00B introduce a compromise setting as well, MID. MID retains much of the resolution of THIN, but lessens the flicker somewhat. Horizontal lines are blurred a little bit vertically, but not as much as THICK does. The result is a compromise that yields high resolution, sharp contrast and not much flicker, but there is more flicker in MID than in THICK.

When shooting on a DVXI00A or DVXI00B in Squeeze Mode, you should always use THIN line detail. The highest-resolution results will come from using THIN line detail. Squeeze mode will not flicker on an interlaced television when using THIN detail because the very process of stretching the footage (to turn it into 16:9 widescreen footage) introduces the same type of horizontal-line vertical blur as THICK mode. This means that using THICK and Squeeze together will result in an unnecessary loss in resolution; the best results with Squeeze mode will come from using THIN.

Also, if you're unsure whether to use THIN or THICK, and you are shooting for perhaps multiple display devices (an example: you intend to transfer to film, and you also want to have a DVD version for release on interlaced televisions); the general advice is to shoot with THIN line detail. THIN preserves the most options, because THICK can be simulated in post, but if you shoot THICK, there's no way to simulate THIN in post. Shooting THIN gives you a master copy with the highest resolution possible (suitable for your potential film transfer or up-rez to high-def), and if the footage flickers on an interlaced television, you can fix that in most editing programs by employing either a "Reduce Interlace Flicker" option, or a very, very mild gaussian blur in the vertical direction.

67

PROGRESSIVE

The DVX was revolutionary for being the first prosumer camera that records at 24 progressive frames per second (24P), in addition to regular 60-field interlaced video (60i). In PAL, it's 25 frames per second and 50-field interlaced video. By changing this menu setting you can choose one of three different shooting rates: either 24P, 30P, or 60 fields per second (in PAL, either 25P or 50 fields).

- The difference between 24/25 progressive frames per second and 60/50 fields per second is startling - 24P looks much more like film, and interlaced looks like video. In the NTSC version of the camera you can also select 30 progressive frames per second: 30.P is somewhat of a middle ground - it looks somewhat filmhke, but the

faster frame rate provides less strobing and smoother pans than 24P. However, 24P can be transferred easily to film or PAL television, whereas 30P cannot. 30P can be used to make for extremely smooth, mild slow motion in a 24P project as well.

When PROGRESSIVE is set to OFF, the camera is set to interlaced video mode. In interlaced mode the camera captures 60 (NTSC) or 50 (PAL) fields per second. Each field is equivalent to half of a video frame either the even lines or the odd lines. Normal televisions are

,

interlaced, and when PROGRESSIVE is OFF, the DVX captures interlaced video just like a regular video camera.

When PROGRESSIVE is set to 24P, 24PA, 25P (PAL) or 30P, the camera doesn't capture fields, it instead captures an entire full-height frame all at once similar to the way a film camera works. Each frame is captured distinctly, and then divided into fie.lds for recording on tape (but non-linear-editing programs can reconstitute the fields back into their full progressive frames).

In the PAL version of the camera, all progressive scan is done at 25P. In the NTSC camera, there are two options for 24P: either 24P Normal (also called 24PN), or 24P Advanced (24PA). There are advantages to each, and they're discussed in more detail in the essay on page 32.

68

The PROGRESSNE setting has a few restrictions. For one, it's the only setting that doesn't change when you change the scene file dial w~ile recording. All other settings (such as Matrix, Color Temp, etc.) will change when you rotate the scene file dial to a new scene file, but the PROGRESS!yE setting will not change unless you stop recording first. Also, the time-lapse Interval Recording feature won't work in 24P mode, although it will work in 30P mode (2SP PAL). And V DETAIL FREQ cannot be changed when in PROGRESSNE OFF mode: it's .always set to THICK, even though the menu setting may say THIN m blue text. 60i/SOi video is always recorded in THICK.

Notes on progressive scan: On the original DVXI00 in progressive mode, the color bar signal cannot be displayed, the gain is always fixed at OdB, and the auto focus cannot be utilized. When the p~ogressive mode ~s first ~elected, the sync signal may be temporarily disturbed and no video will be output for 3 seconds. A shutter setting of lIS0 (OFF) or 1160 is recommended. When using ATW (Auto Tracking White) function the black balance will not function. Since the video is recorded in S frame increments in 24P and 24P A modes the timing at which recording starts may be slightly delayed. . '

NAME EDIT

You can name the scene files and save that name. Select this option and you can use the menu joystick to spell out an individual name for your scene files. This name will be "lost" when you power off the camera, unless you use the SA VEIINIT menu option.

SAVEIINIT

When you change settings in a scene file, they're only preserved until you turn the camera off (or switch to VCR mode). Once you turn the camera off, all settings are re-set to whatever their last saved state . was. If you like your settings and want to keep them permanently, you must save them by using this setting. You have to save each scene file individually. Once you've saved your scene file, it will always hold your saved settings regardless of whether you turn the camera off, or remove the battery, etc. However, you can restore the factory original settings at any time, by using the INITIAL function.

69

CAMERA SETUP SETTINGS

The rest of the menu settings apply to the camera regardless of what the Scene File dial is set to. These settings won't change when you change scene files.

SYNCROSCAN

The DVX has a wide variety of shutter speed settings, from 1124 up to 1/2000 of a second. But by going into the SYNCRO SCAN menu, you can specifically dial in exactly what shutter speed you want, in increments as small as 111000 of a second. The SYNCRO SCAN function is useful for getting alternate shutter speeds, but it's perhaps most handy for when you're shooting a scene with computer monitors in it. Computer monitors frequently have refresh rates that may not synchronize with the camera's standard shutter speed, and that can result in seeing a dark bar on the computer monitors in your shot. To use the SYNCRO SCAN function, first use the SHUTTER and SPEED SEL buttons. Make sure that SHUTTER is not set to SHUTTER OFF, and then scroll through the SPEED SEL button until you see a decimal fractional speed, such as 1136.3 or 1148.0. Those speeds with decimal points are the SYNCRO SCAN speeds. Once you have SYNCRO SCAN speeds selected, you can go into the menu and choose SYNCRO SCAN and change the shutter speed. Each click of the joystick changes the speed by one increment. If you're trying to sync with a computer monitor, point the DVX at the computer monitor and then scroll through the shutter speeds until you fmd one that makes the monitor stop flickering. For television screens in the USA, you'll normally want to use a shutter speed of 1/60, and in PAL territories you may want to use 11100. For computer monitors you will likely have to use the SYNCRO SCAN menu to find a speed that stops them from flickering.

ASPECTCONV

There are two aspect ratios that are common in the world of television: regular 4:3 televisions, and widescreen 16:9 televisions. These numbers refer to the width of the picture as compared to its height; a 4:3 television is four units wide and three units tall, whereas a 16:9 television is 16 units wide and 9 units tall (for comparison, a

70

4:3 television could also be expressed as a 12:9 television, which gives a better understanding of how it compares to a 16:9 television.) Displaying 4:3 video on a 16:9 television will result in video that's horizontally stretched out. Displaying 16:9 video on a 4:3 television will result in a picture that's horizontally squeezed together.

The DVX is a 4:3 camera. Its CCDs (Charge Coupled Devices, the image sensors in the camera) are shaped 4:3. The DVX is optimized to create video suitable for a 4:3 television. When you use the NORM setting you get a 4:3 picture, utilizing the entire CCD.

Many people, especially aspiring indie filmmakers, prefer a more wide screen view, such as you get when you watch a letterboxed movie on a 4:3 television. The DVX provides a LETTERBOX setting for this look. LETTERBOX gives you video that is suitable for display on a 4:3 television, but has black bars on the top and bottom, giving the video a widescreen look like a letterboxed movie.

The DVXlOOA and DVXlOOB add a new setting, SQUEEZE. SQUEEZE actually changes the shape of the video, making it suitable for display on a widescreen television. There are no black bars on the video, the video is digitally stretched to display properly on a widescreen television. SQUEEZE mode uses the central 75% of the CCD (the area that would be between the black letterbox bars if the camera was in LETTERBOX MODE) and then digitally stretches it vertically to fill the full surface of the CCD. When this video is played on a widescreen television, the wider screen stretches the video horizontally, un-squeezing it and giving a proper widescreen picture. SQUEEZE mode is not suitable for display on a 4:3 television, as the image will look tall and skinny and will be lowerresolution than regular 4:3 video.

When choosing which mode to use, you should first decide what type of device your video will be played back on. If it will be played on a 4:3 television, and you want to use the full screen, select ASPECT CONY: NORM. If your video will be played on a 4:3 television, but you like the widescreen look, shoot in LETTERBOX mode. If your video will be displayed on a widescreen 16:9 television, you can shoot in SQUEEZE mode. If you want the highest resolution

71

possible and intend to be transferring to film or displaying on a widescreen TV, you can use the Anamorphic Adapter in NORM.

Also, when using SQUEEZE mode, always use THIN line detail. Using THICK or MID just wastes resolution in SQUEEZE mode.

When using the Anamorphic Adapter, always use either NORM or LETTERBOX mode. If you want a 16:9 widescreen image from the Anamorphic Adapter, use NORM mode. If you want a 2.35:1 "scope" image, use LETTERBOX with the Anamorphic Adapter. Do not use the Anamorphic Adapter and SQUEEZE mode at the same time: the result will be a doubly-squeezed image that can't be displayed on any television set.

When using SQUEEZE mode or the Anamorphic Adapter, the image in the viewfinder and LCD will appear squeezed on a DVXI00 or DVXI00A. On the DVXI00B you can choose to change the aspect ratio of the LCD and viewfinder so that it displays the unsqueezed, widescreen image properly.

COLOR BAR

TheDVX is one of the very few lower-cost cameras that can display actual SMPTE (Society of Motion Picture and Television Engineers) color bars, suitable for use in calibrating a professional monitor. When hooking the DVX up to a professional monitor, you can calibrate the monitor to know precisely what the recorded image looks like. All too often shooters will try to judge color, contrast, saturation or other picture elements based on how the image looks on a television, or on the camera's viewfinder or LCD. Those are not accurate representations of what the recorded image really looks like. The only way to know exactly what the image looks like is to use a professional monitor, or software such as Serious Magic's DV Rack, which provides a professional monitor emulator. Once they are calibrated correctly, informed decisions can be made on picture adjustments and settings. On the DVXIOO, COLOR BAR will not work if the camera is set to shoot progressive-scan video (24P, 25P or 30P). On the DVXI00A and DVXIOOB, COLOR BAR will work regardless of what video mode the camera is set to.

72

SETUP (NTSC ONLy)

Setup is a very confusing subject for many users. It refers to the black level of video, and the fact that in North American NTSC television, "black" isn't necessarily actually "black". In television signals brightness is measured in IRE units, using a scale from 0 (super-black) to 110 (super-white). In this system, "black" is actually pegged at 7.5 IRE, not at zero. The reasons are arcane, having to do with inaccuracies in picture reproduction on the earliest televisions - the TV engineers "fudged" a little bit, raising the level at which "black" would be perceived. With "Setup" added, the black level of the video gets lifted up, which gives some headroom at the bottom end of the scale, but also results in darker elements of the picture looking "milkier". This is akin to raising the Master Pedestal of the camera; in fact, the two are quite tied together: the 7.5 IRE setting is referred to as the Pedestal, and the Master Pedestal control lets you raise or lower that Pedestal setting, which gives you control over how dark/black areas of the picture are processed.

However, what's necessary to understand about setup is that the 7.5 setting is really only relevant for analog signals. In digital recording, such as with the DVX, all processing should be done with the menu setting set to zero. There is no such thing as "setup" in digital recording. In digital recording, the full dynamic range is used, and "black" is properly represented with no setup added. In a proper NTSC system, setup should never be part of a digital recording, or of digital editing, and should only ever enter the picture when making a final dub ofthe finished video over to an analog format (such as BetaSP) or when broadcasting the signal. In top-dollar broadcast cameras, the Setup option only exists as a menu option for analog output of digital video - broadcast cameras don't modify the digital recording to add setup, because adding setup to a digital recording results in a non-standard digital recording with the black level set too high. In a properly configured digital studio, the digital cameras would record with no setup, editing would be done with no setup, and mastering would be done with no setup. Then, when the product is finished and ready for broadcast, the final product would have setup added to it to make it compliant for broadcast. High-end professional decks have a menu setting for setup that will allow you to specify

73

whether you want to add setup to the analog output, but.it doesn't modify the digital recording. That's the "right way" to process setup.

Unfortunately, changing the menu setting on the DVX to "7.5" actually changes the digital data as it's recorded! This is not what. you want. This is not "proper" setup. When you have the?5 setting engaged, the camera lifts all the black levels before recording. ~fyou then transfer this footage to an analog tape format for broadcasting, the analog format will add another dose of setup, and you'll end up broadcasting a product with very milky blacks and flat contrast. Definitely not what you want.

For most purposes, when recording digital video, you should always have the Setup setting at O. There is a time where the 7.5 setup setting does make sense, and that would be if you were us~ng your camera for a "live feed" to an analog deck. For example, If you were using the DVX as a camera-head only (i.e., not recording on tape) and you plugged the s-video port on the camera into a BetaSP deck, then setting the SETUP menu item to 7.5 would make sense.

For digital recording, leave the SETUP setting at zero. In a ho~e digital studio, all capture, editing, and output would be done WIth SETUP at zero. Then for broadcast, the broadcaster would add setup themselves. Or, if authoring a DVD, you'd want to author the DVD with zero setup, because North American DVD players themselves add setup on the output signal.

0% is the default setting for the DVX cameras. Digital :Video uses 0% setup (which is the ITU-R-601 Digital standard setting) and standard analog-recorded interlace video uses a 7.5% pedestal.

74

SW MODE Settings

SW Mode is a confusing name for this menu, but what it really means is, this menu controls what the physical switches on the camera do. SW is short for "switch." This menu controls how the physical switches on the camera affect its operation. The switches that are affected by this menu are: the Gain switch, the White Balance switch, the Handle Zoom switch, the Iris Dial, and the USERI and USER2 (and on the DVXI00A and DVXlOOB, USER3) buttons.

MIDGAIN

Before addressing what MID GAIN does, you should understand what Gain is. Gain is an electronic amplification of the video signal. In other words, it artificially makes the picture brighter. While brightness sounds good, you have to understand that the penalty for making it brighter is that the picture gets "noisier." Electronic noise is a byproduct of electronic gain, and the more gain you apply, the brighter your picture will get, and the noisier your picture will get.

Gain is measured in decibels, or dB. Zero dB means that no gain is applied, the picture is unmodified and no brightness or noise is added. Every 6 dB of gain amounts to doubling the brightness of the picture, so.6 dB of gain would make the picture twice as bright, or 1 f-stop bnghter. 12 dB of gain would be twice as bright again (or four times as bright as zero gain), and 18db is twice as bright as that (for a maximum of eight times as bright as zero gain).

On the original DVXI00, Gain only works in interlaced (50il60i) mode, and is always fixed at OdB in progressive-scan modes, regardless of the switch position. On the DVXI00A and DVXI00B

. ,

Gain works in all modes, but is limited to 12dB in progressive-scan

(and doesn't work at all ifusing the SLOW SHUTTER mode).

The DVX has a physical switch for electronic gain, with three positions: L, M, and H (for "low", "mid" and "high"). The "L" position is always set to 0 dB. The 'M' position can be set to 0 3 6

, , ,

9, or 12 dB. This MID GAIN menu setting lets you select what Gain value to assign to the 'M' position.

75

HIGH GAIN

Same as MID GAIN except that it applies to the "H" setting of the Gain switch. 18dB of gain is only possible through the USER buttons.

ATW

To understand the A TW (Auto Tracking White) function, one should first understand the concept of color temperatures and white balance. In the simplest explanation, light is not all the same color. Even though it may look the same to the human eye, the camera sees light for what it is: reddish, greenish, blueish. Light color is measured in degrees Kelvin, in accordance with what color a hunk of iron will glow when heated to certain temperatures. When heated to about 3200 degrees Kelvin (or 3200K), the iron will glow an orangish-red color (which is pretty much how regular household lamps work: they're small filaments of metal that are heated until they glow that orangish-red color). If the iron is heated more, the color will shift towards the blues, and at 5600K the iron will glow blue-white. These colors correspond to the colors on the rainbow, with red at one end of the spectrum, and blue at the other. These temperatures, and their corresponding colors, are referred to as "color temperature."

In general there are two color temperatures you need to be aware of: 3200k and 5600k. Daylight is typically said to be around 5600k (even though daylight can vary from about 3000K to about 10,000k). And tungsten (or most artificial) lights burn at around 3200k. Our eyes adapt to "know" when something should be white, regardless of what color light is shining on it. The camera sees these colors for what they are. Because of this, we have to tell the camera what "white" should be, by using the White Balance switch.

The White Balance switch offers three basic modes of operation: preset, manual set, and automatic-tracking. In Preset mode, pressing the white-balance button on the front of the camera will toggle between 3200K and 5600K settings. In manual-set mode, you point the camera at the scene, filling the screen with something white in the scene, and hold the button down. The camera will analyze what it's looking at, and figure out how to manipulate the colors so that the object on the screen will actually look white. In automatic-tracking white (or ATW) mode, the camera continuously evaluates the scene

76

and continuously adjusts in a never-ending search for the proper white balance.

The A TW menu setting lets you choose a position on the switch that will trigger the automatic white balance tracking. The A TW is not assigned to this switch by default; normally you have two separate white balance settings you can keep in-camera (A Channel and B Channel), but this menu setting lets you assign ATW to one of those switch positions or to the PRST (Preset) position.

The white balance should typically be re-adjusted as the lighting conditions change. The Auto Tracking White function can be used during the record mode to continually adjust the white balance automatically.

HANDLE ZOOM

The zoom motor on the DVX operates at three speeds: low, mid, and high. The DVX has two zoom rockers that control the zoom motor, one above the record switch, and one on the top of the handle. The one above the record switch is variable pressure: the harder you press it, the faster the zoom goes. But the zoom rocker on the handle only operates at one speed regardless of how hard you press it. It operates at the speed that's set by the 112/3 switch on the camera's handle.

The rocker on the handle has a control switch that selects the speed that will be employed when the handle zoom rocker is pressed. The "I" setting is always the slowest speed, and the "3" setting is always the fastest speed. This menu setting controls what happens when the 112/3 switch is set to "2". It can either be set to "LlOFFIH", which disables the handle zoom (when the 112/3 switch is set to "2") or it can be set to "LIMIH", in which case the handle zoom will run at

1 =L, 2=M, 3=H speed.

IRISDIAL

This menu setting controls the relationship between the direction the iris wheel is turned, and the opening/closing of the iris. You can set it so that the iris opens up when the dial is rotated down, or so the iris closes down when the dial is rotated down.

