3DS Max - Tutoriais de Arquitetura

Welcome to CGarchitect.
com
Modeling a Pillow
Tutorial by: Harjeet Singh (info@virses.com) http://www.virses.com The aim of this tutorial is to model a pillow. This model will be created with simple and easy steps. At the end of this tutorial you will be able to create a pillow as in the image below. In fact, this image is rendered from this tutorial max files only. The max file for this tutorial is available on request.
1. Draw a BOX. Notice the multiple segments. The number of segments can vary as per requirement. These settings were optimum for this tutorial.
http://www.cgarchitect.com/resources/tutorials/misc/tutorial1.asp (1 of 17)29/10/2007 21:13:59
Welcome to CGarchitect.com
2. Goto modifier panel, expand the Modifier list and select "RELAX". Set the parameter values as in the image below. This will smooth the box edges.
3. Goto modifier list and select FFD(box) modifier. Set the number of control points. These values can be chosen differently for different level of control over the resulting mesh.
4. Switch to front view. Select in modifier stack the FFD modifier. Expand it to select "Control Points" sub object. Select the Top right control points (see a in image) by making a selection window. All the control points on this edge would be selected. Hold down CONTROL key and select control points at Top-Left as in (b). You now have both edge points selected.
5. Select "MOVE" from tool bar and move the selected control points down. Bring them near to mid-height of the box.
6. Now select control points at Lower Left and Lower Right corners by making selection window. Make sure that you do not select any control point in mod. Notice the arrow marks A and B in the image below. Then move the control points a bit toward end as marked C and D.
7. Switch to left view and repeat the step 5 and 6.
8. The model should look like the image below. You may consider moving other control points up/down to avoid and bulge if visible. The base pillow is done.
9. We will now add a frill or strip to the pillow to make it look more real. Goto command panel-> shapes -> SECTION . Pick and create a section plane over the pillow.Make sure to place in exactly at the mid-height (pivot point) of the pillow. You can use ALIGN tool on the tool bar to do so. Set Section Boundary to ON (see a) and resize the section to cover the pillow entirely. Now hit the button "Create Shape" (see b). This will create a spline showing exact profile of the pillow at center.
10. Select and delete the SECTION object. Select Pillow, right click and choose "Hide selected". You now can see the profile spline created by the Section object. We need to create an offset of this spline.
11. Select spline goto modifier stack and select Editable Spline -> Spline and select the profile spline again. This will change it to RED. Scroll down the parameter tab on the modifier panel to locate the "OUTLINE" parameter. Enter a value of -10 and hit ENTER. This will add another spline to the profile parallel to the original one.
12 Now select in the modifier stack, the Editable Spline-> Vertex. Clean up some vertexes near end on the outer spline and pull some vertexes to create a sharp cornier like in the image. Do same for all the ends. You must not move or delete ANY vertex on the inner spline.
13. The spline should look something like this now:
14. Select the spline, right click and choose Convert To: -> Convert to Editable Patch. This will create a patch mesh between two splines in the profile.
15. Select Editable Patch-> Vertex. Select some random vertex on outer edge only and move them up or down to get the wavy look.
16. Un-Select the patch, right click and choose "Unhide All:". The base pillow will be visible like shown. Add some nice fabric texture and you have the pillow ready to add more reality to your interior renders.
Modeling a Barcelona Chair

by Montree T. (easyyong@hotmail.com) - Smoke3dStudio
1. Create a Plane or Quad Patch.
2. Assign an edit patch modifier or convert to an editable patch and adjust the bezier
http://www.cgarchitect.com/resources/tutorials/smoke3d/tutorial7.asp (1 of 5)29/10/2007 21:04:43
handles (green dots) as shown. Also change the View Step parameter to be 1 or 2 to reduce the meshing within the patch unit.
3. While still in edit patch mode, go to the patch sub-object and clone the patch, using snapping.
4. Differentiate the number of smoothing groups in each patch unit as to avoid the same smoothing group in whole object after we've welded the vertices in the next step.
5. Turn to poly or convert this patch to editable poly.
6. Select all vertices and weld. This is to making all of the elements become one element.
7. Make the cushion's thickness. In polygon sub-object mode, select all polygons and extrude.
8. Clone the polygons on the top and move down and snap (using 3d snap) with the bottom part's edges. This is to make the bottom part of the chair capped. Flip the normals and select the vertices along the edges and weld.
9. Select the bottom part's polygons then go to a top view and align the selected polygons to the top view (or we can use "make planar" as well)
10. In the Edge sub-object mode, select the edges as shown in the picture and chamfer.
11. Select the chamfered polygons along the chair's edges and extrude or bevel.
12. The seat of the chair is complete as an element in the editable poly. Clone and rotate the seat to become a back rest.