77

USERl and USER2 (and USER3 on the DVXlOOA and DVXlOOB) The DVXI00 has two USER buttons on the side of the camera. The DVXI00AIB adds a third, USER3. These buttons can each be customized to perform one of many different functions, including:

COLOR BAR: when this setting is assigned to a user button, and that user button is pressed, the DVX will generate SMPTE color bars instead of camera footage. However, if you have an original DVXI00, this function will only work in 50il60i interlace mode, it will not work in progressive-scan mode. On the DVXI00A and DVXI00B, this function will work in any mode.

SPOTLIGHT: This function is similar to A. IRIS LEVEL -3.

Pressing this button will cause the iris to stop down about one sto~. If the camera was set in manual iris mode, pressing SPOTLIGHT WIll force it to go to auto-iris mode.

BACKLIGHT: This function is similar to A. IRIS LEVEL +4. Pressing this button will cause the iris to open up about two stops. If the camera was set in manual iris mode, pressing BACKLIGHT will force it to go to auto-iris mode.

BLACKFADE: Pressing and holding this button will cause the picture and sound to fade to black. The longer you hold the . button the more the picture will fade: you have to hold the button m until the' picture is entirely black. For the next shot, if you want ~o execute a fade in from black, you would need to hold the button m while the camera is paused, until the screen is fully black. Then start recording and release the button, and the image will automatically

\ fade up from black.

WHITEFADE: same as BLACKFADE, except it fades

to/from white.

ATW: Automatic Tracking White balance: see the discussion

about ATW in the SW MODE switch settings above.

ATWLOCK: when using ATW, the system will constantly hunt for what it considers the optimal white balance. You can stop the camera from hunting and tell it to lock in the current white balance setting by using the ATWLOCK function. If you use

A TWLOCK to lock in the white balance, and then press ATWLOCK again, it will un-lock and return to hunting in Any mode. .

GAIN: l8dB: The SW MODE switch settmgs for gam allow for setting electronic gain as high as 12dB. The DVX is actually capable of going as high as 18dB of gain, but the only way to get to

78

18dB is to use the USER buttons. 18db will brighten your image by about 3 stops (in other words, it'll be about 8 times as bright as without gain), but 18db is quite noisy and should only be used when absolutely necessary, and when image noise is not a concern. 18dB Gain will not work in progressive scan on any DVX camera.

. INDEX (DVXIOOA, and DVXIOOB only): Assigning this ~ctlOn to one of the USER buttons will cause an Index signal to be ~tt~n t? ~he tape whenever this button is pressed. The index signal IS an InVISIble marker that you can use to quickly locate a particular segment on tape. Using the remote you can perform an "index search", which will fast-forward or rewind the tape until the camera senses the presence of an Index signal. This can be a way to mark "good takes" or especially interesting moments during recording. The DVXlOO has this same functionality, but it's permanently tied to the INDEX button rather than being an assignable menu item.

SLOWSHUT (DVXIOOA and DVXIOOB only): The DVXlOOA and DVXI00B include the ability to set slow shutter speeds, slower than the capture rate. First you have to engage the slow shutter mode by assigning SLOWSHUT to one of the USER buttons and then pressing that USER button. Then you can cycle . through the slow shutter speeds using the SPEED SEL button. Selecting slow shutter speed mode will disable the camera's electronic gain circuitry: you can have a slow shutter speed or electronic gain, but not both. When you push the USER button you have programmed to SLOW;SHUT, the camera will enter Slow Shutter Spee~ Mode, but if you see an INVALID message, this means your Shutter IS ON, thereby locking out the Slow Shutter function.

To get around this, go to Shutter OFF, and then press the programmed USER button to step through the slow shutter speeds.

79

AUTO SW Settings

The DVX has a handy button called "AUTO". When you press the AUTO button, the camera can override all sorts of manual settings and go into fully-auto mode: handy if you need to immediately grab the camera and start shooting something, with no time to adjust manual settings. The "AUTO" setting might help you guarantee that you get your shot, even if it's not the best shot. You can control which functions go into auto mode and which ones stay in manual mode. You can also implement auto picture gain through the use of the AUTO switch. This is the only way to get auto gain on the DVX.

A. IRIS: If you set this setting to ON, then when the "AUTO" button is pressed the camera will perform auto-iris functions, even if you have the camera set to manual iris. If this setting is OFF then the camera will only be able to use auto-iris if you had the camera in auto-iris mode before pressing AUTO: if you had the camera in manual iris mode, it will still be in manual iris mode after you press the AUTO switch (if this setting is OFF.)

A(;C: Normally the DVX only has manual gain control for the picture. In AUTO mode, the camera can actually apply gain when it determines that gain is necessary. This switch lets you control how much gain is allowed: Odb (OFF), 6db or 12db. The camera will decide how much gain to use, up to the maximum that you specify. Setting it to OFF will leave the camera with no ability to apply gain. This setting will have no effect if used on a DVXlOO in Progressive Scan mode, but will be fully functional in interlaced mode, or on a DVXI00A or DVXI00B in all shooting modes. Note that the gain being discussed is only for the picture; the DVX cameras have no audio gain circuitry.

ATW: Two settings, ON or OFF, determine whether the camera will be able to perform automatic tracking of white balance when you press the AUTO button. If it's set to ON, the camera will start automatically tracking white balance. If it's set to OFF, the camera will be forced to use whatever setting the WHITE BAL switch is currently set to.

80

AF: This setting determines whether the camera goes into Autofocus mode (or perhaps Focus Assist on the DVXI00A and DVXI00B) ~hen the AUTO button is pressed. If set to ON, the camera will go mto auto focus mode, regardless of whether the FOCUS switch is set to Auto, Manual, Push Auto or Infmity. None of those switches will have any effect, because the camera will run in auto focus mode. This switch will not work on a DVXl 00 if the DVXI00 is set to record in progressive-scan mode, but will work in interlaced mode, and will work on a D~lOOA or DVXI00B in all modes. When set to OFF, the FOCUS switch controls how focus operates: ifit's set to Auto or Push Auto is held down, autofocus will continue to work (unless on a DVXlOO in Progressive-Scan mode, where autofocus will never work).

When used with slow shutter speeds, Autofocus is much, much slower to respond. Also, when used with a DVXI00A or DVXI00B in pr?gressive-scan mode, autofocus responds quite slowly (which is why It'S called "Focus Assist" instead of autofocus in those modes. Autofocus performs best in interlaced mode, under bright light con~itions,. with a 1148 or faster shutter. Autofocus performs less efficiently m progressive-scan mode, under lower light conditions, or when using slow shutter speeds.

81

RECORDING SETUP Menu Settings

REC SPEED: The DVX is capable of recording on the tape at two different speeds: SP (Standard Play) and LP (Long Play). In SP mode the DVX can record for approximately 63 minutes on a 63-minute cassette. In LP mode the DVX can record for approximately 94 minutes on a 63-minute cassette. The video quality recorded in both modes is identical: the exact same DV signal is recorded whether in SP or LP mode. LP does have drawbacks, however: a higher chance of dropouts, and less compatibility with other cameras/decks. In LP mode the tape moves slower past the recording heads, so the data gets recorded into a smaller space on the tape. By moving the tape slower, it takes longer to run through the camera, which gives you longer recording times. However, the data has to be fit into a physically smaller section of tape, which makes LP recordings much more subject to dropouts or tape error than SP recordings. Also, because LP recordings are so tightly crammed onto the tape, there is no guarantee that an LP recording made on one camera will be able to play properly on another camera or deck: you may only be able to play back an LP recording on the same camera that recorded it. Finally, LP recordings cannot use 4-channeI12-bit audio, they can only use 2-channeI16-bit audio. Most professionals recommend to avoid LP mode whenever possible, and stick with the more reliable (and more compatible) SP mode.

c

AUDIO REC: The DVX can record audio in two different formats,

either 12-bit or 16-bit. In 12-bit mode the DVX records two tracks of 12-bitl32kHz audio. 12-bit works okay for dialogue, but is not optimal for recording music. In 16-bit mode the DVX records two tracks of 16-bit, 48kHz audio. This is the highest-quality audio recording mode the camera offers, and is basically the same specifications as a DAT deck. If you record in 12-bit audio you can add two additional audio tracks to your tape in post (through the Audio Dub function). 12-bit audio is capable of storing four full tracks of audio, although they're all of the lesser-quality 12-bit 32kHz mode, and only two can be recorded live (the other two must be added in post). In high-quality 16-bit 48kHz mode, no audio dubbing is possible.

82

MIC ALC: The DVX camera does not have any sort of audio gain or automatic level control, but it does have a limiter. In other cameras "automatic level control" will try to raise or lower the volume of the recorded sound, depending on whether the camera thinks the audio is too quiet or too loud. In the DVX there is no such automatic level controL There is, however, an Automatic Limiter Control (ALC). When the Limiter is engaged, the camera will try to clamp down excessive volume to prevent clipping. It won't modify the overall signal level, it will just try to keep loud levels from distorting (sort of like the KNEE control for protecting against overexposure). The DVX's ALC starts attenuating signals at approximately-6dB, and tries to limit them to a maximum of -4.5dB. It doesn't catch brief transitory peaks (clapping close to the mic may defeat the limiter and result in overmodulated sound) but if the overall sound level is too high, the DVX's limiter will lower the volume to keep it below the maximum allowable threshold. When using mics directly hooked to the camera, it's usually a good idea to keep the limiter ON. When using an external mixer with its own limiter, you should set the DVX's limiter OFF.

MIC GAIN 1 (and 2): Different microphones have different levels of output. If you find that the audio level coming from your microphone is too low, try changing the MIC GAIN setting to boost the audio levels. -60dB makes the mic input louder, and would be appropriate for use with a less-sensitive microphone.

TIME CODE OPTIONS: Before discussing the timecode options, let's discuss timecode itself Time Code is used for accurate video editing of a recorded videotape. A special internal timecode generator (TCG) stamps an 80-bit code on every recorded frame. The VCR will use this number for individually identifying every frame. This code is recorded with the video and audio signals and is stored invisibly in the sub code area on the DV tape. This code contains the hour, minute, second and frame number of the particular frame it's stamped onto. These 80 bits of time code contain a lot of information, such as drop frame information, frame rate information and user bit information. Drop Frame timecode is used to keep accurate timing between the counted frame time and the actual runtime. The drop frame bit is bit number 10, this bit identifies if the

83

drop frame or the non-drop frame format is being use~. If it is set to "I" the unit is placed into the drop frame mode and will account for the missing frame numbers which are "dropped" to account for the difference in timing between "real time" and ''video time." If it is set to "0" it will be placed into the Non-Drop Frame (NDF) .mode~ where every frame is counted individually, but counted frame time will not match "real time." The 24p mode uses NDF only. There are 32 User Bits in each 80-bit code; these user bits can be used to identify the user information, the time, the date, the time code generator's value,

or the frame rate.

1394 TC REGEN: This menu setting is applicable only when the camera is set in VCR mode, and is recording data from the IEEE- 1394 port. This setting allows the user to clone a DV tap~ from. another camera or deck, and (depending on this menu settmg) WIll either generate new timecode on the new tape in the D,?" or duplicate the timecode from the original tape. Wh~n this me~u setting is set to OFF, the DVX will gene~te new tI~eco~e usmg the values set in its menu settings. When this menu setting IS ON, the DVX will duplicate the timecode it finds in the 1394 data stream. If you hook the DVX up via IEEE-1394 to a deck (or other camera), and-set that deck or camera to "play" and the DVX to "record", the DVX will be able to make a tape that duplicates the timecode on the original tape when this menu setting is ON. Note: on. the DVJ;Cl OOB, setting 1394 TC REGEN to ON will preclude the o~t~on ofusmg the DV IN PRESET function. If you want to use the ability to sync timecode through the 1394 port, you must set this option to OFF.

TC MODE: This menu setting allows you to select Drop Frame (DF) or Non Drop Frame (NDF) timecode. DF and NDF are numbering methods that count the number of frames that have been recorded. They don't affect the recording in any way (i.e., "Drop Frame" doesn't actually drop any frames) but they do affect how the time of recording is counted. Timecode is a system that numbers and counts every frame of video, in the format ofHH:MM:SS:FF (hours: minutes:seconds:frames). In NDF mode, every frame gets counted and numbered sequentially. In DF mode, some timecode entries are skipped in order to make the running ~ime of the v~deo match the timecode display (by way of explanation, NTSC VIdeo runs at 29.97

84

frames per second, but timecode counts at 30 frames per second. Drop Frame counting was invented to resolve the .1 % discrepancy, so when an hour of footage has gone by, the timecode will read

1 :00:00:00, whereas in NDF timecode, after one hour the timecode would read 0:59:56: 12.) When in 24P/25P modes, NDF timecode counting is used regardless of the setting of this menu item. The menu itemwill turn blue and be non-selectable, and it may say DF or NDF, but regardless of what it says, it will be operating in NDF mode. When shooting in 24P or 24PA, you need to be aware that the camera operates in groups of four frames which get written to tape as five frames (after 2:3 pulldown, or 2:3:3:2 pu1ldown). The camera will always start recording on the A frame, the first of the five frame sequence. When you stop the recorder, it may not stop on the last part of the D frame (last frame in each five-frame sequence) and so when you back up to review, the camera may leave a gap between the last recording and where it starts its new recording. The way to address this, if you review your footage as you go, is to leave an extra-long "tail" of footage before you stop the camera, or just record a second of color bars before you stop the camera. Then you can review and park the camera over the "tail" or color bar area and then the camera will be able to find the end of the D frame so that it can continue the sequence, cutting in the new A frame just after the lastrecorded D frame. If you are reading this book and you have the original DVXI00 version of the camera, you can't record color bars in 24P mode, so you would just leave an extra-long "tail" of footage or just record a second of black after the take is fmished. If you do not otherwise use the User Bits, do your editor a favor and set up the User Bits to record the Frame Rate. This way he/she will be able to tell ifthe tape is 24P or 24P A.

TCG: TCG stands for Time Code Generator, and this menu setting determines what kind oftime code generator is employed: FREE RUN or REC RUN. In FREE RUN mode, the timecode clock is constantly advancing whether the tape is recording or not. In REC RUN mode, the timecode clock advances only when the tape is recording. REC RUN mode can make a tape with continuous timecode regardless of how many times you start or stop the camera: the timecode will always pick up where the last shot left off. In FREE RUN mode, the timecode is derived from the clock of the

85

camera regardless of how many times you start or stop ~he tape .. FREE RUN mode can be used to synchronize cameras m a multicamera shoot: for more details, see the essay on page 55).

FIRST REC TC: This menu setting determines how the camera will start timecode when it starts a new tape (or encounters a blank spot on a tape). The choices are REGEN or PRESET. In ~GEN mod~ ~e camera will try to regenerate (i.e., pick up and contmue) the existing time code and continue uninterrupted. If there is no time code, REGEN mode will start new timecode at 0:00:00:00. This is the default menu state: every new tape put in the camera will start at 0:00:00:00 timecode. With PRESET, the camera will not try to pick up existing time code; instead it will start each recording with the . timecode setting held in the TC PRESET menu setting. By changmg the setting in the TC PRESET menu you can ch~ge what starting timecode is assigned to each tape. Every tape will have contmuous timecode thereafter. The camera keeps track of the TC PRESET and advances it automatically, so it will always pick up where the last time code left off, even if changing tapes.

TC PRESET: This menu setting works only in conjunction with FIRST REC TC. If FIRST REC TC is set to REGEN, this menu setting has no effect. If FIRST REC TC is set to PRESET, the~ this

is the value that will be used as the starting timecode the next tune you start recording. This setting gets updated as the timecode runs,

so the timecode on the next tape will continue where the last tape left off (i.e., it won't be reset to what this setting was first set to). By using TC PRESET, you can make a series of tapes wi~h diffe~ent timecode starting values for each tape. For example, If shootmg an all-day lecture, and planning on using 8 tapes throughout the day, you could use the TC PRESET to set the first timecode to 00:00:00:00, and begin taping. At the one-hour mark, stop the.ca~era and change tapes. The next tape will start with timecode beginning where the last tape left off (approximately 01 :00:00:00), and each subsequent tape would have timecode that begins where the last tape stopped. If instead you were to use REGEN, then every time you changed tapes, the timecode would start back at 00:00:00:00 (unless of course you were using FREE RUN timecode, in which case every tap~ would have unique continuously-running timecode based on the time-of-day

86

clock). Note: When the 24P or 24P ADV modes are used you must set the frame value to 0 or a multiple of 5. If you use any other value the time code recorded will shift.

1394 UB REGEN; This function is available only when the camera is in VCR mode, and you are recording information on the DVX's tape from a signal on the 1394 port (i.e., you're using the DVX to clone a tape from another camera or deck). To understand it, take a look at the description for 1394 TC REGEN. This function is very similar, except that instead of controlling how the camera records timecode, it controls how the camera records User Bits in the DV tape's sub-code track. When making a clone of a tape through its 1394 port, the DVX can either clone the source tape's User Bit information, or generate new User Bit information (according to the DVX's UB MODE menu setting). When this setting is set to OFF, the DVX generates new User Bit information. When this setting is set to ON, the DVX will clone any existing User Bit information from the 1394 data stream attached to the DVX's 1394 port.

UB MODE: This menu setting controls what type of information . gets recorded in the User Bits on the DV tape's sub-code track. The User Bits allow you to specify a message to encode invisibly in your video stream. That message can be one that you specify for each camera, or it can be one of a number of pieces of technical information. The options are:

USER: this setting lets you specify an 8-character long message that gets recorded invisibly in the tape's sub-code track. Using the numbers 0-9 and the letters A-F ("hexadecimal notation"), you can write an 8-character-Iong message. Example messages might be the name of your company (assuming you can spell it using just the letters A,B,C,D,E, and F!) or perhaps you'd want to specify a unique identifier code for each camera in a multi-camera shoot. .. that way, in the editing bay, the editor could immediately identify which camera shot which tape. The message that you write is set in the UB PRESET menu item.

TIME: The camera's time-of-day clock gets recorded in the user's bits.

87

DATE: The date from the camera's clock gets recorded in the user's bits.

TCG: The camera records the contents of the time code generator. This would give you two copies of time code on the tape, which is redundant.

FRM RATE: In this mode the camera will record what frame rate the camera is in when it shot the footage. In 24P and

24P A mode, the readout will look something like "UB F2 04 2481". In 60i mode, it will look something like "UB 27 OF 6003". In 30P mode, it will look something like "UB 94 OF 3082".

The first two letters are always "UB", meaning User Bits.

Next are two characters that you can ignore, they're used for "User's Bit value verification information." In the examples above, those would be the "F2", "27", and "94".

The next two characters depend on whether the camera is in 24P mode or not. In 24P mode, these two characters are a constantly cycling pair of numbers from 00 - 01 - 02 - 03 - 04. Those numbers are counting the individual frame within a five-frame grouping. When recording 24P footage, the camera maps four progressive-scan frames onto five interlaced frames. This counter shows where the frames exist within each five-frame block. For example, when recording 24P A mode, the frame enumerated "02" will always be the interlaced "split frame". When in 25P/30P or 50i/60i mode, these characters are always "OF".

The next two characters show the frame rate. In 60i, they'll be "60"; in 30P they'll be "30"; in 24P modes, they'll be "24".

The last character is a constantly-changing digit that toggles when the camera is paused, and advances when the camera is rolling.

UB PRESET: This setting works in conjunction with the UB MODE setting of "USER". When the UB MODE setting is set to "USER",

88

the camera will record the contents ofUB PRESET in the tape's subcode track. You can encode a specific message or identifying tag into your camera, so that all tapes shot with that camera will bear that message. This could be useful in a multi-camera shoot, to identify which camera shot which tapes. The contents of the UB PRESET message can be 8 characters long, and those characters are encoded in hexadecimal notation (which means that you can only use numbers from "0" to "9" and letters from "A" to "F" to make up the message).

INTERVAL REC (DVXIOO and DVXIOOA only) and ONESHOT REC (DVXIOOB only): The DVXIOO and DVXIOOA are capable of performing interval recording, which can be useful for time-lapse photography (such as photographing buildings under construction, clouds moving across the sky, or flowers opening up) or for stop-motion animation. There are two components to establishing an interval recording: how frequently the camera shoots, and for how long it shoots. You can set the camera up to just record for a fixed set of time whenever you press the "REC" button, or (on the DVXIOO and DVXI OOA) you can set it up so that it will automatically trigger itself to record for a brief burst offrames every so often. You camiot use 24P modes to record in INTERVAL REC: you have to use 25P/30P or 50il60i.

There are two options: ONE-SHOT or ON (the DVXIOOB only allows for ONE-SHOT). In ONE-SHOT mode the camera will start recording when you press the REC button, and will continue recording for the duration set in REC TIME, after which it will automatically stop. This is useful for stop-motion animation: set the camera in ONE-SHOT mode, set the REC TIME to 0.5s, and press RECORD. The camera will shoot haifa second worth offootage (IS frames) and then automatically pause. You can then go in and move your subject a little bit, and press RECORD again, and the camera will record another half-second of footage. Repeat until you've fmished your sequence. Then, you can import the footage into your editing program and run it at 15x speed, and you'll have frame-forframe accurate stop-motion animation. Always use the remote control to trigger the camera when doing stop-motion animation, as you might bump the camera if you try to push the camera's record buttons, which will disrupt your stop-motion framing.

89

For the DVXlOO and DVXIOOA, the ON mode will cause the camera to automatically trigger itself to record bursts of footage. It works the same as ONE-SHOT mode except that the camera will trigger itself to start recording, at intervals specified in the INTERVAL TIME menu setting. Using this mode you can record time-lapse photography.

REC TIME: When using INTERVAL REC mode on a DVXIOO or DVX100A, or when using ONE-SHOT REC on a DVXI OOB, this setting controls how long each burst of recording will be. There are four settings, either half a second, one second, one and a half seconds, or two seconds.

INTERVAL TIME (DVXIOO and DVXIOOA only): When using INTERVAL REC mode, this setting determines how long the camera will wait between automatically triggering bursts of recording. You can set the camera to start recording every 30 seconds (IS seconds on the DVXIOOA), every minute, every five minutes, or every 30 minutes.

90

A V IN/OUT SETUP SETTINGS

This menu is only active when the camera is set to VCR mode, and deals with settings related to the RCA (red-yellow-white) and s-video input/output jacks.

A.DUB INPUT: The DVX offers an audio dubbing feature, which allows you to add narration or additional sound to a tape after it's already been recorded. This feature is ONLY possible if you recorded the original recording in 12-bit, 32KHz audio mode, and SP video mode. Audio dubbing is not possible if you used 16-bit/48KHz audio, or if you recorded in LP mode. When dubbing, this menu lets you select where the source of the dubbed audio will come from: either from the camera's built-in microphone, or from the red and white RCA audio in/out jacks.

DV OUT: This setting determines whether the DVX functions as an on-the-fly analog-to-digital video converter. When set to ON, the video and audio signals from the analog inputs (the red, white, and yellow A/V connectors and the S- Video jack) will be converted to DV video and output on the 1394 port. Using this menu setting you can use the DVX to convert analog video to digital, which could let you digitize analog sources into your 1394-equipped non-linear editor without having to record to tape first. When importing or converting footage through the analog video inputs, understand that it is not possible to convert that footage to 24P in the camera. The 24P mode only works on footage shot through the lens. Any dubs of footage will be copies of what the footage already looks like; no frame rate conversions are possible when the DVX is in VCR mode, and scene file settings and other options are not used when dubbing footage.

91

DISPLAY SETUP SETTINGS

This menu controls the text and graphics that show up on the camera's LCD monitor and viewfinder, as well as whether that information shows up on the analog video outputs. (Note: there is no way to get this information to show up on the 1394 video output).

ZEBRA DETECT 1 and 2: The "zebras" are an exposure guide you can use to judge the overall exposure level of your picture. When you enable the zebras, the camera will display a diagonally-striped black & white line pattern over sections of your video that exceed a specified brightness level. Using the zebra pattern will let you know what overall brightness levels your video is at, and help you to get proper exposure and more dynamic range out of your video. The zebras can be set to 80%, 85%, 90%, 95%, and 100%, and in the DVXI00A and DVX100B they can also be set to 105%. While this sounds impossible (how can brightness be at 105%?) keep in mind that the percentages are just a naming convention. Video brightness is measured in IRE units, with pure white registering at 110 IRE, so setting zebras at 105% means having the zebras trigger at 105 IRE.

Th~reare two zebra settings, Zebra 1 and Zebra 2. They display different diagonal patterns on the LCD so you can tell them apart:

Zebra 1 draws its lines from the upper-right to the lower left; Zebra 2 draws its lines from the upper-left to the lower right. Common practice is to set Zebra 1 at 80% and Zebra 2 to the maximum value, 100% (or 105% on the 100A/B). Then use the zebras to guide your overall exposure level. Typically when shooting faces, you'd like to see the brightest spots on a Caucasian face (usually the highlights on the forehead and maybe the nose) just bright enough to trigger the 80% zebras, but nothing should be so bright as to trigger the 100%/105% zebras (unless you have something overly bright in the scene, like a light bulb or the sun or something that simply must be extremely bright). You use Zebra 1, at 80%, to guide overall exposure, and you use Zebra 2, 100%/105%, to protect against "blowouts". This is especially vital when using the CINEGAMMA mode, which offers no knee protection against blowouts. Blown-out video happens when the overall brightness is too high for the CCD to

92

be able to resolve it, so the video signal overloads the CCD and the blown-out portion burns to pure white (110 IRE), losing all detail. It looks ugly and should be avoided; the Zebras are a guide to letting you know when you are approaching blowout.

Of course, the zebra display is only drawn on the LCD and viewfmder, itdoesn't get recorded on tape in the video signal.

This menu item lets you set a zebra level (or OFF) for zebra 1 and 2. You toggle between the zebras (and the MARKER) using the ZEBRA button on the camera's body (66, or 54 on the DVXlOOB).

MARKER: The DVX offers an excellent tool for measuring the video level of different parts of your shot: the MARKER. The MARKER acts something like a lightmeter's spotmeter: it tells you the brightness of whatever's in the frame within the box in the center. This menu item lets you turn the MARKER ON and OFF.

To use it, press the ZEBRA button (66/54) to cycle through the Zebras until the Marker box comes up (it'll draw a white box in the" center of the screen). The MARKER measures the video level within that box and reports it using a numerical percentage readout in the lower left comer of the screen. The Marker readout is specified as a percentage of brightness, with 0% representing solid black, and 90% (on the DVX100) or 99% (on the DVXlOOA and DVX100B) representing the brightest possible luminance level. Note that these numbers are relative to the aperture and shutter speed that you currently have set: changing either the aperture or the shutter speed will affect the brightness level of the picture, and correspondingly all the MARKER values will change as well. The percentages the MARKER reports are percentages that are relevant to the shot that it sees at the time. If you change the light level or the iris or shutter speed, then the MARKER will of course report new numbers that reflect the new brightness levels the camera is seeing.

You can use the MARKER to see if the camera is discerning detail in the shadow areas of the picture, or to determine brightness ratios between key and fill lights, etc. Another excellent usage of the MARKER is to check for the overall evenness of lighting of a

93

greenscreen, for example. Using the Marker you can locate hot spots or dark areas that may not be easily discerned by eye, but will guide you towards creating a more overall evenly-lit surface, thus improving the quality of your greenscreen keys.

On the DVXlOOA and DVXlOOB a second, smaller box has been added to the display, so when MARKER is selected you'll actually see two boxes in the center of the screen. The smaller box represents the area the marker is sampling and reporting the brightness of

VIDEO OUT OSD: This menu item dictates whether the text and graphics in the LCD display also get displayed on the RCA/composite and S-video outputs. OSD refers to On Screen Display. If you set this menu item to ON, any text or graphics (menu displays, marker settings, level meters, etc) will also be displayed on a monitor or VCR connected to the analog video outputs.

DATEITIME: This setting is obvious, it controls whether the date, the time, or both are displayed on the LCD and viewfinder. However, it's important to note that this setting WILL be displayed on the composite video output and s-video output, regardless of what the VIDEO OUT OSD menu item is set to. It will not, however, get recorded on the DV video, like some consumer cameras do.

LEVEL METER: This menu setting governs whether the DVX's audio meters are displayed or not. The audio meters show the overall strength of the audio signal, for left and right channels. The display is a horizontal graph, mainly white dots but with red dots for louder signals, and two small vertical bars, one at the transition from white to red and one at the absolute peak of red. Each dot corresponds to 2 dB of signal strength. The transition point from white to red happens at -12dB. Absolute peak (the most red showing) is at OdB, and should be avoided - audio that is so loud as to "peak" will likely be distorted when recorded. The smallest level shown is -34dB. For normal audio recording you want to have a good strong signal, with your loudest peaks going into the red but not all the way to OdB. Many DVX users are leery ofletting the audio dip into the red section. It would perhaps have been more informational had Panasonic included a yellow section, perhaps from -12dB to --6dB,

94

but they didn't, so don't be afraid of the first few dots of red on the level meters. Just don't let your audio peak or it will clip and become distorted. You can use the ALC limiter to get some protection against peaking, but there is no provision in the camera for automatic audio level control (i.e., the camera cannot boost the audio levels during quiet portions, nor can it lower the audio level if the overall level is loud, other than the protection the ALC provides). You manually control the audio levels by using the two audio level potentiometers (30/36). The only automatic intervention is the limiter.

ZOOM - FOCUS: This menu item lets you display or hide the zoom and focus readouts. The DVX cameras feature percentage readouts that give you precise feedback about the mechanical position of the zoom and focus rings. With these readouts you get precision and control that are rarely found on cameras in this price range.

The zoom ring is purely mechanical, and because the zoom ring is an actual physical linkage to the lens, it's important to not try to tum the zoom ring when the camera is in servo zoom mode, or you risk damaging the camera. Only manually tum the zoom ring when the . zoom/servo switch (44/68) is in manual mode.

The zoom readout ranges from ZOO to Z99. The numbers let you know exactly where you are in the zoom range, which is especially handy for establishing focus: you establish your shot composition and make a note of what Z setting you're at, then zoom in all the way (Z99) on your subject to grab critical focus, and then you can zoom back out to the same zoom setting you started at. The zoom numbers are not a direct mathematical progression (i.e., while ZOO corresponds to 4.5mm, and Z99 corresponds to 45mm, Z50 doesn't correspond to 22.25mm! Halfway through the zoom range, mathematically, is reached at approximately Z81.) There are markings on the zoom ring that show what the exact focal length is at certain increments, but the camera doesn't directly report the focal length. For a direct relationship between Z numbers and the lens' focal length, refer to the following chart.

95

z DVX 35mm still
ghotogragh¥
eauivalent
00 4.5 32.5
05 4.8 34
10 5.2 37
15 5.8 41
20 6.4 46
25 7.0 50
30 7.6 55
35 8.2 59
40 8.8 64
45 9.5 69
50 10.3 74
55 11.3 81
60 12.5 90
65 14.1 101
70 16.1 116
75 18.7 135
80 22.1 159
85 26.4 190
90 31.8 230
95 38.5 278
99 45.0 325 The focus ring is not a pure mechanical linkage, but it does give you precise control over the focus mechanism and repeatable focus marks. The MFOO-MF99 readout lets you know exactly where the focus mechanism is at, and provides for easily repeatable rack focus moves (where you shift focus from one point to another). This type of control is simply not possible on any other DV -format prosumer camcorder, with the only exception being the Canon XL series when using an optional fully-manual lens.

The focus readout is displayed in numbers from 00 to 99, prefixed with either "AF" or "MF". "AF" means the camera is in auto focus mode, and MF means manual focus mode. MFOO means the lens is focused as close as it can possibly get; MF99 means it's focused at its furthest possible point (which is actually past infinity, as infinity focus is achieved at MF95). Panasonic chose a numerical readout rather than listing actual distances, because the numbers-todistance relationship changes depending on whether you have any additional lenses or close-focus diopters attached to your camera. For example, with no lens attachments, MF50 delivers proper focus for an object that's 3' 10" away from the focal plane (as measured from the rear of the light shoe). But if you attach the anamorphic adapter, MF50 would be the proper focus point for something that's 2' 10 W' away from the lens. And if you use the .6x wide-angl~ adapter, . infinity is reached at approximately MF30, whereas WIthout a wideangle adapter, infinity focus is set at MF95. Using close-focus diopter lenses also changes the focus point, and using a close-focus diopter is a way to take advantage of the lens' ability to seemingly focus past infinity: the full range up to MF99 is usable.

96

The focus ring also continues to work in auto focus mode: if the camera is set to autofocus, and you turn the focus ring, it will change the focus position of the lens.

The focus ring doesn't have hard stops at minimum or maximum focus points, but it is precise, and repeatable. Century Optics makes an optional focus gear/ring that can be attached to the camera that provides for hard stops at each end of travel, as well as providing industry-standard pitch gear teeth for interfacing with devices such as follow focus units, or focus motors on professional jib arms, etc.

For focus numbers and their relationship to distance, you can refer to the chart included in the Anamorphic Adapter User's Guide section at the end of this book.

TAPE - BATTERY: This menu setting controls whether to display the remaining Tape Left in the camera and the remaining Battery Life. Battery Life is shown with a little graph of a battery in the lower right comer, Tape Left is shown with a numerical readout right next to it. Some people confuse the "Tape Left" number with Battery Life, thinking that the number being displayed refers to the number of minutes of battery life left.

OTHER DISPLAY: The camera is capable of displaying many pieces of information in the viewfmder. Displaying all of them makes for a quite cluttered display, but not displaying enough of them may leave you without vital information during your shot. The OTHER DISPLAY menu item lets you selectively show this additional information. There are three settings: ALL, PARTIAL, and OFF.

97

The display items that PARTIAL and OFF control are: recording speed (SPILP)

Auto-iris (blank for manual or STD for auto-iris operation) Shutter speed (which disappears if you have the Marker turned on)

Audio sampling frequency (48KHz/32KHz)

Scene file name (in the upper right comer) Progressive display (blank, 30P, 2SP, 24P, or 24PA)

A WB display (P3.2K, PS.6K, ATW or blank for manual) Iris readout (OPEN, F/1.7 through F/22 or CLOSE)

Gain (OdB through 18dB)

ND filter (blank, ND1I8 or ND1I64)

Recommended ND filter (flashing ND1I8 or ND1I64)

Image stabilization status (a hand with "shake" lines around it, or blank)

Microphone limiter display (either ALC or blank) Status of the AUTO button (either AUTO or blank).

CAMERA DATA: This menu item is only applicable in VCR playback mode. When you record footage on the DVX, the camera automatically records some information about the camera's setup during the shot. The camera will record whether or not the Optical Image Stabilizer was used during the shot, what f-stop the camera was at, and what gain setting was used (OdB to 18dB). Knowing this information could come in handy if trying to recreate the look of a certain shot, for intercutting a pick-up shot into existing footage.

EVF COLOR (DVXIOOA and DVXIOOB ONLY):

The DVXlOOA and DVXI00B allow you to set the viewfinder (but not the flip-out LCD panel) to display in black & white or color. Setting this menu item to ON will enable color in the viewfinder; setting it to OFF will turn the viewfinder to black & white. Disabling the color in the viewfinder does not make the viewfinder the equivalent of a shoulder-mount professional video camera's CRT viewfinder; it will still be an LCD. However, with color disabled, many people find focusing to be a bit easier, and the effects of the EVF DTL "peaking" circuitry may be easier to see without the distraction of color in the picture. Turning the color off, and EVF DTL on, may make the viewfinder a more accurate tool for focusing.

98

DISPLAY ASPECT (DVXI00B only): The DVX100B includes the ability to switch the aspect ratio of the LCD and viewfmder between regular (4:3) and widescreen (16:9). There are three menu options:

AUTO: Selecting AUTO will instruct the DVX100B to change the aspect ratio of the LCDIEVF based on what mode you're shooting in. For NORM or LETTERBOX, the camera will select 4:3. When shooting SQUEEZE mode, the camera will select 16:9. Also, AUTO will allow the camera to automatically adjust the aspect ratio when playing back footage - SQUEEZE footage will cause the display to automatically switch to 16:9 mode; LETTERBOX and NORM footage will be displayed in 4:3 mode. For most purposes, AUTO is the most useful setting.

4:3: You can force the display to always be 4:3. This will cause SQUEEZE footage to display improperly (it will be stretched vertically, making everyone look tall and skinny). However, it may allow a little more precision in focusing, because the footage will be displayed in higher resolution (the 16:9 aspect ratio conversion process lowers resolution slightly).

16:9: You can force the display to always be 16:9. This would be most useful when using the LA-7200G Anamorphic Adapter, for example. However, be aware that footage shot in NORM or LETTERBOX mode will be displayed "squashed" if you set this menu item to 16:9.

The rest of the settings in this menu are adequately described in the camera's manual, and will not be duplicated here.

99

OTHER FUNCTIONS SETTINGS

REMOTE: The camera can be set to respond to two different remote control units. Since the DVX's remote control is cross-compatible with Panasonic's other cameras, it's possible you may run into circumstances where more than one camera remote control is in use at the same time. If so, you can configure the remote (by pressing a certain button combination) and the DVX (using this menu item) to

. correspond to either VCR1 or VCR2. You can also set this menu item to OFF, which will cause the DVX to ignore any and all remote control signals. When using more than one DVX camera, this setting could be used to allow all the camcorders to be controlled from one remote control. This item makes sure that only that one button (VCRI or VCR2) will operate this camcorder from the remote control.

DV CONTROL: The DVX offers the ability to issue recording control commands through its 1394 connector. IEEE 1394 is an extremely high speed, high performance Serial Bus that provides for transferring data between digital devices, including audio and video

data, time code, and also provides for remote control of a VCR. If you were to attach a 1394 cable to the DVX, and attach the other end to a deck or camera (or a hard disk recorder unit, such as the FireStore™ or a computer software hard disk recorder such as DV Rack™) then the DVX can send commands to that external recorder whenever you press the DVX's record button.