13. Use an FFD box to add a curve to the chair.
Details in Materials
by Montree T. (easyyong@hotmail.com) - Smoke3dStudio I have shown step by step how to light an interior scene in my recent tutorials. This chapter will go into the details about materials to enhance them to create a more realistic scene. 1. Quickly flash back to the recent tutorial, step 1, create the ambient light of the area by placing omni lights (ambient only) in the middle of the room as the main ambient color. Make the far attenuation range fit to the room and place skylight ambient lights (light blue omni with ambient only) on the roof top. This will create the effect of the skylight scattered around the roof and walls. The scene below show the space bring lit up with the "AMBIENT ONLY" omni. The ambient pass is successfully done.
Step 1, create the whole ambience Tips: In this step, we may not put in all of the real materials into the scene first as some materials have reflections or whatever that make the rendering time longer . I assign a white color material to all the objects first. This is to make the entire scene lighting easier to control, as we can see the brightest and darkest part of all the object clearer and more accurate with white color.
Step 2, diffuse and specular pass created by adding: 1. Free direct (diffuse and specular) to cast indirect illumination shadows (soft shadow) 2. Skylight by placing blue omni (diffuse and specular) use cast shadow. We can replace the white color with the real materials in this stage to correct all detail of specular and diffuse color and levels.
Step 2, diffuse and specular pass created Step 3, The brightest part will be added in this stage. Sunlight and indirect illumination of sunlight from the floor and walls. See how and where I place these lights in the scene below.
Step 3, Brightest part to be added in this stage 2. Details in mapping and materials. Sometimes if we re-use materials in our library they don't work well in another scene. It doesn't go well or match with the current lighting ie. no specular or reflection is too bright. Making the reallistic material in a big scene or a scene that is lit up with many
lights is not as easy as a still light image that require less number of light. I suggest that we should go into the material details after we've finished lighting up the scene. The concept is to fake the material's physical condition to make it more realistic. In this tutorial I will mentionded a few examples so that you can be apply it to your own files. Step 4, Caustics. I found that the sunlight illumination from the floor to the environment is not reallistic enough. I placed 2 omnis at the small wavy water and applied a projector map into it, using cellular maps. The cellular map on a projector give the effect nearest to real caustics.
Step 4, Caustics
Cellular map as a projector map Step 5. One day I walked in to a huge building lobby with an all granite floor. I noticed that the granite floor had an ood reflection even though it looked like a smooth plain floor. The reflection looks distorted and scattered into pieces and doesn't join together. I went back home and tried to fake this effect but no such option existed in the raytrace material (or I don't know about it). I solved the problem by drawing the granite piece by piece and joining them together, but I found I was wasting so much time and the edge of each granite did not join correctly with our eye's estimation. 2.1 Distortion effect. In this scene, I drew a plain as another layer of the foreground floor with 8x8 segment, assign marble material to it and a UVW map. I also added a noise modifier to it. This is to give every piece of marble some texture so that it is not too flat. When rendered, we will see the distorted reflection on the marble continually in every piece of marble because the whole polygon on plain is in the same smoothgroup. The distortion appeared on the marble because of the noise modifier. I added editmesh modifier to give each polygon a different smoothing group numbers. I then extruded some rectangle polygons randomly, and very slightly and rendered again. The distortion of the reflection now seems independent just like in real life... and that's the effect I want.
Normal reflection of flat marble floor
Distorted reflection appears continuous on the marble floor as the whole polygon is in the same smoothing group number.
picture. Distortation appears independently piece by piece as from the different smoothing grou 2.2 Faking physical conditions of materials to match the lighting in the scene. At the begining of this drawing, I used the material from my library but when I made the final rendering, the specular level appeared acceptable but didn't have enough details to make the scene more realistic. Actually the specular level of the original material is adjusted correctly. I faked the rendering again by increasing the specular level to make the apperance of scratches (I used scratch map for specular level map) appear more obvious. Not only was the specular wrong, but the diffuse level or bump map parameter may be changed to be more obvious in other scene. Remember that material follows light whenever there are more lights in the scene but light can follow material when there are less lights in the scene.
Normal material in library
Increasing the specular level to enhance the realism and to match the light.
Final rendering
Material Tips and Tricks Part 3

by Montree T. (easyyong@hotmail.com) - Smoke3dStudio 3. Tips for reflection and refraction maps 3.1 Perpendicular and parallel The anisotropic effect produces different kinds of reflection. The reflection will change more or less depending on the viweing angle. The least reflection will appear on a suface that is perpendicular to the camera and the most reflection appear on surface that is parallel to the camera. This effect is called the fresnal effect. 1.You can see this effect on most of the anisotropic objects ie. damp/wet road, water surface on swimming pool, polished leather, some kind of metals and etc.