There are four possible settings:

OFF: When set to OFF, the DVX will not try to control the external unit (camera, deck, or hard disk recorder). However, the data stream is still sent over the 1394 cable, and if you manually trigger the external recorder to start recording, it will be able to record the 1394 DV data stream regardless of whether the DVX is controlling it (and, in fact, regardless of whether the DVX is recording or not!)

EXT: When set to EXT, the DVX will not try to record anything on its internal tape. It will only tell the external unit

100

to record or pause. If you have a hard disk recorder or deck, this can save wear and tear on your camera's recording heads. Also, this is a way to get much longer continuous record time: you can hook your DVX up to a large-format DV cassette deck, and be able to record for 276 minutes on one tape without stopping. And, when recording to an external deck, the data recorded will be at the full quality the DVX can provide, at the frame rate the DVX records at. This means that even if you're using a little cheap $300 DV camera as an external recorder, it will still record the full quality and 24P or 30P image that the DVX would record!

BOTH: When set to BOTH, the DVX will record to its own tape, as well as instructing the external recorder to start recording. This way you'll have an automatic backup. The recording will be redundant, as you'll be recording the same image on two different devices; however, this is the best way to avoid dropouts on your tape. If a dropout (or "glitch") happens on one tape, you'll have a backup that will almost definitely not have a glitch at that same exact spot, so you'll have a clean copy you can use to repair any glitches in post.

CHAIN: This feature is a powerful option the DVX brought to prosumer camcorders. In Chain mode, the DVX will start recording to its internal tape, until that tape is almost full. When the camera gets close to the end of the tape, it will automatically trigger the external recording deck to start recording. That way you can make a seamless recording across multiple tapes. You can use the external recorder to cover the time while you're changing tapes in the DVX. Using the CHAIN method, you can theoretically record for hours and hours and hours without missing a single frame, although you would need to patch the tapes together in post to make a seamless recording.

DV eMD SEL: This menu setting works in connection with a recording device attached to the 1394 DV output jack. In other words, this command only works if you have a camera, tape deck, or hard disk recorder hooked up to your DVX via a 1394 cable. What

101

this menu option does is control what type of signal gets sent down the 1394 cable to the other device whenever you press the RECORD button. Once you're already recording, and you want to stop recording, you can tell the other device to go into either PAUSE or STOP mode when the RECORD button is pressed. Obviously PAUSE mode will allow for quicker restarting of recording when you press RECORD again, but it will also keep the other device's tape heads engaged. For the least wear and tear on your backup recording device, STOP mode will tell the backup camera to stop, rather than pause. The downside to using STOP is that when you tell it to start recording again, it may take a few seconds for the recording heads to get up to speed and recording actually begins, so you may lose the first few seconds of your shot.

REC LAMP: The camera has two "tally" lights, one at the front and one at the rear, that light up red whenever the camera's recording. This menu option lets you choose how those lights behave: you can have them both tum on when recording, or neither, or just the front, or just the rear. If your on-camera subject gets nervous when they see the red light, you may want to tum the front light off. Another reason to disable the front light would be if you were shooting into a somewhat reflective surface and the red light is visible in the shot. Or, if you want to shoot in "stealth" mode, with nobody knowing the camera is running, you can disable both lights. Some on-camera subjects will prefer to see the red light active so they know the camera's rolling and they're being recorded.

BEEP SOUND: The DVX is capable of "beeping" when certain events happen,like consumer cameras do. The default state ofthis menu item is OFF, but if you'd prefer to have audible alerts, you can enable the beep. The beeping can be considered unprofessional, but in certain circumstances it may come in handy; for example, if you find that you don't always succeed in activating the camera when you press the RECORD button, you can enable the beep which will give you audible feedback (one beep when you start recording, two beeps when you pause recording). One thing to be aware of though: if you have an analog device connected to the RCA jacks of the camera (such as a VHS deck or DVD recorder, etc.) and the camera BEEPS, that beep will also be output through the analog jacks.

102

TAPE PROTECT (DVXIOO and DVXIOOA) or POWER SAVE (on DVXIOOB): Most consumer and prosumer video cameras have an irritating habit of turning themselves off after about five minutes of inactivity. This can be very vexing when you're setting up a shot, for example: you're taking the time to frame your shot, set the lighting, and when you go to review how it looks on the monitor you find that the monitor is a black screen because the camera's turned itself offl This menu item lets you avoid that - instead of powering itself off, you can set the camera to go into standby mode (STBY). The default setting is POWEROFF, which powers the camera down, saving wear on the heads and saving battery life. The other option is STBY, which means the camera stays powered on, but the tape drive mechanism is powered down. This relieves stress and tension off the tape heads and protects them from excessive wear. If the camera

goes into STBY mode, it will take a bit longer to start recording the next time you press the red RECORD button: the tape heads need to power up and get back up to speed, and that can take several seconds, so always watch the indicators in the viewfmder to verify that the camera is actually recording so you don't miss the first few seconds of your shot. On the DVX100B this menu item is named POWER SAVE, and the menu items are ON or OFF. A setting of ON means the camera will power itself off after about 5 minutes; a setting of OFF means the camera will not turn itself off.

H.P. MODE (DVXIOOB only): H.P. stands for Head Phones. This menu item lets you control the source of audio that gets sent through the headphone jack - either LIVE, or from TAPE. LIVE means that the headphone jack will output the audio directly from the audio input circuitry with no "buffering." TAPE means the audio plays back using the same delay that the audio recorded to tape receives (in progressive-scan mode). For more discussion on this audio-buffering topic, see the article on page 27. In practical terms, TAPE can mean that you may hear an echo in the headphones when shooting progressive-scan 24p; LIVE eliminates the echo.

USER FILE: The "user file" refers to all the camera settings that are not in the scene files. The DVX lets you save all your settings, or restore them to the factory defaults. Each scene file can be individually saved, and the rest of the camera settings are all saved in

103

the ''user file". If you were going to be using your camera in a mixed-camera environment, or loaning it to someone else, you might want to save all your settings before loaning it out. That way the new user could restore the camera defaults by going to the USER FILE menu and choosing "INITIAL" - that will restore the camera's settings to the factory original. Then when you get the camera back, you can go to the USER FILE menu and choose "LOAD" - this will restore all the settings that you had saved. When the LOAD or INITIAL operation has been performed, turn the power OFF and then ON in order to enable the settings.

FILE TRANS and FILE RECEIVE (DVXIOOB only): The DVX100B includes the ability to transfer scene files or the user file to another DVX100B, using the 1394 cable. This transfer system only works between two DVXI00B cameras; you can't transfer files from a DVXI00A to a DVXI00B, for example. File transfers can make matching multiple cameras an extremely fast and efficient process. FILE TRANS is available only in camera mode, and FILE RECEIVE is available only in VCR mode.

To transfer files, first connect two DVXIOOB cameras together using a 4-pin to 4-pin 1394 cable. Set the camera you want to send the file FROM (the "source" camera) into "camera" mode. You'll also want to turn DV CONTROL to OFf (in the "OTHER FUNCTIONS" menu). Set the camera you want to send the file TO (the "receiving" camera) to VCR mode. Go into the RECORDING SETUP menu on both cameras. On the receiving camera, select the FILE RECEIVE menu item to YES (its LCD will start flashing "RECEIVE MODE"). On the source camera, select one of the FILE TRANS options.

The FILE TRANS menu has three options:

SCENE: This setting will let you transfer the current scene file to the receiving camera. Using SCENE will transfer only one scene file across. The scene file that gets transferred will be whichever scene file the source camera is currently set to (i.e., whatever the Scene File Dial is set on, FI through F6). The scene file will be transferred to the same position on the receiving camera, regardless of what the receiving camera's Scene File Dial is set to (i.e., Scene File I on the source

104

camera will always transfer to Scene File 1 on the receiving camera; you cannot send a scene file to a different scene file slot in the receiving camera).

SCENE ALL: This setting will transfer all six scene files from the source camera to the receiving camera.

USER: This setting will transfer the source camera's User File settings to the receiving camera. All camera setup information (such as timecode mode, SW mode settings, etc) will be transferred to the receiving camera.

After transferring files successfully, both cameras will display a FILE TRANS OK message, indicating that the transfer is complete (if for some reason the transfer doesn't work, the cameras will display FILE TRANS NG, where NG = "no good"). The receiving camera will request that you turn the power off and then back on to complete the process.

105

PHYSICAL CAMERA SWITCHES AND JACKS

This section will describe some of the features of the camera and observations about how those features work and how they can be best employed. The switches and other physical buttons and jacks on the camera are referenced here by name as well as by number; in the camera's manual, each switch or jack is identified by a number on a diagram of the camera. Unfortunately, those numbers are different for the DVX100 & DVX100A manual, as compared to the DVX100B manual (for example, the HEADPHONES JACK is referenced as "6" in the DVXlOO and DVX100A manuals; it's referenced as "10" in the DVXlOOB manual.) Those numbers in bold correspond to the identifying numbers given to the items in the original DVX100 and DVX100A camera user's manual; the numbers in italics correspond to the identifying numbers given in the DVX100B camera user's manual. Refer back to the camera's manual for information on where each numbered switch, button, or jack is located on the camera.

5 & 11: CAM REMOTE JACK & ZOOM SIS. On the back of the camera, hidden under a flap next to the battery, there is a small

2.5mm mini-plug socket named "CAM REMOTE" on the DVX100 and DVX100A; on the DVX100B it's labeled "ZOOM/SS". This socket allows for a remote zoom controller to be used. It is NOT a LANC-type of jack, and there is no provision for remote focus

control through this socket. The only functions that can be controlled through this socket are zoom and record start/stop. The DVX has three zooming speeds (low, middle, and high) and this jack can trigger the camera to start zooming at any of those speeds. There is no provision for "feathering" the zoom, or for getting smoother zoom transitions than would be possible with the zoom levers on the camera itself. Other cameras use LANC protocol, which includes the ability to control the focus motor in the lens, but the DVX doesn't offer that protocol, and this remote jack has no provision for controlling focus. The only way to control focus remotely with a DVXlOO or DVX100A would be to use a physical motorlgear combination that actually physically turned the DVX focus wheel.

106

11: CAM REMOTE FOCUSIIRIS (DVXI00B only). The DVX100B adds another remote jack right above the ZOOM SIS jack. This is a 3.5mm mini-plug socket labeled "FOCUS/IRIS". This jack allows the use of a remote controller to control the focus and iris settings. Using this jack allows owners of optional remote control accessories to have remote control of focus and exposure (iris).

6 & 10: PHONES JACK. On the back of the camera, hidden under a flap next to the battery, there is a sma1l3.5mm mini-stereo socket labeled "PHONES". This obviously allows you to plug in headphones and monitor what the DVX is recording. Many people complain about the volume level coming out of the DVX100 headphone jack, saying that the.volume level is too low, and that you need an external amplifier (such as a "Boostaroo™") to get adequate volume levels out of the headphone jack. Actually, the headphone jack provides plenty of volume, but what's happening is that people are using consumer-style headphones, whereas the DVX's jack is designed to drive professional headphones. There's an impedance mismatch between the two. The DVX100 headphone jack is rated at 32 ohms at 77mV. If you use professional headphones with the DVX (such as Sony MDR-7506's) you will find that there's plenty of volume available. Using cheaper "consumer" headphones, you may fmd that there's such an impedance mismatch that not much volume gets through. Match the headphone impedance to the camera and you'll be much happier with headphone performance.

One of the changes made on the DVX100A was to amplify the headphone jack output. A side effect of this is that there's a noticeable hiss in the headphone output on a DVX100A and DVX100B. The hiss is not recorded on tape, it exists only in the headphone circuitry, and is a byproduct of having amplified the headphone jack output. The DVX100A and DVX100B headphone jack is rated at 100 ohms.

7 & 4: REMOTE CONTROL SENSOR: The DVX has two sensors for the infrared remote control. While the remote is not typically used very frequently with the DVX, there are two circumstances where it is invaluable:

107

1) When doing stop-motion animation or other lock-down shots where it's critical that the camera not be moved, you can trigger recording without touching the camera by using the remote control.

2) When using multiple cameras in a multi-camera shoot and you want to ensure that timecode matches perfectly between the cameras. For more info, see the article on page 55.

9 & 16: VIEWFINDER: The DVX's viewfinder is quite large and versatile. You can configure the eyecup for left-eye or right-eye viewing, focus it sharply using the diopter adjustment dial, and you can configure the camera to enable the viewfinder to stay active even if the flip-out LCD panel is active. This can come in handy if the camera operator needs to be viewing the scene and someone else (director, focus-puller, etc) needs to be viewing the LCD panel: you can tum the LCD panel so that it snaps into the side of the camera body, screen facing outwards, so others can view it while the camera operator has exclusive use of the viewfinder. Configuring the camera to use both the viewfinder and the LCD at the same time will result in the camera drawing more power, thus resulting in shorter battery life. On the DVX100A and DVX100B, the viewfinder can also be configured to be either black and white or color. Selecting black & white may make it easier to discern the effects of the EVF DTL function, which may make for easier critical focus when using the viewfinder. On the DVX100B the viewfinder is significantly higher resolution, and it also shows the full ''underscan'' frame (not just the "overscan" area, as on the DVX100 and DVX100A and most every other under-$5,000 camera).

12 & 18: DC INPUT CONNECTOR: The DVX includes a 7.9V DC input jack in the battery compartment. If you have a compatible DC power supply you can power the DVX from this jack rather than from batteries or the AC power supply. Another way to get alternative power to the DVX is to use something like the Bescor CLC-D120, which allows you to plug the DVX into a l2V battery belt through the use of a cigarette-lighter plug.

108

15 & 38: SCENE FILE DIAL: The DVX provides a unique way to change the look of your video instantly, or pre-load several different looks. The Scene File Dial (also called the CineSwitch™) is a sixposition dial on the back of the camera that lets you store six customized scene files. The scene files are named Fl through F6, and come pre-loaded from the factory with certain presets that Panasonic thinks users would find useful. You can customize many settings in each scene file, and switch quickly between them even while recording. All the settings in the scene file will be updated on-the-fly except for frame rate: you cannot switch between 24P, 30P, and 60i while the camera is actually recording. If you want to change the frame rate, you have to stop recording first. This means you can't switch from 24P to 30P in the middle of a shot, for example. However, all other settings will change, so you can switch from CineGamma to Low Gamma, high detail to no detail, etc. all on the fly. Scene files can be customized and saved; on the DVXIOOB they can even be shared between cameras by using the 1394 port.

16 & 39: END SEARCHlEVF DTL BUTTON (EVF DTL only available on DVXIOOA and DVXI00B): when the camera is in VCR mode, you can use the END SEARCH button to find the end of the tape. The camera will automatically fast-forward or rewind until it fmds what it thinks is the last recorded segment on the tape, and then position the tape to start recording after the last shot. Using this feature can help avoid timecode breaks by making sure the camera picks up from the end of the last shot (also see REC CHK for avoiding timecode breaks). On the DVXIOOB you can perform an end search even while in camera mode; just hold down the button for a few seconds and it will begin the end search process.

On the DVXIOOA and DVXl OOB this switch also controls the EVF DTL ("peaking") function. When in CAMERA mode, instead of working as END SEARCH, the switch engages or disengages the EVF DTL feature. This feature can help you focus more accurately by providing a professional "peaking" feature in the viewfinder and on the LCD. When you have peaking turned on, any items that are in sharp focus will be outlined with a subtle white outline. The effect isn't dramatic, but it is noticeable, especially when you know what to look for. You may have an easier time seeing the effect if you switch

109

the viewfinder to black and white mode (an option only available on the DVXIOOA and DVXIOOB). Note that this "outlining" effect is not recorded on tape, it's only displayed on the viewfinderlLCD. On the DVXIOOB, press and release the button quickly; holding down the button too long will start an end search instead.

17 & 32: GAIN SWITCH: The DVX has a 3-position toggle switch for controlling picture gain. Gain is an electronic amplification of the signal, which means that by using gain you can make the picture brighter than it otherwise would look. The downside to using gain is that it introduces noise into the picture. The more gain you use, the brighter the picture becomes, and the noisier the image gets. The switch provides for three positions: L( ow), M(id) and H(igh). LOW means that no gain is applied. The factory defaults for MID and HIGH are 6dB (twice as bright) and 12dB (four times as bright). You can change those settings in the SW MODE menu. The DVX also provides for a very strong 18dB of gain, which would give you a picture eight times as bright as with no gain (but the picture will be very noisy). You cannot assign 18dB of gain to the GAIN SWITCH; the only way to get 18dB of gain is to assign it to one of the USER buttons. It should be noted that Gain doesn't work on the original DVXIOO when in progressive-scan modes (24P, 24P A, 25P or 30P). Gain only works on the DVXIOO when it is in interlaced mode. Gain will work in all modes on the DVXIOOA and DVXIOOB, but 18dB only works in interlaced mode. Also, gain doesn't work in combination with the DVXIOOA or DVXIOOB Slow Shutter modes: you can have slow shutter speeds OR gain, but not both at the same time. Another thing to understand about Gain is that it can only amplify the signal that the camera is currently seeing; it cannot add detail that wasn't there. If you're shooting under low light conditions and need to employ gain to get the picture bright enough, you should understand that your video is in all likelihood underexposed, and using gain will artificially brighten up the picture, but it will not restore detail that wasn't properly captured due to the underexposure. Gain is usually used as a "last resort" - when shooting under dim conditions you should take other measures to increase the brightness of the scene first, including removing all neutral density filters, opening up the iris to its maximum opening, and adding light whenever possible.

110

18 & 33: WHITE BAL SWITCH: The DVX offers a 3-position switch for white balance. The positions are PRST, A, and B. Using this switch, in connection with the menus and the A WB button on the front of the camera, you can choose from a wide variety of white balance possibilities: either a 3200k preset, a 5600k preset, up to two channels of manual white balance, or an automatically-tracking white balance mode (ATW). For a discussion on color temperatures and white balance, see the discussion under menu item ATW (page 76).

To white-balance the camera, first decide if you want to use one of the existing presets or if you want to use a manual white balance.

The presets are selected by setting the WHITE BAL switch to PRST, and then toggled by pressing the A WB button on the front of the camera (19 & 28). The presets are P3.2K and P5.6K. Those correspond to indoor lighting (3200 Kelvin) and outdoors (5600 Kelvin). While the presets are perhaps a good starting point, there are many circumstances where a preset will not deliver the most accurate color rendition. For example, many incandescent and halogen lamps burn at color temperatures different from 3200 Kelvin; some may burn as low as 2700 K. If you're using 2700 K lamps to light your . scene, and you have the white balance set to P3.2K, your lamps will not look white, they'll look orange-ish. Also, daylight varies tremendously in color temperature, from around 3000 K during sunrise/sunset to over 10,000 K on an overcast, cloudy day. So the presets are a good starting point, and good for on-the-run shooting, but if you have the time to take a manual white balance you can get more accurate color rendition. Also, using the presets will disable the COLOR TEMP feature in the scene file menus.

The DVX provides two channels of white balance, A and B. Both function identically. To set a white balance, place a white card (or other white object) into the light where you intend to be shooting. Ideally you'd have your subject hold a white card up in front of their face, etc. Zoom in until that white card fills the screen. Set the WHITE BAL switch to either A or B, and press and hold the A TW switch until the camera displays "A WB Ach ACTIVE" (or "A TW Bch Active"), which lets you know that either A-channel or Bchannel is being calibrated to an accurate white balance. Any time

111

your lighting conditions change, you'll need to re-white balance if you want your colors to continue to be rendered accurately.

Another white balance option is to set one of the WHITE BAL switch options to ATW in the SW MODE menu. You can set anyone ofthe WHITE BAL switch settings to ATW mode (either PRST, A or B). For example, you could configure your camera so that PRST = presets, A is reserved for manual white balance, and B is set to ATW. Then when you move the WHITE BAL switch to "B", the camera will automatically start tracking white balance by itself, updating as lighting conditions change. For professional shooting situations you may not want that, but for run on' gun type situations it may come in handy. However, the ATW response rate is fairly slow (it can take a few seconds to respond to changing lighting conditions).

The manual specifically says to not block the White Balance Sensor "or the ATW function will not operate properly." During manual white balancing the camera makes its determination by looking through the lens (which is why you want to zoom in on a white object). But during auto-tracking white balance (ATW), the camera doesn't get its color temperature information through the lens; rather it relies on a little window near the 1394 port. If you block that window then the camera won't be able to track the white balance, but it will have no effect on your ability to manually white balance. Furthermore, understanding the different ways the camera gathers light to judge a white balance can explain why auto-white balance might fail under certain circumstances. If the scene the camera's lens is seeing is lit with a certain color oflight, but the light striking the White Balance Sensor is a different color, the ATW system will balance towards the light striking the White Balance Sensor, which will render your scene in the wrong colors. An example may be if you were standing outside in full sunlight, shooting into a house through a window, and the scene inside the house is lit with tungsten lighting. In this case the daylight will be hitting the White Balance Sensor, causing the camera to believe that daylight is the prevailing lighting color, whereas your scene indoors is lit with tungsten light. In circumstances such as that you would be much better off to perform a manual white balance.

112

21 & 34: IRIS DIAL: The iris dial allows you to set the f-stop manually. You can even use it when the camera's in AUTO-IRIS mode; the iris dial will let you override the auto setting, although the camera will eventually compensate. A common technique is to let the auto-iris set the overall exposure level, then switch to MANUAL IRIS and fine-tune the exposure according to taste (or according to the zebras or referencing a production monitor).

In order to understand how the iris dial works, you first need to understand what f-stops are. F-stops are basically a way to describe the amount oflight the iris STOPS from getting into the lens. Put in simple terms, fll would be admitting as much light as the lens is possibly capable of (think of it as "fll" = "f divided by one" ... "f' = the maximum amount oflight, so "f' divided by 1 would still be "f', still the maximum). F-stops are numbered according to the following sequence: fll, fll.4, fl2, fl2.8, fl4, fl5.6, fl8, fill, and fl16. Each additional f-stop cuts in half the amount of light admitted by the previous f-stop. So fll.4 admits half as much light as fll does. F/2 admits half as much light as fll.4, and so on and so on. F-stop numbers are based off of two base numbers, fll.0 and fll.4. Each new flstop number is seen by doubling the previous number. For example:

1.0 1.4 2.0 2.8 4.0 5.6

8.0 11 (rounded down from 11.2) 16 22

So if you remember 1.0 and 1.4, you can calculate the rest of the sequence easily.

You can think of the flstop notation as a diameter formula for the lens iris. Whatever "f" stands for, when expressed in the term of fl2.0, would mean an iris size of "f" divided by 2, which would let in 114 as much light as an "f" divided by 1. Think in terms of a square: if you divide the length of a side of a square by 2, the overall square is 114 the size as the original: a 2" x 2" square has an area of 4 sq inches, but cut that size in half and you get a 1" x I" square, with an area of 1

113

square inch, which is 1/4 the area of the 2x2 square.

So to get half as much, you don't divide by two, you need to divide by 1.4 (or the square root of2). If you take the 2" side of the square and divide that by 1.4, you'd get a square of 1.42" x 1.42", which has an area of2 square inches (1.42 x 1.42 = 2.0). And 2 sq. in. is 112 as much area as the original 2x2 square's 4 square inches. So to get half as much coming in, you need to divide by 1.4.

Therefore, the numbers you divide "f" by are: 1, 1.4,2,2.8,4,5.6,8, 11, 16,22

Each successive number lets in half as much as the previous number.

The iris wheel on the DVX shows those whole f-stop numbers, but it also shows other numbers in-between. The other numbers the DVX iris wheel shows are half-stop increments, so between 2.0 and 2.8 (which are both "whole" stops) you'll fmd 2.4 listed, which is a halfstop, which is equivalent to halfway between 2.0 and 2.8, and lets in 75% as much light as fl2.0 does (since fl2.8 lets in 50% as much light as fl2.0 does).

The full sequence of f-stop numbers that the DVX can display are:

OPEN (fl1.6), 1.7,2.0,2.4,2.8,3.4,4.0,4.8,5.6,6.8,8.0,9.6, 11, 14, 16, and CLOSE

The DVX iris wheel is actually calibrated much finer than half-stops though. The DVX iris is calibrated in 116 stops. By very carefully rotating the iris wheel, and watching a monitor, you can discern that there will be three changes of brightness before the f-stop display changes (i.e., when at fl2.0, you can stop the lens down two visible ticks and the readout will still show fl2.0, but when you stop down one more tick then the display will change to fl2.4). Knowing this, you can make very fine and very precise exposure adjustments.

The DVX's iris can be more open at wide-angle than it can at full telephoto. At full wide angle (ZOO) the maximum iris opening is fl1.6, but at full telephoto (Z99) the maximum iris opening is fl2.8. If you're shooting in extreme low light conditions and need the

114

brightest picture you can get, you may want to avoid zooming in very much, as the most telephoto position of the lens is a full 1.5 f-stops slower than the full wide angle position.

22 & 29: FOCUS SWITCH: On the side of the lens is a switch named "FOCUS", with three possible settings: "A", "M", and "00". This switch lets you control the auto focus capability of the camera. When set to "A", the camera operates in full autofocus mode, automatically hunting for the best and sharpest focus. When set to "M", the camera is set to strictly manual focus. When pushed to "00" the camera is instantly set to "infinity" focus, the setting where even the furthest possible objects (such as the moon) should be in sharp focus. Infinity is not really a switch setting, it's just a momentary push - when you push the switch down, the lens gets set to infinity and then the switch automatically returns to manual focus mode.

Full-fledged Autofocus works only in 50i/60i interlaced mode. When in progressive-scan mode, autofocus doesn't work at all on the original DVXlOO, and on the DVXIOOA and DVXlOOB it works in a limited capacity, called "focus assist". The performance of the autofocus system is best in interlaced 50i/60i mode.

On the original DVXIOO you could use autofocus to assist you in focusing your shots even in progressive mode, by temporarily swapping the scene file dial away from progressive mode. For example, scene file setting F5 is factory-default set to 24P progressive mode, and scene file setting F4 is factory-default set to interlace mode. If you wanted to use auto focus to assist you in getting good focus, you could set the FOCUS switch in "A" (auto) mode, and then switch the scene file dial toF4. The camera would go into interlaced mode, and the autofocus motor would start working. Once the camera "locks in" its focus (which is known by the fact that the AF readout in the viewfinder locks to a steady number) you would then switch back to F5 to shoot your scene. The focus setting will stay the same regardless of what scene file the camera is set to, so sharp focus set by auto focus in the interlaced mode will still be sharp focus when you swap back to progressive mode.

115

On the DVXlOOA and DVXIOOB, "Focus Assist" was added. Focus Assist works like regular autofocus, except that it is much slower to respond. The reasoning is simple: in interlaced mode, the autofocus system gets fed sixty (or fifty in PAL) updates per second (due to the 60-field-per second speed of the CCD). But in progressive scan mode, the updates come far more slowly: in 24P mode, the autofocus system only gets 24 updates per second, which is only 40% as often as in interlaced mode. With less information to work from, the autofocus system responds to changes much more slowly (mainly because it doesn't see those changes nearly as frequently). The result is that when in 24P or 25P/30P mode on a DVXlOOA or DVXlOOB, the autofocus response isn't satisfactorily quick enough to be called true "autofocus." Instead it can be used as a "focus assist" to help you determine proper focus.

Autofocus works best under brightly lit conditions. Under low light conditions, autofocus has to work much harder, and will respond much more slowly, and will be more prone to "hunt" for proper

focus. Another factor to consider in auto focus performance is that the auto focus system relies on measuring contrast to determine proper focus points. Ifthe scene you're shooting is very low in contrast, the autofocus system will have a much harder time determining the proper focus point. Autofocus works quickest when it can easily discern a transition between dark and bright elements, especially vertical elements (i.e., on a black and white picket fence, the

auto focus system would perform superbly. Trying to find focus on a sold white sheet of paper would be extremely challenging for it.)

When the FOCUS switch is set on "M", the DVX enters manual focus mode. The focus readout in the LCD display changes from AF to MF to signify the change from Auto Focus to Manual Focus. In Manual Focus mode, the DVX offers the most precise focus system of any under-$4,000 camera. The focus ring moves with precision and an accuracy that is just as precise as a true physical-connection manual focus ring. Focus marks are definite and repeatable. Rack focus moves are easily accomplished. The only substantial difference between how the DVX's focus ring works, and how a true manual linkage works, is that the DVX's focus ring doesn't have physical hard stops at infinity and at minimum focus distance. However, you

116

can buy an accessory ring from Century Optics that adds hard stops, follow-focus gear teeth, and a markable surface (for marking focus points for rack focus moves) that gives you a fully manual focus solution.

When the FOCUS switch is pushed towards the Infinity symbol ("cd'), the lens is set to infinity focus (MF95). However, as discussed in the section on ZOOM-FOCUS, infinity focus may need to be adjusted depending on what accessory lenses you have installed on the camera. For example, when using the anamorphic adapter, infinity focus is achieved at MF93, not MF95. And when using the Century Optics .6 wide-angle adapter, infinity focus is achieved at approximately MF29. So the Push- To-Infinity button is only practical if you have no additional lenses installed on the camera.

Also of note, the auto focus system works by comparing vertical lines, and ignoring horizontal lines. This is why the auto focus system delivers incorrect focus numbers when the anamorphic adapter is attached: the anamorphic focuses at different planes on the horizontal and verticals, and the auto focus system only looks at the verticals.

26 & 31: INDEX BUTTONIUSERJ BUTTON: On the DVXIOO, next to the USER buttons is a button labeled INDEX. On the DVXIOOA and DVXIOOB this button was replaced by USER3; however, you can assign the same INDEX function to USER3 to get the same ability as on the DVXIOO. The INDEX function writes an invisible, searchable tag onto the DV tape. When shooting, you can mark a spot on the tape for easy locating back in the edit bay. An example might be if you're shooting multiple takes of a scene, and you fmally get the best take: if you press INDEX at that point, you can easily perform an "index search" on a deck or by using the camera's remote control, to fast-forward exactly to that point on the tape. Note that the index won't be physically written to the tape until the camera is stopped; if you're using INDEX to mark your scenes, the mark will be occurring after the scene, so you may have to rewind to see the actual scene.

27 & 22: AUTO BUTTON: The DVX cameras have a dizzying array of manual controls and capabilities. But what if you're in a situation

117

where you just have to grab the camera and shoot (say, a breaking news story?) For circumstances like that, where you simply don't have time to set all the manual settings in time to get the shot, you can press the AUTO button and the camera will take over many of the functions automatically. You can configure the camera to instantly switch into auto-focus, auto-exposure, auto-iris, auto-gain, and autowhite balance ... or you can tell it to go into auto-mode for any combination of those settings (see the discussion on the AUTO SW menu for more information). Another good use ofthe AUTO button might be if you need to hand the camera to someone who's not skilled on the DVX, to get a shot you need. The results will likely not be as good as if you'd manually set up the camera, but the AUTO button may make the difference between getting a shot, and not getting it.

28 & 31: USERI and USER2 (and USERJ on the DVX100A and DVX100B): These buttons allow you to instantly switch in certain features, such as color bars, 18dB gain, spotlight or backlight compensation, or fade to white or black, etc. Further details are discussed under the menu options on page 78.

29 & 24: ND FILTER SWITCH: The camera comes with two Neutral Density (ND) filters. Neutral Density filters are used to control exposure, and ND filters act like "sunglasses" for your camera: they help cut down the amount of light entering the camera, so in bright conditions you can engage the ND filters to lower the light level and get proper exposure. They're called "neutral" density filters because they add no color shift to the image: they're a neutral shade of gray, so the only image effect is to lower brightness.

The switch has three settings:

OFF: no ND filters (best used in lower light conditions); 1/8: one stage ofND filter, reducing the amount oflight coming into the camera by approximately 3 and 113 f-stops 1/64: two stages ofND filter, reducing the amount of light coming into the camera by approximately 7.5 f-stops.

The ND filters can be thought of as two 3-stop filters. ND filters are named according to how many thirds of an f-stop they reduce the incoming light, and typical ND strengths are ND .3 (one stop), ND .6

118

(two stops) and ND .9 (three stops). An ND .3 reduces light by 3 thirds of an f-stop, or one full f-stop. Put another way, an ND .3 reduces the amount of light coming into the camera by half. The exposure compensation of an ND .3 is the equivalent of closing down the lens by one f-stop; for example, a camera shooting at fl4 with an ND .3 filter will deliver the same exposure as a camera shooting at fl5.6 with no ND filter.

The DVX includes two (approximately) 3-stop ND filters (or, ND.9 filters). Three stops of light reduction results in the amount of light being cut to 118 its intensity (one stop = Y2 the light, another stop = ~ the light, and a third stop = 118 the light). So the switch is named ''NO 118" for position one. The second position results in three more stops ofND being applied, so the light is reduced to approximately 1164 its intensity: From ND 118, one stop would be ND 1116, two stops would be ND 1132, and three stops delivers ND 1164.

The DVX is an extraordinarily light-sensitive video camera, rating at anywhere between 400 and 1280 ASA depending on what mode you shoot in. Because of this high speed you need to use the ND filters to control the amount of light that enters the camera. For indoors shooting you'll usually want the ND filter off, but outdoors will almost always dictate using at least ND 118 and usually ND 1164. Follow the recommendations of the auto-iris and your zebra display to determine which ND setting to use.

For those who need to know the exact f-stop difference, it turns out that the ND filters are actually a little stronger than is indicated on the ND filter switch. ND 1/8 is approximately three and 113 stops (i.e., an ND 1.0) filter, and ND 1164 is another four and 116 stops, for a total of seven and a half stops of total ND power.

38/39 & 57/59: AUDIONIDEO INIOUT CONNECTORS: The camera features an s-video port and "RCA" yellow-red-white jacks for connecting to video devices, such as external monitors or VCRs. These ports are bi-directional for input/output, so you can display the DVX's image on an external monitor, or you could feed the DVX a video signal from an external deck or camera. You can also configure the camera to perform on-the- fly analog-to-digital

119

conversion, by feeding a video signal into the video connectors, and the DVX will output a digital DV signal from its DV port. The svideo connector takes precedence over the composite (yellow) video port, and in order to record analog audio there must be a video signal present as well (you can't record audio without video).

42 & 66: XLR AUDIO CONNECTORS: The DVX features two XLR audio connectors for attaching microphones, wireless mic receivers, mixers, or other professional audio devices to the camera.

, These XLR connectors are input-only, no audio can be output from them. The DVX also includes two LINEIMIC switches that control the signal level sensitivity of the XLR connectors, for mating the connector to the type of component attached to it. For example, when you have a mixer attached to the DVX, the mixer may be outputting a LINE level signal (at 0 dBu), so you'd want to flip the LINEIMIC switch to LINE for that channel. If instead you hooked up a dynamic microphone, that mic would be outputting a MIC level signal, so you'd need to flip the LINEIMIC switch to MIC to match levels to

the microphone. The MIC level is calibrated to -50 dBu. If you attach a device to the XLR input and you can't get satisfactory audio levels from it (like the audio is way too low, or way too loud) then try changing the setting of the LINEIMIC switch for that audio channel to get a better level match.

Also, the XLR connectors only function for input when in camera mode. You cannot record an audio signal from the XLR connectors when in VTR mode.

44 & 68: ZOOM SWITCH: On the front of the camera, under the lens, is a switch labeled ZOOM: SERVOIMANU. This toggles between servo zoom (i.e., power zoom) and manual zoom operation. When in Servo zoom the zoom rockers on the camera (or the remote control) will drive the zoom motor, but when in MANU mode, the zoom rockers (47/48 & 6/8) will have no effect. When in MANU mode, zoom is controlled by turning the manual zoom ring. The DVX is the first low-cost camera to offer a true manual zoom, just like on a movie camera, where the zoom ring is physically connected to the lens elements. On the DVXlOO it was a little difficult to get smooth manual zooms, but one of the nice improvements on the

120

DVXI00A and DVXI00B is that more friction/tension was added to the zoom ring, making for much smoother manual zooms. Smoother manual zooms are also possible by adding a longer zoom lever, which will effectively increase the diameter of the zoom ring and thus give finer control. When the lever is in SERVO mode you shouldn't try to turn the manual zoom ring, ?r you risk breaking the camera.

46 & 5: REC CHECK BUTTON: When the camera is in CAMERA mode the REC CHECK button will rewind and review the last few

,

seconds on the tape. This is handy for knowing that you've cued up the camera to the end of the tape, past your last "known good" take. It also helps prevent "timecode breaks" by making sure the camera picks up the timecode off the last shot of the tape. You can also manually review the tape when in camera mode by using the fast forward!rewind joystick (37 & 26): pushing to "rewind" will cause the tape to play backwards at real-time speed, pushing the joystick to "fast forward" will cause the tape to play forward at real-time speed. For more control, switch to VCR mode, but for quick review/rewind! play you can use the joystick to do tape review in camera mode.

47/48/49 & 6/7/8: ZOOM BUTTON: The power zoom is controlled by one of the zoom rockers, either the handle zoom (48 & 7) or the main zoom rocker (47 & 6). The DVX is capable of three zoom speeds, slow ~ mid ~ fast. The main zoom rocker is pressuresensitive, the harder you press it the faster it zooms. The handle zoom rocker is not pressure-sensitive, you tell it what speed you want to zoom at (by using the handle zoom switch, 49 & 8) and it will zoom at that fixed speed and only at that fixed speed. Zoom can also be controlled by the remote control (fixed at the "medium" speed) or by the CAM REMOTE/ZOOM SIS jack, at any ofthe three fixed-rate zoom speeds.