The reflection is changed in every angle. The bird eyes angle has less reflection and more reflection at the worm eyes view (where the camera angle is almost parallel to the wet floor) 2. Curve/freeform object ie. Cylinder, and sphere and etc.
The least reflection is at the middle of the cylinder which surface is the area perpendicular to the camera and the more reflection is on the suface which is parallel to the camera.
You can create this effect by using a fall off map (perpendicular/parallel or fresnal) over raytrace. Both perpendicular/parallel and fresnal produce almost similar effects, but I found perpendicular/parallel to be more apparent. 3.2 Shadow and Light Falloff "shadow and light" seems far different from the real world reflection. I am not sure if there are any theories about this, but I just use it to accerelate the reflection at the light area and dim down the reflection in the shaodw areas. This will create "subdued mood" in your scene. Imagine the reflection on a black granite floor in a pub. The bright area will be blow the light and will then disappear into the shadows.
Shows the parameters and settings of Shadow and light.
A rendering with Shadow and light falling off. The reflection give a subdued mood. 3.3. Some examples of complex scene refractions This is quite off-topic in that doesn't lend itself to our architecural tutorials, but we basically need to know how it works so that we can apply it on our own . 3.3.1 Create water drops on a glass
I will skip mentioning about details on lighting as this is a simple scene. Let's just look at the wire frame only. After a few time tests on the materials, I decided to use raytrace material as the effect is much easier to create than standard material. Also, I found the splash map is the more suitable than cellular as the effect appears clearer. I used blend material for the glass model which contains 2 materials ie. galss material and water material and a splash map for masking.
The rendering time was about 32 minutes and it quite long for a dual processor.
Glass material legend
1. Glass material. It seemed to be very easy as I didn't even put any map on it. Just change the color of diffuse, transparency and reflection and change the IOR to be 1.5.
2. Water drop material. I just copied the same mateial from the first slot and changed the IOR to be 1.333 for the water, and copyied the same masking map (slash map from slot#3) and placed it on spec.level and Bump.
3. Masking map : splash
This is the floor materail which uses a similar technique to the blend material.
The image above is very close to the 1st one but uses only material#2 which is the water. There are bump and specualar effects except that it has only 1 IOR that is set to 1.33. If you don't mind about the IOR. The rendering time is quite fast at 3.32 minutes. 3.2 Diamond Someone had asked me how to draw a diamond. Rendering a diamond is not as difficult as what you might think. Just use raytrace material and chnge the transparency to white, IOR=2.417 and facet, that's all. The diamond model and its surface is also important that you need to find a blue print to see how it looks before you draw.
This is roughly how the edge of the diamond looks like.
Put seveal color of lights around the diamond

by Montree T. (easyyong@hotmail.com) - Smoke3dStudio 2. Self illumination map
Just imagine how the lampshade's illumination looks like first. A self illumination map can be easily done with many layers of gradient ramp.
1. Black and white cross hatch bitmap put on mix map and change the color for the lampshade's diffuse color.
2. Gradient ramp map tree for self illumination maps.
2.2. Fluorescent diffuser (light box)
We can create this map by using two layers of gradient ramp and putting in selfillumination maps.
2.3 Fake translucency using falloff map on self illumination map
From Max help file: "Translucent shading is similar to Blinn shading, but it also lets you specify translucency. A translucent object allows light to pass through, and also scatters light within the object. You can use translucency to simulate frosted and etched glass."
Translucent shader
There is an unwanted shadow appearing on the sphere which is the translucent shader or translucent color in the raytrace material. I have not yet found a solution to solve this problem.
Fake Translucent, using falloff map (shadow and light) on self-illumination map in Blinn shader.
Though the effect of this falloff map is not as good as the real translucent shader, the effect is still too closed and without the weird shadow appearing on the object. An example of faking translucent, using falloff.
1. Diffuse
*2. Self illumination
3. Opacity
4. Reflection
5. Refraction
The beer bottles in the image are just small items compared with the whole scene, so it's non-sense if we put the material above for these as the rendering time are surely super long. So, for these bottles, it's enough to put only the self illumination and opacity map on them. 2.4 Vertex illumination Vertex color illumination color overlay on base map. The vertex illumination uses a similar technique as the vertex color on diffuse but instead, we put this map on the self illumination map. The default color of vertex is white. So if we use the vertex color map for self illumination, we first need to change all the vertex colors to be black.