54 & 40: LCD MONITOR. The DVX's LCD flip-out monitor is a large, high-resolution screen useful for monitoring your shots and gauging overall exposure and color levels. It is not as high-resolution as a pure dedicated CRT production monitor, nor is it going to be as accurate at gauging exposure or contrast or focus as a dedicated CRT production monitor, but for an on-camera LCD it's one of the best. The LCD monitor can be rotated 180 degrees so that it can be seen

121

from the front of the camera, and it can also be closed against the body of the camera in either screen-in (normal closed state) or screenout position. With it screen-out against the body of the camera, an assistant can be viewing the footage from the side while the camera operator uses the viewfmder, and an assistant camera operator could track focus (using the MF manual focus display).

On the DVXI00 and DVXlOOA, the LCD monitor (and the viewfinder) do not display the full image that the camera is recording. They both display "overscan", which is similar to how most CRT television sets display the picture. If you view your image on a computer editing station, or on a production monitor that has ''underscan,'' you'll find that the camera is recording a little bit more than the LCD or viewfinder are displaying: there's a zone, maybe about 5% around the picture, that you can't see in the viewfinder or the LCD, but which is being recorded on tape. All prosumer and consumer cameras operate this same way. It's very handy to have a production monitor that lets you see the full ''underscan'' frame, but if you don't have access to one of those, the DVXI00 and DVXlOOA give you a guideline to see what would be in the full frame (if only you could see it): there's a groove that surrounds the DVX's LCD screen, a thin rounded-comer groove that surrounds the actual LCD itself That groove corresponds almost exactly to the size of the full recorded frame. If you can imagine that the LCD was actually the

full size of the grooved rectangle, you can picture what will be showing up in the underscan area of the frame. On the DVXI00B, the LCD panel is slightly higher resolution and it does display the full frame (''underscan''). The DVXI00B's LCD panel also rotates even further away from the body of the camera (up to 120 degrees) for more flexibility in positioning.

The LCD SET menu gives you some control over the LCD display, including brightness, contrast and color. While it is tempting to think that you could calibrate the LCD to the color bars to match a professional monitor, it's not really practical because a slight change in viewing angle will drastically change how the LCD's display looks to you. It's much better and safer to rely on a true CRT production monitor to gauge color, exposure and contrast. The LCD is good for

122

reference, but no on-camera LCD is up to the task of showing what your recorded video will really, truly look like on a television.

57/58 & 45146: SHUTTER BUTTON and SPEED SEL: The DVX gives you extensive control over the exposure, through gamma curves, the iris dial, the ND filters, and also through complete control of the shutter speed. The SAUTTER button lets you choose between default ("SHUTTER OFF") and user-controlled shutter speeds. When in "SHUTTER OFF" mode, the DVX selects a default shutter speed depending on what mode the camera is in: 1160 for 60i and 30P, and 1150 for 24P (in PAL, SHUTTER OFF always defaults to 1150 for progressive or interlaced). SHUTTER OFF is really a misnomer, as it implies that there is no shutter, when in fact there is a default shutter speed being selected. Perhaps it's best to think of it as "custom shutter speed off." If the camera is not set to "shutter off' then it's able to have the shutter speed set by the user. There are stock shutter speeds (1160, 11100, 11120, 11250, 11500, 111000, 112000) and SYNCRO-SCAN (discussed in the menu options, page 70). It should be noted that these shutter settings are relative to the progressive-scan mode that's selected: for example, if you select 111000 as a shutter speed when in 24P mode, but then you change to 60i mode, the shutter will no longer be 111000, it'll be 1I2000! Keep an eye on the shutter speed whenever changing modes, to make sure you're getting the shutter speed you want.

Shutter speed affects motion blur, pure and simple. Aperture affects Depth OfField, and shutter speed affects motion blur. Both can be used to control exposure, but not without side effects (i.e., changing motion blur or changing depth of field). In almost all normal circumstances, you'd want to use the default shutter speeds (1160 for 60i, 1150 for progressive or 50i) for normal-looking video. Small variations in shutter speed won't affect the look of your video much (i.e., 1160 will pretty much look the same as 1150 or 1148). Film cameras use anywhere from 1143 to 1160 as a standard, and it all pretty much looks like film motion, so small variations won't matter much.

Shutter speed affects exposure. If you cut the shutter duration in half (i.e., use 11120 instead of 1160) you will need twice as much light. If

123

you double the shutter duration (i.e., 1124 instead of 1148) you'll need half as much light for the same exposure.

There are certain circumstances where you'd want to use alternate shutter speeds:

1) the "Saving Private Ryan"I"Gladiator" stutter effect: try 11500.

2) Slow motion in 60i: try 11120. A film camera running at 60 frames per second would have a shutter speed of about 1/120, so if you're shooting 60i to slow down in post, you probably want to match the shutter speed for film-style motion blur.

3) Sync'ing with monitors: use the Syncro-scan, or 1160, to match the DVX's refresh rate with computer monitors or televisions in the shot and stop the "rolling dark band" syndrome.

4) Special blur effects: the opposite ofthe "Saving Private Ryan" effect. Use a slower shutter speed (like 1124 or slower) to add smear and blur to the motion in your shot.

5) Minimize strobing: If you think there's too much strobing in your progressive footage, you can try a slower shutter speed to introduce a little blur into your footage. 1143 or 1136 are popular choices.

6) Extreme low light situations: when in 24P mode, using 1136 instead of 1150 will gain you half a stop of low-light performance, and still look reasonably like film. Using 1124 will gain you a whole stop of light performance, at the expense of the footage looking quite smeary on motion.

7) To simulate additional frame rates. See page 37 for a discussion on how to simulate frame rates such as 12P and 15P.

8) When shooting under older.magnetic-ballast fluorescent or HMI lights, you may have to select certain shutter speeds to avoid pulsing or cycling of the lights. Similarly, when shooting NTSC under 50hz lights, or PAL under 60hz lights, you may have to adjust the shutter speed to avoid flicker or pulsing.

124

9) Freezing water droplets or rain: for specific instances like shooting a food commercial where someone squirts a lemon, etc., and you want to show the individual droplets clearly, you might try using a very short shutter speed (like 1/1000). Typically these shots are done using strobe lights, but you may be able to get a satisfactory facsimile of the effect by using a super- fast shutter speed.

Some people question why the DVX uses a default shutter speed of 1/50 instead of 1/48, when running at 24P. The reason is, the DVX is emulating a film camera with a 172.8-degree shutter. If it was emulating a film camera with a 180-degree shutter it would use 1/48 as an exposure, but for 172.8 degrees, 1/50 is the proper exposure. Movie cameras don't all default to 180 degree shutters, and they don't all default to 1/48 exposure (at 24 fps). Some cameras run at 1/48. Some run at 1/50. Some run at 1/60. Some run at 1/43. Most modem cameras have variable shutter angles that let them run anywhere from 1/90 down to 1/40th.

1/48 is the shutter speed you'd get if you used a 180-degree shutter. Not all cameras have a 180-degree shutter. Different shutter angles cause different shutter speeds. The Cinema Products CP 161R, for example, comes in three varieties: 144-degree shutter, 156-degree shutter, and 170-degree shutter. Three different exposure times, none of which are 1/48 of a second. The Krasnogorsk K-3 has a 150-degree shutter. The Panaflex L W, Arri SRII HS and Arri SRIIE have 172.8- degree shutters (which perfectly match the DVX's 1/50 exposure time). Most modem movie cameras have variable shutters, with shutter angles anywhere from 11 degrees to 200 degrees. Most have the option of being set at 180 degrees, but can be at many other shutter angles as well. 180 degrees gives a 1/48 of a second exposure time, and looks like film. And the DVX can be set to 1/48.0 in the Synchro-Scan menu, although the difference between 1/48.0 and 1/50 will be nearly impossible to tell under most circumstances.

On the DVXI00A and DVXI00B it's possible to select slow shutter speeds as well (see discussion on page 79). The slow shutter speeds in progressive mode retain full resolution to the picture, unlike on other cameras where setting a slower shutter speed results in getting

125

only one field's worth of resolution. The slower the shutter speed, the more motion blur you'll see as well.

59/60 & 49/52: AUDIO CHlICH2 SELECT: The DVX camera has the ability to record audio from three potential sources: the built-in microphone, XLR connector 1, or XLR connector 2. You can mix and match these to suit your circumstances (i.e., connect a wireless microphone on audio channell, and use the built-in microphone on audio channel 2 for ambiencelbackup purposes). You can also choose to have one input be recorded on both audio channels, by plugging in an audio source into XLR connector 2 and setting both inputs to INPUT 2. When using the built-in stereo microphone, only the left side can be recorded on channell, and only the right side can be recorded on channel 2 (you can't get three tracks of audio at once.)

66 & 54: ZEBRA BUTTON: the Zebras are discussed in the menu settings section, page 92.

67 & 55: OIS BUTTON: The DVX includes an Optical Image Stabilization system, which helps smooth out shaky handheld shots. The OIS system consists of a series of prisms and gyros in the lens that actually move and redirect the image coming into the lens, to detect motion and compensate for it. The OIS effect can be most easily seen when at full telephoto: at full telephoto it's harder to hold the camera still without some shake, as the picture is magnified and so any corresponding shake will be similarly magnified. The OIS tracks any movement ofthe frame and moves the prisms to track the original framing, trying to keep the frame as still as possible. There's a limited amount of compensation it can do, but it works well at full telephoto.

There are times when you will want to tum the OIS off. First, when using the anamorphic adapter, tum OIS off. Since the anamorphic adapter uses "bent glass" to squeeze the image onto the CCD, the adapter is designed such that the-camera should always be looking out through the dead center portion, and OIS would interfere with that, making the camera look out through different parts of the adapter, perhaps resulting in some image distortion.

126

Another time you will want to disable DIS is when using the DVX on a tripod (or other image stabilizing device). When mounted on a trip~d, the DVX im.age will already be adequately stabilized, and any motion that occurs m the frame will be intentional (i.e., if you start panning the camera). However, the DIS system will see that motion and try to ':compensate", actually canceling out your panning motion. The result IS that the DIS will try to "stabilize" your shot, and the further. you pan the further it will try to stabilize it, up until the point where It can no longer compensate (remember that the prism can only move so far before it reaches its limit). The result will then be a noticeably jerky motion in the pan. If you're using a tripod, there's no reason to have two image stabilization devices trying to do the s.ame job, so tum DIS off and you'll get cleaner, smoother pans and tilts,

Also, ifusing a lens adapter such as the P+S Technik Mini35™ or simi!~r d:vice, you mo~t definitely need to tum optical image stabilization off. Leavmg DIS enabled while using one of those ground-glass image adapters will result in a very weird effect of the whole frame "floating." When using those adapters, definitely tum the DIS off.

127

ANAMORPHIC ADAPTER USER GUIDE

With the introduction of the LA-7200G Anamorphic Adapter, Panasonic has produced a solution for getting a native, full-resolution widescreen image from the DVX cameras.

Why anamorphic?

The DVX shoots video at a 4:3 aspect ratio, which is the same shape as a normal television screen. This aspect ratio can also be expressed mathematically as 1.33:1, meaning that the picture is 1.33 times as wide as it is talL It's basically a square, 33% wider than it is tall. Practically all video cameras (at the time of this writing) with CCD's smaller than 2/3" have 4:3 CCD's.

Movies at the theater are shown ''widescreen'', at an aspect ratio of 1.85:1, and some movies are shown in "scope" ratio at approximately 2.35:1. Both of these aspect ratios are much wider than a normal TV. And widescreen 16:9 televisions are becoming more popular. HDTV sets and Plasma TV's are usually 16:9. "Anamorphic Widescreen" DVDs are optimized for display on widescreen televisions. The 16:9 aspect ratio is approximately 1.78:1, which is very, very close to the movie theater ratio of 1.85: 1. Accordingly, when viewing a movie from DVD on a widescreen television, the entire movie fills the entire screen of the widescreen TV with no "letterbox bars." Even many

4:3 televisions are now including an "enhanced 16:9'~ mode, where all the lines of resolution of the TV are scrunched down into a 16:9- shaped area, which effectively makes the image look letterboxed, but uses all the lines of resolution to provide a picture competitive in resolution with a native widescreen television. Widescreen televisions are very popular in Europe and are becoming more and more prevalent in the U.S. as well.

Many DVX owners would prefer to shoot widescreen imagery because they prefer that aspect ratio and believe it is more "cinematic". Many DVX users also want to release their movies and videos on DVD, and would like to make an "Anamorphic Widescreen" DVD, enhanced for display on widescreen televisions. And many DVX filmmakers want to shoot in the widescreen aspect

128

ratio because they intend to (or hope to) blow their final project up to 35mm film for projection in a theater, and shooting widescreen makes for a much cleaner blowup. The Panasonic Anamorphic Adapter is an option that can convert your DVX camera from shooting native

4:3 footage to shooting native 16:9 widescreen footage, perfect for anamorphic DVD's, widescreen TV's, up-rez to High Definition, and film transfers.

Anamorphic vs. Squeeze vs. Letterbox vs. Convert in Post

On the DVXI00 there are two ways to get an image in-camera that looks like a widescreen image: either using the optional anamorphic adapter, or shooting in Letterbox Mode. The DVXlOOA and DVXI00B add an additional option: Squeeze Mode. Additionally, there is a fourth option, which is to shoot in Letterbox Mode, but "stretch" in post to convert your footage to be anamorphic. There are certain circumstances where each mode is appropriate.

Letterbox Mode: Letterbox mode is not the same as shooting 16:9 footage. Letterbox mode adds black bars at the top and bottom of the footage, leaving a middle patch that is shaped 16:9. The result is similar to watching a letterboxed movie on your 4:3 TV set: black bars at the top and bottom, with a widescreen-shaped image in the center. Letterbox Mode adds 54-pixel-high black bars at the top and bottom of the image, leaving the central 372 lines to record the widescreen image (note: these are NTSC numbers). Letterbox Mode is appropriate for displaying a widescreen-looking image on a regular 4:3 TV. It is not appropriate for display on a widescreen TV, wide screen HDTV, widescreen plasma TV, for making an anamorphic DVD, or for blowup to film (although Letterbox would be preferable to Squeeze mode for a film blowup or HD up-rez - see Footnote 1). Letterbox Mode footage is still 4:3 footage, and displays properly on a 4:3 TV set, and does not display properly on a widescreen 16:9 television set. Letterbox mode is appropriate for making VHS dubs of your footage for playback on a normal 4:3 TV, for example, or for making a non-anamorphic DVD.

129

Squeeze Mode: The DVXI00A and DVXI00B add a new option: Squeeze Mode. Squeeze Mode is a way to get an anamorphic 16:9 image in the camera at the time of shooting. Squeeze Mode footage is appropriate for displaying on a wide screen television, or for making an anamorphic DVD. Squeeze mode is not appropriate for displaying on a normal 4:3 television, and not optimal for making a film blowup or HD up-rez (see Footnote 1). Squeeze Mode uses a similar image as Letterbox Mode, but Squeeze Mode digitally stretches the central image to become 480 (NTSC) pixels tall. The result is a significant loss in resolution as compared to anamorphic, but the image is now suitable for play on a widescreen television.

Stretching In Post: The original DVXI00 doesn't offer Squeeze Mode as an option, but it can be simulated in post production, by cropping the frame to 720x372 (NTSC) ~nd then instructing the editing program to stretch the 372-lme image to fill the full screen. Depending on the quality of the resizing algorithm in your editing program, the results can be quite comparable to Squeeze Mode. I personally have tested Sony Vegas 4.0d and 5.0b and found the stretching algorithm to be comparable to Squeeze Mode, but the results in Premiere 6.0 were much less satisfactory. The results from stretching

in PhotoZoom Pro can be quite good.

Anamorphic Adapter LA-7200: The final way to get widescreen footage is to use an optical anamorphic adapter. At the time of this writing (9/2005), the only optical anamorphic adapter that works with the DVX cameras is the Panasonic LA-7200. The Anamorphic Adapter utilizes the full resolution of the camera to record the image (as opposed to both Squeeze and Letterbox Modes, where 25% of the frame goes unused.) The anamorphic adapter optically bends light rays to squeeze 33% more horizontal information into the frame, while preserving the fun height of the frame. The result is a full-frame 720x480 image that is immediately suitable for display on a widescreen television, for making an anamorphic DVD, for up-rezzing to High Def or for film

130

blowup. Footage shot with the anamorphic adapter is not suitable for display on a standard 4:3 television, although you could always "un-squeeze/letterbox" your footage in post to create a 4:3 image with black letterbox bars. Using the anamorphic adapter gives you a camera that is roughly equivalent to one that has native 16:9 CCD's: you'll get a fullresolution image across the entire surface of the CCD that is widescreen in nature and suitable for play on a widescreen television. The anamorphic adapter potentially provides the very highest resolution 16:9 imagery possible from these cameras, but not without some limitations (discussed below).

How does Squeeze compa.re to Ana.morphic?

You cannot directly compare Squeeze Mode to Letterbox Mode, because they don't produce video that's the same shape. Letterbox Mode produces video that is shaped 4:3, with black bars and a widescreen-Iooking image in the center. Squeeze Mode produces video that is 16:9 in shape .. You can convert Letterbox Mode into something directly comparable to Squeeze Mode, caned "Stretched Letterbox". For this discussion "Letterbox Mode'; will have limited applicability, but we will include "Stretched Letterbox" in the discussion since it represents a way to accomplish 16:9-shaped video.

The Anamorphic Adapter provides potentially the highest resolution possible from the DVX series of cameras. Squeeze Mode (or Stretched Letterbox) will also provide anamorphically-shaped video, but with a loss of resolution. What type of resolution loss are we talking about with Squeeze Mode (or Stretched Letterbox), and how much higher-resolution can the anamorphic adapter provide? For comparison, here are three resolution charts. To read the chart, look at the wedges oflines, and notice the point at which you can no longer distinguish four individual lines (the charts are marked with arrows to point where the "mushing together" begins. (All charts were shot in progressive-scan, with THIN line detail).

131

converge into a gray mush. On the larger wedge (at the bottom) the lines start to converge at about the 400 tine mark. On the smaller wedge (which starts at 400 lines, and should consist of four distinct lines) it's impossible to discern four individual lines.

Also, notice that the circle at the center of the image is basically a wide oval (owing to the 4:3 aspect ratio, and the way DV pixels are non-square) .. This image is not re-sized or re-shaped, it's straight from the camera (and then cropped and with an arrow added to it). NTSC DV images come out of the camera at a size of 720 x 480 pixels, which is mathematically a wider aspect ratio than 4:3 (instead of 1.33: 1, it actually mathematically reduces to 1.5: 1). The TV screen takes care of resizing the image and the pixels to make it look the right shape.

LETTERBOX MODE

the that

up

I I Ii

132

The horizontal wedges near center show the camera can cleanly resolve to about 380 lines of resolution before the lines start to

STRETCHED LETTERBOXISQUEEZE MODE

/

I

Notice on the above chart that the shape of the video is different. The image has been stretched vertically to fill the full frame top to bottom, so the shape of the circles is different. The circles are now vertical ovals, but if you imagine this image being displayed on a widescreen TV, you can see how the TV would stretch the image horizontally, which would make the circles round again.