by Montree T. (easyyong@hotmail.com) - Smoke3dStudio This 3 part tutorial will explain how to add more details to material easily while saving time, especially by adding another layer of 2D and 3D maps which are available in the program. The first thing before we start: If you don't know how to create good detail on your material, just use the same bitmap on Diffuse, Specular map, Specular color map etc. 1.Adding details on Diffuse 1.1. Vertex color map Vertex color maps help to add a layer of color overlaying the base map. We need no UVWmap for vertex color as the overlay color will change by adding a color on the vertex.
A simple bitmap for the diffuse map on a wood material. Some lighting techniques make objects in a scene appear very flat when rendered. The rendering of this wood cupboard looks flat.
Copy the wood material to another slot, adding mix map on the diffuse map. Put the same wood material in the middle(color#2) and put vertex color map on top (color#1) and bottom (mix amount).
Select Polygons and assign the new material to them.
Go to vertex subobject, select vertices and change the default color of the vertex (that is white) to be dark brown.
The revised rendering which the cupboad looks more 3 dimensional. 1.2. Changing wood's color by using mix map You can combine two colors or materials on a single side of the surface with a mix map. We only have one single wood bitmap but we can produce many types of wood. This may save you time rather than having to scan a new one.
We just have a black and white map of wood and overwrite a mix map on it so you can change the color freely.
1.3. Falloff map on diffuse From Max help file: "The Falloff map generates a value from white to black, based on the angular falloff of the face normals on the surface of the geometry. The direction used to specify the angular falloff varies, depending on the methods you choose. However, with the default settings, the map generates white on faces whose normals point outward from the current view, and black on faces whose normals are parallel to the current view." This technique is suitable for curve/wavy object.
The shaded area of the white curve wall, which is lit with ambient only light, seems to
be flat.
Solving the problem by using falloff map on diffuse map.
The rendering look more 3 dimensional now. 1.4. Overlay brick map on a simple marble (in max 6, called tiling map)
Some scaned bitmap, used as a diffuse looks like tiling/unseamless or so plain. Put the same bitmap on a brick map to differentiate the tone of the marble's tiling.
Editable poly for Interiors Part 2

by Montree T. (easyyong@hotmail.com) - Smoke3dStudio Drawing a large scale interior 1. Prepare a multi-subobject material for the building. ie. walls, windows, floors, glass, ceiling etc.
2. Import the building Autocad drawing, group and freeze it.
3. Start by tracing a line following the building contour.
4. Work with the segment subobject and divide/refine each line segment to reflect the number of windows on it.
5. Extrude spline and make the segment = 3. The polygons in the middle will be the building's windows.
6. Convert to poly, flip normals and create a camera inside. Assign the material ID onto the walls, ceiling and floor first.
Flip normal
Camera view
7. Select the polygons for window and deselect any of the walls. Assign material ID of the selected polygons in the middle = 4 ( from the material ID 4 = window frame) and adjust their height, following the window's size and then inset them all, using by polygon type. This is to create the width of all windows.
Select polygons in the middle and assign Material ID
Inset to create the width of every window frame. 8. Create frame's thickness by extruding the polygons and change the ID to 5, that is glass.
9. Select polygons and apply UVWmap on ceiling, floor etc.
10. Create columns and building core using the same polygon technique.
11. Place some lights into the scene.
Now you have prepared a scene that is ready to add in the furniture and design and runing a walkthrough. I am quite sure if you are familier with this technique. You shouldn't spend more 2 hours on the modeling and lighting.
Editable poly for Interiors Part 1

by Montree T. (easyyong@hotmail.com) - Smoke3dStudio For this tutorial we need to know how the multi-subobject and polygon ID works. 1. Prepare a multi-subobject material before drawing. Material ID 1 = wall Material ID 2 = ceiling Material ID 3 = floor . . . and etc.
2. Draw a box, convert to poly and flip all polygon's normals.
3. Create a camera inside and change a viewport to be the camera view. Assign material IDs to each polygon by following the materials that we have already prepared. In this stage I put ID 1 for whole wall, ID 2 for the ceiling and ID 3 for the floor. The important thing is to assign material IDs to polygons in the beginning rather than later, as the finished model might be more complex and it could be more difficult to select polygons.
4. work on edge sub-object, select edge and connect. This is for creating window and door
Tip: Avoid using slice tool in editable poly/mesh by using connect edge instead, as incorrect uvwmap coordination may sometimes appear on the sliced polygon when we apply face UVWmap to it. 5. Connect edge again. 2 segments divided horizontally for a window in the left and 1 segment divided for a door in the right.
6. Adjust the selected edges height to the same as window and door.
7. Go to polygon sub-object and extrude the selected polygons to create wall thickness and then assign Material ID =4 (Material ID 4 = window frame) to the extruded polygons.
8. Connect the selected edges
9. Select the polygons and inset. This is to create the width of window and door frame.
10. Go to the door polygon and move it down as we don't need the door step.
11. Extrude polygons to create the thickness of door and window. Then assign number ID 5 for the extruded polgons window and number 6 for polygon door.
12. Select the polygons and delete the overlaping polygons at the door step that we don't need. Go to the Border sub-object and cap the hole again. A new clean polygon will be created at the foot of the door.
13. The model has been finished and we can now check the material IDs on each polygon.
14. Select each polygon and apply UVWmap.
15. The finished scene lit with lighting techniques from the first 3 tutorials.
Creating a Stainless Steel Material

This tutorial was written from my experience working on metal materials personal obsevations and from photographs. There are 3 important things to remember when you create a stainless steel material. 1. Strong Highlight Although the metal shader is the easiest way to create a stainless steel material, it has limitation in creating the highlight, so I prefer to use either multilayer or anisotropic, as we can play with its several highlight shapes.
2 layers of specular level on multi layer shader
The different between the highlight shape of the blinn shader and the anisotropic shader Besides the really strong highlight on the stainless steel, that we can create by changing the basic specular level and glossiness parameter, I found that there is another strong illumination of light caused by the reflection on the stainless steel surface (the highlight appear inside the reflection) that also looks like another highlight. As this illumination doesn't have an exact shape, depending on the light in the environment of the reflection. i.e. Some appear like thin vertical lines and we need to use anisotropy to create such effect.
Faking the highlight in the environment reflection using anisotropic Use First specular layer that we have in multilayer shader as a real highlight, so we can adjust only the specular level to be higher and lower and use glossiness to make the highlight look more obvious.
The First specular Level is the real specular and the Second specular Level is the one that appear inside the reflection. I also make the Second specular level to be as strong as the First specular layer, but this time I adjust the anisotropy parameter to be higher so the highlight shape is thinner. The orientation will be changed as well if the effect of the anisotropy doesn't appear correct. ( The orientation is a value in degrees that can range from -9,999 to 9,999. Default=0.)
The different shapes of specular highlights caused by changing the orientations. Put a scratch or brushed metal map in the specualar map slot for more detail in the material highlights.
2. Low Diffuse level The nature of a metal is that it has strong specular highlight and reflection but it has a diffuse level scale (in 3ds max) lower than a normal shader (the brightness of the difuse color is less than normal). If we are creating the stainleess steel with an anisotropic shader, we have to make the anisotropy parameter lower than default. I normally use is around 25 - 70.
3. Motion Blur Reflection The reflection on stainless steel is very refractive and blurry (for example it looks like a motion blur effect in photoshop). The best reflection map in 3dsmax is to raytrace, but I found no such effect in raytrace that is similar to the effect seen on stainless steel. Though we can use "multi resolution adaptive antialiaser blur", and the effect is similar to using a motion blur, it is not quite the same and it takes a long time to render. Put a motion blur filter on the original environment map. (Note: This effect does not work well on flat surfaces.)
An alternate method is to skip almost all the detail in environment map and only recreate the strong illumination of the reflection highlight by creating a new reflection map with a linear gradient ramp. As the most obvious reflections on a stainless steel surface are the lines, this gradient ramp map should use the "explicit map channel".
Spherical environment gradient ramp Tthe reflection still does not look realistic as it appears in every part of the surface equally, so we use a falloff map to specify the fading of the reflection area.
Apply a falloff map over a reflection map.
This picture shows the reflection falloff when using falloff in the raytrace material.
The reflection is masked by falloff map. The fading distance is specifed by the falloff type. In the picture I used "perpendicular and parallel" No reflection map produces realistic effect on a flat surface besides Raytrace. I sometimes use gradient ramp maps on anisotropy to fake the effect on flat surface. As the anisotropy appears in the specular highlight only, the gradient ramp will appear on the specular area only and fading into the diffuse and ambient area.
Anisotropic map
The anisotropy that appears on the specular highlight looks like a reflection.
kitchen sink studios

8 1 5 N . P h o e n i 6 0 2 . 2 5 w w w . k i
llc 1 s t A v e , S u i t e 5 x , A Z 8 5 0 0 3 8 . 3 1 5 0 t c h e n s i n k s t u d i o s . c o m
F A L L I N G
W A T E R
t u t o r i a l
# 2
CHALLENGE
To create a convincing waterfall within 3DStudio MAX.