There is no additional resolution though - the image tops out at about 350 lines of resolution, a bit less than straight Letterbox Mode. The slight loss of resolution is due to the digital stretching process. Letterbox footage is "cleaner", because it hasn't been stretched yet (which also means it's not really 16:9-shaped). Squeeze mode footage is a Little bit lower resolution, because it's been through a digital stretching process, but it's 16:9-shaped.

133

/

ANAMORPHIC ADAPTER

I I I

On this chart, shot with the anamorphic adapter, the extra resolution is immediately apparent. On the horizontal wedge you can clearly discern four individual black lines all the way to about 540 lines of resolution; after that it starts to blend into mush. 540 lines is the theoretical limit of DV-format resolution, and the DVXlOO with the anamorphic adapter delivers full DV resolution in a 16:9 shape.

Always Use TIDN or MID Line Detail, Never THICK

When using Squeeze Mode, never use THICK line detail. Always use THIN or MID, preferably THIN. THICK lowers resolution by employing a vertical blur filter in order to overcome thin-line flickering when displayed on a television set, but the stretching process used in Squeeze Mode also blurs the image enough to overcome that problem. If you use THICK mode with Squeeze Mode, you'll get about 270 lines of resolution maximum, vs. the 350 you could get from using THIN mode. When using the Anamorphic Adapter, if you use THICK mode, you'11 get a maximum of360 lines of resolution- practically no better than if you'd used Squeeze Mode/THIN in the first place! You'd end up with al1 the restrictions of the Anamorphic Adapter but none of the benefits. So always use THIN line detail when shooting with either Squeeze Mode or the

134

Anamorphic Adapter. If your Anamorphic Adapter picture ends up being too high-resolution for proper display on an interlaced television set (a very distinct possibility), you can always employ a "reduce flicker" filter in post to introduce a slight vertical blur and bring the picture into spec.

How does it work?

"Anamorphic" is a term that has Greek roots, which means (roughly translated) "to reshape". An anamorphic lens literally changes the shape of anything viewed through it. An anamorphic lens is like a wide-angle lens, but only on the horizontal axis: the vertical axis remains unmodified. The amount of image change depends on the lens: there have been cinema lenses with squeeze factors of 1.5:1, 1.75: 1, and 2: 1, to name a few. For video, anamorphic adapters use a 1.33:1 squeeze factor. The anamorphic adapter sees the same field of view as the normal camera lens (on the vertical axis) but gathers in 33% more image on the horizontal axis. The result is a picture that (when properly displayed) is 33% wider than the stock camera.

Anamorphic lenses have been used in movies since the 1950's, when the introduction of television was viewed as a threat to movie producers, who experimented with many different techniques to keep the theater-going experience unique. One thing they tried (that stuck) was making movies wider. Television was introduced with a 4:3 aspect ratio, because movies were made in a 4:3 aspect ratio. Nearly all films made prior to the 1950's were shot at 4:3. But to encourage people to keep going to the theater, producers started making widescreen movies.

Two methods are used to make movies wider: either cropping down the image, or using anamorphic lenses. The majority of movies are shot using the cropping method: a 1.33: 1 film frame is cropped to 1.85:1 for "flat" projection (the same technique used for Letterbox Mode) or, for Super35, the 1.33:1 film frame is cropped almost in half, all the way to 2.3 5: 1, which is then optically stretched to fit an anamorphic film frame (a very similar technique as used by Squeeze Mode). The third method is to use optical anamorphic lenses, which optically squeeze the image at a 2: 1 ratio, fitting twice as much horizontal image on the film frame (the film is cropped to 1.2:1 in the camera, and the optically squeezed image results in a full-frame

135

image that uncompresses to approximately 2.39: 1, which is the standard for wide screen anamorphic motion pictures). This technique is the same principal as used by the Panasonic anamorphic adapter: an optical squeezing of the image that fills the entire frame of the recording medium.

The Super35 cropping/squeezing technique results in a widescreen image with the same aspect ratio as an anamorphic film, b~t with a softer image and more grain as compared to the anamorphic film. Shooting Super35 for widescreen has its advantages (quicker setups, cheaper lenses, etc), but compromises image quality; shooting anamorphic requires more work on the part of the cinemat~gr.apher . and crew, but there's no denying that the optical anamorphic Image IS cleaner and sharper.

The same can be argued of the Panasonic adapter: shooting Squeeze Mode is quicker, faster, and involves only one compromise: lower resolution. Shooting with the Anamorphic Adapter involves several compromises, but the one place you don't compromise is on . resolution: the image is higher resolution when properly shot WIth the anamorphic adapter.

The resolution advantages become more apparent the larger the screen you display your work on. On a 9" television, for exam~le, it would be very hard to tell the difference between footage shot ill Squeeze Mode and footage shot with the anamorphic adapter. On a 20" TV, the difference may be more noticeable. ~n a 32" TV the anamorphic footage will be significantly sharper ~If shot properl!'). Projected on a 10' diagonal screen the anamorphic's advantage IS more dramatic, and moreso for a film blowup.

Limitations and Restrictions

Using the anamorphic adapter involves several restrictions that you need to be aware of. While the potential resolution is superior to Squeeze Mode, there are a number of restrictions placed on the camera when the adapter is installed. Squeeze Mode faces none of these restrictions, since its.image manipulation occurs entirely within the camera. Some of the anamorphic adapter restrictions include:

136

No Additional Lenses: the anamorphic adapter is a physical lens attachment that fits on the front of the camera. You cannot use it in combination with an additional wide-angle adapter or telephoto adapter. While this may appear to be a significant restriction, in practice it's hardly an issue: you don't normally need a wide-angle adapter because the anamorphic already gives you a 33% wider field of view, which makes it equivalent to using a .67 wide-angle adapter. As for telephoto, there is only one telephoto adapter available for the DVX cameras, and it provides only a modest 1.6x extension, only allows for using part of the zoom range, and costs over $650, so many users don't even have it. While it would be nice if additional lenses were possible with the anamorphic, this hasn't really proven to be a major restriction.

Few Matteboxes: because the anamorphic adapter has a square lens hood (instead of a round lens front), very few commercial matteboxes will fit it. As of the time of this writing, only the $1700 Chrosziel411-53PK system will fit the Panasonic anamorphic adapter, or the CineTactics Anamorphic MatteBlox. A camera using Squeeze Mode could of course use just about any mattebox, so this limitation applies to using the anamorphic adapter. Again, it sounds like more of a restriction than it truly is. Matteboxes provide two main purposes: a lens hood, and a way to mount filters. As for lens hood, the anamorphic adapter includes its own lens shade, which obviates some of the need for a mattebox. If additional shading is needed, you can use the Cine Tactics MatteBlox French Flag and some Velcro™ to attach it, or you can use an inexpensive Flarelsuster=, a repositionable flag on a flexible extension arm that lets you provide shade for the lens exactly where you need it. Neither solution is quite the same as having a "real" mattebox, but they can work and they're much less expensive as well. If you need a "real" mattebox, the aforementioned Chrosziel is one possibility; another possibility is to rent an Arri 6x6 mattebox.

Optical Image Stabilizer: It is recommended in the user manual to not use the OIS when using the anamorphic adapter. The reasons are not given, but a thoughtful analysis by Neil Rowe on DVXUser.com shows what the likely cause is: the optical image stabilizer works by using a series of gyros and motors to position

137

prisms to actually redirect what portion of the lens the camera "sees" through. While that's fine for cameras that don't have the anamorphic adapter installed, the anamorphic is designed to bend light in a certain way on a certain axis, and if the gyros and motors move the internals of the lens so that they're not looking out through the optimal portion of the adapter, some image distortion may arise. While the loss of the image stabilizer is regrettable, again it can be overcome: keep in mind that no film camera has an image stabilizer, and the movies seem to look pretty good even without it! The image stabilizer is most useful during handheld shooting, and should always be disengaged when shooting from a tripod or other camera stabilization device (see Footnote 2). Since the focus/aperture restrictions that accompany the anamorphic adapter make it not well suited to "run On' gun" handheld-style productions, the loss of the image stabilizer becomes a relatively moot point. For best results with the anamorphic adapter, careful attention to focus and shot setup are necessary, and those types of setups will also often include a tripod or dolly or Steadicam™ or other type of camera support - and in those cases, the optical image stabilizer should be disengaged anyway (see Footnote 2) so it's not really a significant limitation.

Filters: Perhaps the most significant restriction when using the anamorphic adapter is the restriction on using filters. The anamorphic adapter mounts directly to the lens threads of the camera and precludes the use of screw-in filters. There is no provision to mount screw-in filters in front ofthe adapter, and the design of the adapter makes it impossible to mount filters between it and the camera. So cheap, plentiful 72mm screw-in filters are not an option when using the anamorphic adapter. However, this is, again, less of a restriction than it may seem. First, many shooters prefer to shoot unfiltered, and apply all effects in post. Second, the camera can perform some filter functions in-camera (for example, the effect of a warming 81 filter, or a cooling 80A filter, can be accomplished by using the COLOR TEMP setting in the camera, and color-shift filters may also be simulated by using the CHROMA PHASE setting). Finally, for certain effects that can only be done by using filters (such as a polarizer or a Neutral Density-Grad), some options do exist:

138

1) The Chrosziel 4ll-53PK 4x4 mattebox allows the use 4x4 or 4x5.65 filters. This option also involves modifying your anamorphic adapter by shaving off most of the lens hood.

2) You could rent the aforementioned Arri 6x6 mattebox and use 6x6 filters.

3) For a no-budget option, it should be noted that the interior dimensions of the anamorphic adapter's lens shade are ALMOST big enough to accommodate a 4x4 filter, as long as it is fastened securely. If it wasn't for the slightly angled comers on the adapter, you could easily wedge a 4x4 filter in there, and add a little Velcro™ retaining strip to keep it fastened. If you wanted to use Lee or Cokin Z-Pro resin filters, you could trim just the comers off the filter and it would fit snugly into the adapter's lens shade (again, make a restraining strip ofVelcro™ to keep the filter from slipping out.)

4) Bob Watson at Evergreen Pictures has also adapted a Cokin X-Pro series filter holder to fit the front of the anamorphic adapter - you can find detailed pictures at http://www.evergreenpictures.com/DVX100IDVX16x9.htm. Cokin X-Pro filters are huge resin filters that will easily cover the surface of the adapter. Cokin filters are typically much less expensive than glass filters of a similar size, although some believe the glass filters to be of superior optical quality to the Cokin resin filters.

FOCUSING ISSUES

Perhaps the most important thing to understand about the anamorphic adapter is how it affects the camera's focus mechanisms, and why, and how to work with it. Simply put, it is more difficult to achieve proper focus with the anamorphic adapter installed. However, with a little effort, you can easily achieve perfect focus under most conditions.

139

Why is focusing weird with the anamorphic?

Remember that the anamorphic adapter "bends" light on the horizontal axis, but on the vertical axis it remains unmodified. This unequal curvature introduces a rather unique property to the lens: it focuses on the horizontal axis at a different plane than it focuses on the vertical axis! You cannot get both the horizontals and the verticals to focus at the exact same point, because the curvature of the lens makes for differing focal points. These different focal points change based on how close the camera is to the subject, and the telephoto setting of the lens.

Autofocus Will Not Work

Compounding the problem, you cannot use autofocus with the anamorphic adapter. The autofocus system in the Panasonic cameras focuses on one axis only: the verticals. It ignores horizontal data and only takes vertical information into account when determining the proper focal point. This is not a problem when using the stock lens, because with the stock lens both the horizontals and the verticals will focus at the exact same plane. However, with the anamorphic adapter installed, only the verticals will be focused properly (and they will be focused very sharply indeed) but the horizontals may be completely out of focus, and that makes the overall picture look out of focus. If you enable autofocus with the anamorphic adapter installed, you are basically guaranteeing that your footage will be out of focus.

Needless to say, this can make focusing quite a challenge. Many have found that it's extremely frustrating to try to focus using just the camera's LCD. But there are ways to guarantee perfect focus with the anamorphic adapter, provided you take the time and care and that you control the aperture.

Depth Of Field Overcomes The Astigmatism

The key to getting sharp focus when using the anamorphic adapter is to control the depth of field of the camera in order to bring both the horizontal and the vertical plane into focus. Depth ofField (DOF) can be a complex subject, but in this guide we will give you the tools you need to easily compensate/or most any circumstance and know that you're going to get sharp focus on your subject.

140

Depth of Field (DOF) refers to the area in front of and behind a given subject, the range of which all falls into sharp focus. When focusing a lens, only one infinitely-thin plane is ever truly "in focus" (the "focal point") - everything else is progressively more and more out of focus the further it is from the focal point. The trick to focusing the anamorphic adapter is to determine the focal point for the horizontals, and the focal point for the verticals, and then split the difference to

get the proper focal point. Then you use the depth of field to bring both planes into sharp focus.

You can't pick just any combination of distance, zoom setting and aperture, and expect to be able to focus. That works for Squeeze Mode, but not for the Anamorphic. When using the Anamorphic Adapter, for any given distance, combined with any given zoom setting, there is a maximum aperture opening that you're allowed to use if you want to keep the shot in focus (see Footnote 3).

To get proper focus with the anamorphic adapter, you have to a) determine the proper focal point, and b) determine the proper minimum aperture. Unfortunately, neither is easily set by eye or by the camera's viewfinder and LCD. The anamorphic image, when properly focused, delivers a fu11480 pixels of resolution in the frame, and the camera's LCD only shows about half that much resolution. The LCD cannot be relied on as your sole focusing aid. Fortunately we have included charts and reference cards that tell you how to focus with much greater accuracy than is possible by using the LCD alone!

Examples Of The Different Focal Planes

The picture at left shows the anamorphic adapter focusing on a resolution chart, where the horizontals are sharply in focus, but the verticals are completely out of focus. And the impression of the overall picture is that it's out of focus.

141

The first step in the process of getting proper focus is to determine the right focal point, which is a compromise position between the best focus for the horizontals and the best focus for the verticals. The picture at left shows a compromise between the horizontal and the vertical focus. It looks better than the other two (notice the number "200" on the left side), but still looks out of focus. This is about the best you would expect to get if you were focusing strictly. by the LCD or by a high-resolution monitor, and didn't take aperture into account.

Determining The Right Focal Point .

There are several ways to determine the right focal point. One that IS often mentioned is to use a high-resolution (500+ lines) CRT field

142

The picture at right shows the exact same frame, the exact same chart, only this time the focus was set to sharply focus on the verticals. The horizontals are now completely blurred and out of focus, but the vertical lines are much sharper and more clear.

The picture at right shows the result of using the proper focal point AND the proper aperture. By stopping down the lens far enough so that both the horizontals and the verticals fall within the depth of field, you can bring out the best results possible with the anamorphic lens.

monitor. That solution can work, but it's also expensive, clumsy to haul around in the field, and difficult to power in the field, It can also prove frustrating and inaccurate, if you don't know the charts and numbers included in this guide. When you have circumstances where the verticals and the horizontals are focusing at different points, eyeballing a monitor to determine the best focal point can lead to guesswork and inaccuracy. In this guide we have developed three ways to precisely determine the proper focal point in any given situation: a distance chart, an autofocus-to-anamorphic conversion chart, and even some techniques that you can use in the field when you've forgotten the printed charts.

The Charts

When you have the time and the ability to drag a tape measure between the subject and the hot shoe of the camera, you can use the distance measurements on this chart to get the right focus setting. This chart was inspired by the excellent chart first produced by Evin Grant and published on the DVX forum at 2-pop.com.

We have gone through and verified the original distances, added close-focus distances for the DVX100A and DVXlOOB, and then determined the proper anamorphic adapter settings for each distance .. There is obviously a little wiggle room here - the lens ring can move quite a bit between MF numbers -- so when using the distance guide, double-check your settings on a high-resolution CRT monitor, or stop down an extra stop or so to have adequate depth of field.

To use this distance cbart, place your tape measure at the back of the camera's accessory shoe and measure to the subject. Crossreference that distance to the "Anamorphic" column on the chart, and that's what you set your focus point to (the MF setting in the viewfinder) .. For example, if your subject is 11' 6" away from the back of the camera shoe, set your MF to 80 when using the anamorphic. Keep in mind that you'll still have to set your aperture: getting the right focal point is just the first step in properly focusing the anamorphic adapter.

143

Using Autofocus

Dragging a tape measure around can be fairly precise, but it can also be a burden. An easier way would be to let autofocus try to figure out what the proper setting is, and then use this chart to look up the right setting for the Anamorphic. To use this chart, put the camera in Autofocus mode (on a DVXIOO you'd have to switch out of

Focus Setting Focus Se,tting
Distance Normal Anamorphic Distance Normal Anamorphic
1 ' 11.5" 37 41. 5 6' 0" 66 69.5
2' 0.75" 38 42.5 6' 31' 67 70
2' 1.5" 39 43 6' 6" 68 71
2' 2.25" 40 44 6' 9" 69 72
2' 3.25" 41 44.5 7' A" 70 73
2' 5.25" 42 45 .. 5 7' 4" 71 73.5
2 ' 6 .. 75" 43 46.5 7' 8" 72 7 4.
2' 8.25" 44 48 8' on 73 75
2' ro: 45 50 8 ' 6" 74 76
3' 46 51. 5 9 ' 0" 75 77
3' 2.25" 47 53 9 ' 6" 76 78
3' 4" 48 54 10' 0" 77 78.5
3' 6,75 49 55.5 la' 8" 78 79
3 ' 10" 50 57 II' 6'i 79 80
3' 11" 51 58 12' 4H 80 81
4 ' 0" 52 59 13' 4" 81 81
4 ' 1" 53 59.5 14' 7" 82 82.5
4 ' 2.5" 54 60 16' 0" 83 83
4' 4" 55 61 17' 9" 84 84
4' 5 'I~ 56 62 19' 11" 85 85
4 ' 6.5" 57 62.5 22' 9" 86 86
4 ' 8" 58 63 26' 6" 87 86.5
4 ' 10" 59 64 31' 10" 88 87
4 ' 11.5" 60 65 39' 9" 89 88
5' 0" 61 65.5 52' 11" 90 89
5' 2" 62 66.5 79' 5" 91 89.5
5' 4:" 63 67 160' 8'" 92 91
5 ' 6" 64 68 inf 95 93
5 ' 9,5" 65 69 144

progressive scan mode to do this), and wait for the autofocus motor to choose a focus point (which will be wrong, but we'll use this as a starting point). It will report this as an AF number in the viewftnder/LCD). Look up that AF number setting in the "Normal" column. Then switch to manual focus and spin the focus ring to the corresponding entry in the MF column. For example, if the camera came back with an AF setting of 68, you'd then look in the chart under "AF" and find 68, then look to the right to the "MF" column and see that the proper MF setting should be 63. You'd then set the camera on Manual Focus and dial the focus ring to 63, and you now have the proper focal point.