PARTICLE FALL
by Mike Heuschele
So we need to make a waterfall in the circled area. No problem. Let's use the Snow particle system and a reflective material behind the snow. Create plane objects to recieve the water material where the water needs to be. After that, we need to create the water material. Start by creating a noise map in the diffuse channel of an unused standard material. Set all the parameters to match what is shown to the right. Once you have your noise parameters adjusted, copy the map from the diffuse channel to the bump channel. Be sure to make a copy ond not an instance so you may edit the parameters independently. Under the bump map noise parameters, set the #1 color to white and the #2 color to black. Once completed, copy the bump map to the specular channel. In the reflection channel of your water material create a flat mirror map. Be sure that Apply to Faces with ID:1 is checked or else the relection will not work. Apply the material and do a test render.
SOLUTION
Use a particle emitter in conjunction with a reflective procedural texture to generate the effect.
k i t c h e n
s i n k
s t u d i o s
l l c

8 1 5 N . P h o e n i 6 0 2 . 2 5 w w w . k i
F A L L I N G
W A T E R
t u t o r i a l
# 2
Now lets make the particle systems. Create a Snow particle emitter and line it up with the edge of the waterfall. Also create a gravity space warp with a strength of 15. Bind the particle system to the gravity space warp. Next create a particle deflector at the base of the waterfall. Bind the particle system to the deflector. Now set the deflector's bouce parameter to approximately .14. Copy your particle system to the other edge of the waterfall. Also place a third copy at a 45 degree angle over the corner of the two adjoining planes. This will soften up that 90 degree angle. Ok, let's create a material for the particle systems. First make sure all of your particle systems are set to render Facing the Camera. It's just a little checkbox in the parameters roll-out. Next, create a noise map in the opacity channel of any available material. Be sure the face map box is checked at the top of the Standard material panel. Set the noise parameters to match the image to the top left. Apply the material to all three particle systems. To make the effect more convincing, we need to add motion blur to the particle systems. Right click and choose properties on each of the emitters. Enable the motion blur at the bottom left and close the window.
k i t c h e n
s i n k
s t u d i o s
l l c

8 1 5 N . P h o e n i 6 0 2 . 2 5 w w w . k i
F A L L I N G
W A T E R
t u t o r i a l
# 2
Do a test render and you'll see that you have a fairly decent waterfall effect.
Questions or comments on this tutorial? Email Mike at
k i t c h e n
s i n k
s t u d i o s
l l c
Lighting and Shadow Basics - Part 1

Ambient Light Ambient light is one of the most important parts of creating a realistic rendering becuase it sets the time and mood of picture that you are trying to create. The complete scene should be composed of a specular pass, a diffuse pass and an ambient pass. Some artists use only the default specular and diffuse settings, but I always like to use ambient light only as a main ambient light.
The examples, below are done without the ambient only light seeting checked.
3D images by Layer3d@hotmail.com, www.thai3dviz.com
Compare with the scene, done with ambient only lights (below)
3D images by Narin Assawapornchai, gio_rin@hotmail.com(top) and Peangporn Boonmema, peangporn@yahoo.com (bottom)
So what color should we use for the ambient light in the scene? The color of the ambient light should be the average color of light affecting the scene + the average color of the entire room. For example, skylight will influence a room with a large window or opening the most as the skylight (a kind of area light) will scatter around a large portion of the room or even throughout the entire room. So if the wall is a white or cream color, the ambient color should be blue(skylight) + white or cream which results in a yellow-green color. Also, don't forget to dim down the intensity of the color. The sunlight itself will only affect the room in some parts, that is only the floor and walls as the sun is a directional light and it will not scatter around like the skylight, with exception to its indirect illumination from floor or wall. However, this indirect illumination will not have as much of an effect as the skylight.
If you still don't know what color you should use. I suggest that you go with a browngrey first. I don't use pure grey as its lack of color will make the image look dead. I have some suggestions about color below, but keep in mind these colors are not set in stone. You may use this as a guideline. Example 1: The color of a room with a small opening
Ambient only R= 142 G=139 B=116 Example 2: The color of a room with a bigger openning
Ambient only R= 168 G=175 B=159
If we have any downlights or diffusers in the room, the skylight and sunlight are still considered stronger. If I give skylight multiplier = 1, the multipier of downlighs or any diffusers should be less. I always start creating lights whose multiplier is the strongest first.
The 2nd picture shows that the multiplier of skylight is equal or lesser than downlights. Example 3: The color of atmosphere of a room without natural light. Suppose the color or these down light is yellow.
Ambient only R = 163 G =128 B = 23 Example 4: The color of atmosphere of a room without natural light as well, but the light sources are more diffuse.
Ambient only R = 138 G =143 B = 117 Another interesting way to understand how to use color to make our renderings look better and more colorful is to understand color theory. I have read a number of books about painting written by artists. They always try to avoid black and grey becuase this causes images to look dead and colorless. Complimentary color is also interesting. An artist (I am sorry that I don't remember his name) says "Cool light(color) warm shadow(color) and warm light cool shadow." I am not too sure if this conflicts with
our present science theory but as long as your work turns out very well and you call it "Art" , nothing is right or wrong. For example, if we see orange and blue (complemantary colors), you may think about evening time, as the color of sunlight appear orange in the evening and the ambient light is blue. Another example that is yellow and purple, yellow light goes well with purple ambient light.
Example 5: The scene is filled with blue and orange light

Area Lights and Area Shadows Another thing that a new user should not ignore is the beauty of area light. Area lights cause specular scattering on objects and blend with its diffuse area, creating a very soft look with beautiful soft shadows. The bigger the light sources is, the softer its shadow will be.
Indirect light is bounced light that reflects from an object when it is affected by direct light. The color of that light is affected by the color of the object. I also consider indirect light as a kind of arealight as well as the light is scattered from the object surface. The area light is not that difficult to fake, but its soft shadow is what we need to pay more attention to. In Autodesk Viz 4, we have a new shadow type called the "area shadow" but the problem is that it takes a really long time to render. Some use many lights to fake this effect, (ie. skydome light) but that is for a still light or exterior rendering. For interior scenes, we can not use this method as there are too many objects in the scene to consider.
Figure 1: Area light, divided into small units and the units are replaced with free spots
Figure 2: A shadow, casted by 1 single point light
Figure3: Faking arealight effect, using spotlights
Figure 4: The shadow before changing size and sample range
Figure 5: After adjusting size and sample range, The shadows looks smoother There is another technique I have found using a negative parameter for shadow density to fake an area shadow. Sometimes if we use a lot of bright ambient-only lights in a scene, the default shadows may look faded or pale. The negative shadow value also helps to make the shadows darker.
Figure 6: Shows the black color shadow with the default density parameter of 1 or more. You will see in the middle of the shadow area is brighter than the edge of the shadow.
Figure 7: Shows the shadow with a negative density (-1,-2,-3). Due to the negative value, you also need to change the color of shadow to be whiter or light grey to inverse. The middle of the shadow appears darker than the edge. The result is a much softer look.

Step by Step Let's start with a question: How many kinds of light affect the scene that we are going to create and which one is the affecting scene the most? In the last teo parts we leanred there are skylight from openings, Sunlight and indirect light. The skylight may have the most effect on the entire atmosphere color if the openings are big or it may be the color of the objects in the room. 1. Creating ambient Once I know the color of atmosphere, I create an omni light as the main ambient light in the scene and enable "ambient only". Place the omni in the middle of the room then adjust the attenuation range. The "start" of the far attenuation should be fit to the room and the "end" of the attenuation is out of the room as shown in picture. I use uniform scale to scale down the Omni to fit to the height of the room so that the ratio between start and end should be around 1:2 or 3:5.
2. Creating soft shadows from area lights and indirect light I usually make 2 layers of shadow for my interior scenes. The first layer is for the shape of the shadow from light source and the second layer is for soft shadows from indirect light and area light. The different look between soft and sharp shadows makes the scene look more fuzzy and realistic. Basically, I always have one free direct under the ceiling and make the multiplier as small as possible. A typical value might be 0.1-0.2 and the color as dark as possible. What I want is only the soft shadow of every object in the room on the floor. Make the shadow color light grey and change the shadow density to a negative number like
-1, -2 or -3 with a sample range as large as possible ie. 28-50. The negative shadow seems to be the easiet way for me to fake the area shadow. The advantage of the negative shadow is that it saves time during rendering. The negative shadow will appear darker or brighter depending on the color of the ambient light, so if the shadow appears very dark, you can reduce either the color or shadow parameters. Casting these types of shadows is not limited to the floor. Sometimes I also place this kind of direct light to cast the shadow onto the ceiling or even on the wall.
3. Add specular and diffuse pass from direct light 3.1. Creating skylight When the skylight come into the room from opening ie. windows or doors, I consider those openings that the sky light came through as area light sources. Put new omnis as blue skylight outside every opening and adjust the far attenuation ranges to fit in the opening area. You may use cast shadow from these lights and the shadow should be soft as well.
3.2. Sunlight The sunlight itseft is not an arealight as the sun is only a small light spot in the sky, compared with the sky. (The sky is considered as a big dome of an area light) The sunlight shadow has a sharp edge. Sunlight is also considered as a directional light source. Place a target directional light as a sun outside of the room, and set the shadow to a shadow map or raytrace and change the edge of the shadow to be sharp.