AUTOFOCUS TO ANAMORPHIC CONVERSION

(AF 46 through 62 apply to the DVX100A and DVXlOOB only)

83 83

84 84

85 85

86 86.5 87 88

88 89

91 93

46 42
47. 43
~:i.5 43
4,13 44
45
46
47
48
50
52 66
67
68
69.5
71
75 72
76 74
75
76
77
78 58 59 60

iiS7 61

[67.5 62

1',,::;, :~

1:;69 5 65 L"';;,;;·!'

Note: the chart begins at AF46/MF42, which is the closest that a DVXlOOA or DVX100B can focus at full telephoto. The numbers AF46IMF42 through AF62/MFS6 are outlined because they are only applicable to the DVX100A and DVXIOOB. The DVX100 chart begins at AF64/MFS7 because that's the closest focus point you can get with the anamorphic adapter on a DVXIOO at full telephoto. It corresponds to a distance to subject of about 3' 10". At 3' 1 0" the verticals are in focus at MF64 and the horizontals are in focus at MFSO, which is the closest distance possible to focus on a DVX 1 00 at full telephoto - which makes the proper focal point MF57. You can't get closer than that with a DVXlOO, although if you use wide-angle

145

When using Autofocus it is VITAL that you zoom in all the wayan your subject and open the iris all the way. Use the ND filter switch to compensate if your subject gets too bright. The autofocus motor will only return the proper number if you are at Z99 and F/2.8. Any wider on the lens (i.e., wider-angle shot, or lower Z-number), or any aperture other than f/2.8, will increase the depth offield and fool the autofocus motor into possibly returning incorrect values.

settings you can focus all the way down to the surface of the anamorphic adapter's lens.

On the chart there are some "halfway" numbers, like AF77.S. Sometimes the autofocus motor will try very hard to determine between two AF numbers, and will fluctuate rapidly between two numbers (like 77 and 78). In cases like that where the AF motor isn't "absolutely sure", we split the difference and added a decimal point

to the chart. Additionally there are some decimal-point numbers in the MF column (such as 86.S). This is a result of there being a little "free play" on the focus ring between numbers, so we compensated and did the best we could to spell out the exact point on the focus ring where best focus was achieved. To hit 86.5 you'd move the focus ring between 86 and 87 until the point just before the number turns to 87. As before, it's always best to double-check your work on a highresolution CRT monitor, and stop down to the proper aperture (discussed below).

Focusing If You Forget The Chart

Focusing by using the charts is very simple and fairly foolproof, but what if you're in the field and you forgot the chart? Here's a handy trick to manipulate autofocus into delivering the right numbers. First, make sure the anamorphic adapter is properly attached and "squared up", and then zoom in to Z99 and open the aperture to F/2.8 (use the NO switch and/or shutter speeds to compensate if the picture gets too blown-out). Then switch to autofocus mode, and let the camera lock in on what it thinks the proper setting should be. Remember, the

auto focus system only calculates focus based on the verticals, so the number it reports will be wrong for the overall picture, but we still need to know it. Once the camera settles down and gives you an AF

146

number, make note of that (let's say it's 64). That's the number for the verticals.

N ow we have to determine the proper focus point for the horizontals. Do that by using the anamorphic adapter to "lie" to the autofocus system. Loosen the leveling screw, and rotate the adapter 90 degrees, to where the leveling screw is pointing straight up, directly at the center of the on-board microphone (see the picture at the right). Then take another autofocus reading. This time the autofocus system will again be looking at the verticals, but with the bent glass "out of the way", it'll report a different number. For our example, let's say the autofocus system gives you a new number of AF50. We now know that the horizontals are at AF50, and the verticals were at

AF64. To get the proper focal point, just choose the number that's right in the middle (in this case, MF57). Switch to manual focus, dial in the midpoint between the two AF focus numbers (i.e., dial to MF57 in this case), and re-level the anamorphic adapter. Then re-set your zoom setting, select the right aperture and you're ready to shoot!

Other Focusing Techniques

Neil Rowe ofDVXUser.com offers two additional techniques for achieving proper focus with the anamorphic adapter when in the field and without the charts or graphs. First, always focus for the subject's eyes. The eyes are the point we as viewers are drawn to, and are most important to he in focus. Neil suggests adding an eyelight, a pinpoint oflight reflecting in the subject's eye. Then focus based on that pinpoint - when you're in proper focus, the pinpoint should be perfectly round .. The more out of focus you get, the more distorted and oval it will be - focus too far and it'll be a vertical oval, focus too close it will be a horizontal oval.

147

Secondly, Neil recommends the use of a focusing aid, a picture of vertical and horizontal lines with an "X" in the center. Have the subject hold it up exactly where their face would be. If only the horizontal lines are in focus, or only the vertical lines, then you know the overall picture is not sharp. When you hit the proper combination of horizontal and vertical focus, and aperture, then the "X" in the center should pop into sharp focus (because the "X" is made of diagonal lines, it can only be sharply focused when both the horizontals and the verticals are sharply focused.)

Determining the Minimum Aperture

Once you have the right focal point, you then need to extend the camera's depth offield so that it will cover both the horizontal and the vertical plane. This is dictated by the zoom setting, focus setting, and aperture. Your shot framing will determine the zoom setting, and you already know the focus setting, so now you can determine the proper aperture. Too wide-open (see footnote 3) and the depth of field will be too shallow, resulting in out-of-focus pictures. You have to pick the right aperture to make sure the depth of field is deep enough to get both the verticals and the horizontals in focus.

Adam 1. Wilt, one of the legends of the DV Community, has published a chart that shows the apertures you must use at a given zoom/focus setting. The closer your subject, or the more zoomed in your shot, the smaller the aperture you need to use. If you focus using the techniques in this guide, and follow this chart, all your shots wi11 be razor-sharp and take full advantage of the highest resolution the DVXlOO can deliver in l6:9!

148

ANAMORPHIC ADAPTER MINIMUM APERTURE CHART

6.8

NG = never sharp; - = can't focus DVXIOO that closely. © 2004 Adam J. Wilt

The chart was originally published at http://adamwilt.coml24p/7200chart.html and is reprinted here with permission of the author.

To use this chart, you need to know what your F (Focus) and Z (Zoom) settings are, and then cross-reference them on the chart above. Look up your F setting on the vertical bar, then follow the chart across to find your Z setting on the horizontal bar. Where they intersect, that's the aperture you have to use to guarantee sharp focus. For example, for an F of 50 and a Z of 30, you would need an f-stop of2.8 in order to get sharp focus. For an F of70, and a Z of99 (full telephoto) you would need to set the camera all the way to fl14 to get adequate sharpness. For an F of 60 and a Z of 20, you could get adequate sharpness with just fl2. When using this chart, keep in mind that bigger numbers = sharper picture, so if you're on the border between fl3.4 and fl2.4, err on the side of caution and use fl3.4. Also, if the chart says you need fl2.8 but the autoexposure system in the camera wants to use fl5.6, that's okay: bigger number = safer, use fl5.6. But if the chart says that focus needs fl6.8, and the camera wants fl2.8, that's where you have a problem, and you have to solve it through adding lighting, or through reframing your shot (using a wider-angle shot). Adam acknowledges that the chart is probably conservative, and you may find that you get adequate focus with slightly wider apertures than the chart recommends, but use the chart recommendations for safety.

149

Shallow Depth Of Field Problems With The Anamorphic

Getting a shallow-depth-of-field effect is more difficult with the anamorphic adapter than it would be with Squeeze Mode. Because of the need to use deeper apertures to extend the depth of field across both the horizontal and the vertical planes of focus, it's very difficult to get the wide-open aperture and long telephoto shots necessary for a shallow DOF and a soft out-of-focus background. The anamorphic excels at wide shots, giving much more detail in wide-angle shots, but can be difficult to get acceptable shallow DOF shots with. In cases where shallow DOF is a must, you may try removing the anamorphic adapter and using Squeeze Mode (or StretchedLetterbox) for those shots. The shallow depth of field effect is most noticeable on closeups, and the resolution loss from going to Squeeze Mode may be offset by the artistic look of the out-of-focus background. It's an artistic choice for you to make, but in circumstances where shallow depth of field is absolutely mandatory, Squeeze Mode can do a better job. The other way to accomplish it is to use a SoftScreen soft-focus background device, available at http://indietoolbox.com. The SoftScreen will give the shallow-depth-of- field look to any background, regardless of the camera's aperture setting, so if you simply must have the anamorphic adapter on a close-up head shot with a shallow depth-of-field look and soft-focus background, the SoftScreen is a way you can accomplish it.

Unsgueezing The Image For Framing

When shooting with the anamorphic adapter (or with Squeeze Mode) on a DVXI00 or DVXI00A, the image in the viewfinder and on the camera's LCD is shown "stretched". Everyone will look tall and skinny. This is because the LCD and the viewfinder are 4:3 display devices, whereas the video is expecting to be displayed on a widescreen television. This "stretched" display can take a little getting used to. There are ways to ''unsqueeze'' the image for proper

viewing: .~

A 16:9-capable monitor attached to the camera, either a 16:9-shaped monitor or a monitor or television that's switchable between 16:9 and 4:3. Unless it's a professional production monitor with 500+ lines of resolution, the monitor may not.be adequate to judge focus with, but it will let you see the unsqueezed image for proper framing.

150

A portable DVD player with a widescreen display and a video input port. There are many clamshell and tablet-style portable DVD players available now that include 16:9-shaped LCD screens. Some of them have video input ports, which would allow you to hook up the composite monitor-out of the DVX to the video input of the DVD player, using the DVD player's screen as a widescreen monitor. Portable DVD players don't typically have high-resolution screens so this isn't necessarily a high-quality way to monitor your footage, but it will unsqueeze your footage so you can see it in the proper aspect ratio.

Century Optics also makes an attachment that fits onto the eyepiece of the DVX cameras and optically unsqueezes the image from the viewfinder. This eyepiece is perhaps the most convenient way to unsqueeze the image, involving no additional cables or monitors to track. It's Century part # VSWSEP-DVX; they also make a widescreen LCD magnifier that fits over the LCD panel and unsqueezes the image.

Using a laptop computer can be another way to view the footage live, unsqueezed. There are various video capture programs available that can preview your footage on your computer screen, live from the DV port, and some of them (such as Vegas 5.0 and certain Macintosh programs) can be configured to display the live footage in an ''unsqueezed'' wide screen image. This is not necessarily a convenient way to go, but it can be effective, and having a laptop computer on the set can prove handy in other ways as well. If you use Serious Magic's DV Rack software, you'll get a widescreen monitor as well as several other valuable production tools, including a digital disk recorder. DV Rack makes a great companion to the DVX and can be a good external-monitor solution for the anamorphic adapter.

The DVXIOOB's LCD and Viewfinder can also be configured to unsqueeze the Anamorphic image.

151

APPENDIX

ANAMORPHIC ADAPTER FAQ

(originally published on 2-Pop.com and updated for this guide)

The LA-7200G is an anamorphic adapter for the DVX that attaches to the front of your camera's lens. It optically "squeezes" the camera's image so that it becomes 16:9. Having this adapter on your camera gives you effectively a native 16:9 camcorder.

How does it work?

The anamorphic adapter uses "bent glass" to squeeze a 33% wider picture onto your camera's CCD. There is no change in the vertical size but on the horizontal axis there is 33% more image. You get a much wider picture, and one that allows you to make 16:9 DVD's, or up-rez to HD or transfer to film with higher quality.

What's the difference between the anamorphic adapter and using "letterbox" mode?

Both are ways to get an image that looks like 16:9. Letterbox Mode shoots 4:3 video and simply puts black bars on the top and bottom of your image. If you're playing back on a 4:3 display, Letterbo~ mod~ will give you letterboxed video. Also, because much of the picture IS covered with black bars, you end up with only about 372 lines of available resolution (in NTSC mode).

The anamorphic adapter actually changes the shape of the video, making it appropriate for widescreen televisions. If you play back

anamorphic footage on a 16:9 widescreen television it'llloo~. .

"proper". If you play back anamorphic footage on a 4:3 television, It will look "squeezed". The anamorphic adapter also uses the full surface of the camera's CCD, so you get widescreen video with a full 480 lines of vertical resolution (576 lines on PAL).

How does it attach to the camera?

The anamorphic adapter has industry-standard 72mm lens threads on it. You screw it into the front of your lens just like any filter. It can also attach to any other 72mm-lens-thread camera.

152

Once the adapter is on your camera, you have to "square it" -- make sure that it is perfectly in square with the CCD of your camera, so that the lightbending effect is working on the horizontal axis only. The way you do this is by using the anamorphic adapter's white lens cap. The lens cap is translucent and has a grooved square outline on it. Point the camera at a light source and zoom out to ZOO, and you'll be able to see the grooved square (see the picture at right for what this looks like in the viewfinder/LCD). You use this square to line up the adapter. Loosen the leveling screw on the side of the adapter, then rotate it until the square of the lens cap is perfectly squared-up in your viewfinder. Then tighten the screw.

(note: you can also use the adapter for some wacky effects, if you purposefully rotate it during shooting --loosen that screw, and rotate that lens around, and you'll get some wild distort-o-vision effects on your video).

It's easier to attach the adapter if you hold the camera pointing straight up, so the adapter sits level on the surface of the camera mounting threads. It's easier to position it and it doesn't fall off while you're manipulating it into place.

How do I remove the anamorphic adapter? It just spins but won't come off!

This is a frequently asked question. The answer is, tighten the leveling screw even more before trying to unscrew the lens. If the screw is not tigbt enough, the lens will behave as if you're leveling it instead of trying to remove it. Tighten the screw and it'll engage the lens threads, allowing you to unscrew and remove the adapter. DO NOT OVERTIGHTEN THE SCREW, or you run the risk of stripping the screw threads.

153

Should I use "Letterbox" mode when using the anamorphic adapter?

No, normally you'd use full-screen mode with the anamorphic adapter, which will result in 16:9 widescreen video. But if you also use letterbox mode at the same time, the result will be a 2.35: 1 widescreen image -- the same aspect ratio as widescreen movies.

Should I use "Squeeze Mode" when using the anamorphic adapter?

Definitely not. Using the two together would result in a "double 16:9" effect, squeezing the image twice. You would end up with an aspect ratio of approximately 2.35: 1, like CinemaScope™, but there would be no proper way to display it. There is no such thing as a 2.35: 1 television. So to display it on a television, you'd have to unsqueeze the image, and what's the point of squeezing it just to have to unsqueeze it to look at it? It's much better to shoot anamorphic + Letterbox Mode, if you're going for a 2.35:1 look Anamorphic + Letterbox will give you the proper 2.35: 1 aspect ratio while also being able to be made into a DVD and played back properly on widescreen televisions. If going for a film transfer you also wouldn't want to shoot in the Squeeze Mode (see Footnote 1); you'll get better results from Letterbox.

154

FOOTNOTES

Footnote 1: For a film blowup or an up-rez to High Definition resolution, Letterbox Mode will produce superior results as compared to Squeeze Mode. The reason is because for both film transfer and for up-rezzing to HD, a computer algorithm will stretch the 720x372- pixel image into a much larger size, such as l280x720, 1920x1080, or more. Letterbox Mode and Squeeze Mode both start with approximately the same 720x372-pixel image, but Squeeze Mode does a small digital stretch to convert the 720x372 image into 720x480, whereas Letterbox Mode leaves the 720x372 pixels alone and adds black bars over the rest. So Letterbox Mode's central original image is untouched, unmodified, at the full resolution that the camera can deliver, whereas Squeeze Mode's image is now digitally stretched and modified. There will need to be much more stretching to get the image to fill an HD or film frame. Letterbox Mode provides a cleaner "starting point", whereas Squeeze Mode has already introduced artifacts and blending into the footage by performing its intermediary stretch. You will get a cleaner result by starting from the original, unmodified footage (i.e., Letterbox Mode) than you would from the partially-stretched Squeeze Mode. Of course, the best results would come from the optical anamorphic adapter, which gives you 33% more information in the source frame as compared to either Squeeze or Letterbox Mode.

Footnote 2: The DVXlOO user manual recommends that when you're using the camera on a tripod, you should tum the optical image stabilizing (OIS) system off Because a tripod is already an image stabilizing device, there's no need to have a second one (the OIS) active as well, and the two can actually "fight" each other, resulting in jittery footage. For example, think of a simple pan being executed on a tripod: the camera starts moving to the side, but the OIS sees the motion and thinks it's "shake" that it has to compensate for, so it starts using its gyro's to reposition the prism in the lens to keep the image stable. Well, you didn't want the image stabilized, you were trying to pan, right? But the image will stay stabilized while you continue to pan, and at some point the OIS will exceed its ability to compensate (it can only move the prism so far before it runs out of room) so it will "give up", and the result is that your nice smooth

155

Of course, the smaller the aperture is (i.e., the bigger the number is) the more light that will be necessary in order to get adequate exposure. It may be tempting to just open up the aperture some to get enough light, but doing so may result in out-of-focus pictures. The camera needs the smaller apertures to get adequate sharpness when using the anamorphic adapter. Always go by the aperture chart and make sure that the aperture you use is as small ( or smaller) than the one recommended on the chart.

tripod pan will encounter a large "bump" in the motion. So when you're on a tripod, you should tum the OIS off. Since most anamorphic adapter shots should be executed from a tripod, the loss of OIS is really a fairly minor issue.

Footnote 3: Aperture numbers that are smaller are considered more "open". For example, fl8 is more open than fill, and fl4 is more open than both of them. F/2.8 is the most wide-open you can get with a DVXlOO at full telephoto, and fll.6 is the most wide-open the DVXlOO can get at full wide angle. The more open the aperture is, the shallower the depth of field. You have to ensure that your aperture is closed down enough to provide adequate depth of field to make sure that the anamorphic picture is in focus. Adequate depth of field is crucial to getting sharp-focus pictures with the adapter.

156

ABOUT THE AUTHORS

Barry W. Green is an Emmy®-award-winning producer with four Emmy nominations for writing and producing television commercials and public service announcements. His technical background includes 13 years as a professional computer programmer and producer for Westwood Studios, creating some of the most popular video games in history. Since leaving the video game industry in 1999, he now writes and produces award-winning corporate and industrial films, commercials, screenplays and films for Fiercely Independent Films Inc., a Las Vegas-based production company. He also serves as a moderator for DVXUser.com, one of the world's largest online communities for users ofthe DVXI00 and DVXI00A cameras.

Jan Crittenden Livingston has been with Panasonic for almost 21 years. She started with the company as a salesperson in the Washington, DC area and then progressed to a Group Manager for Training. As the opportunity opened up, she moved into product management where she currently serves as the Product Line Business Manager for DVCPRO, DVCPR050 and the AG-DVXlOOB.

Harry W. Foulds is Panasonic's National Training Manager. He has been with Panasonic for over thirty years. Having served in a number of capacities in his early career with Panasonic as bench technician and salesperson, he progressed through the ranks into the field of training where he has been for nearly 20 years.

157