3.3 In case of we have any interior lights sources ie. Downlight or any diffuser, we also add them in at this stage. Don't forget that the bigger the light source, the softer the shadow will be. Note: Sometimes, I finish the rendering without the sunlight and this may make the atmosphere look cooler. With the sun, the interior might look hot and look like it was in the middle of the afternoon. 4. Indirect light and illumination Take a look around the whole scene again. I notice that there are 2-3 main indirect illuminations that we need to add. 4.1 Illumination from the big orange wall. I place an orange omni light behind the orange wall and make the attenuation as big as the wall size. Use ambient only. The ambient only light will also make the orange wall illuminte but if we put normal specular and diffuse light, it will illuminate only the rest of the wall, floor and ceiling. The advantage of ambient only is it will affect the opposite side of an object. 4.2 Illumination on the floor, affected by sunlight. When the sunlight touchs the floor, it will bounce and scatter to its surroundings. I put a color of blue (floor's color) + a bit of yellow tint omni (ambient only) under the floor and set the attenuation size to fit into the sunlit area. If you see the scene is too full of ambient only light, it will make the scene lack dimension. You may change the omni to diffuse and specular instead. This method is flexible, it depends on your sharp eyes.
In a bigger interior scene that is more complex and has more indirect illumination, we might not put indirect light on all the illumination. Just placing a few main indirect illumination lights is enough as we are not the machine that can calculate all the illumination correctly. The look of the scene is based primarily on your sharp eyes. I also compare light to a brush that we are playing with on our artwork.
Semi-Outdoor Lighting
by Montree T. (easyyong@hotmail.com) - Smoke3dStudio The semi-outdoor lighting uses a similar technique to the interior lighting in my previous tutorial. The semi-outdoor is slightly different in that the opening is bigger and divided by columns. 1. Creating a main ambient light I use 1 brown/grey omni, ambient only, placing it in the middle of the interior and the start of the far attenuation is fit to the room.
2. Creating sub-ambient light Put light blue omnis, ambient only, in the openings between the columns. The start of the far attenuation is also fit to the opening. The sub-ambient blue color is a skylight ambient which scattered around the area.
3. Creating specular and diffuse passes and soft shadow 3.1 Creating skylight Copy the skylight omnis and change ambient only to diffuse and specular. This is to create specular and diffuse pass of skylight that affect the scene.
The opening of a semi-outdoor scene is slightly different from an interior opening in that the opening is bigger and divided by columns. I put an omni as a skylight outside an opening/ window. This will also use same technique in that we consider an area between the columns as a big opening. Tip: Shadow is affected by playing with the attenuation 3.1.1 The darker area in the highlight area on the picture is the area between start and end of the far attenuation.
3.1.2 Place omnis as a skylight, the start of far attenuation is fit to the area in between the columns, so the light in the area beyond that will be darker.
The picture above is a rendering without a main ambient. Only the skylight affects the scene. 3.1.3 The area between the start and end attenuation of each omni in the scene when intersected, will still be darker and look like column shadows. So with this effect, we sometimes don't need to make all the skylight cast shadows.
3.2 creating soft shadow Place 2 direct lights up and down to ceiling and floor and make the shadow density negative and color to be grey. The hot spot and fall off of the lights will be fit to the building as well.
Now we can see the soft shadow of the chairs and another layer of soft shadow of columns. 4. Sunlight Create a target direct light with raytrace or advanced raytrace shadows set to cast sharp-edge shadow directly from sun.
5. Sunlight illumination
The scene with large openings has direct sun illumination . I place omnis down to the floor at the same position of the sunlight on the floor to create bounced light which illuminates the area. I also want the light to affect some parts of the timber ceiling so I make the omni's attenuation touch the ceiling.
EVALUATING LIGHTSCAPES ACCURACY FOR PREDICTING DAYLIGHTING ILLUMINANCE COMPARED TO AN ACTUAL SPACE.
By MOHAMED ADEL SAMY SHALABY
A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF INTERIOR DESIGN UNIVERSITY OF FLORIDA 2002
Copyright 2002 By MOHAMED ADEL SAMY SHALABY
Dedicated to my father Samy Shalaby, to my mother Samia Abdel Moneim, to my brothers Ahmed and Amin Shalaby, and last but not least my wife Rania. Thank you.
ACKNOWLEDGMENTS I acknowledge with gratitude the inspiration and guidance provided by professors Greg Hall and Marjua Torres, who gave me insight to work on this project. A grateful appreciation goes to professors Janine King and Martin Gold. I also express grateful thanks to professors Mary Jo Hasell and Helena Moussatche in providing me with the appropriate research tools and patiently guiding me in utilizing those tools in my research process.
iv
TABLE OF CONTENTS
Page ACKNOWLEDGMENTS ................................................................................................. iv LIST OF TABLES ........................................................................................................... viii LIST OF FIGURES .............................................................................................................x ABSTRACT......................................................................................................................xvi CHAPTER 1 INTRODUCTION ............................................................................................................1 1.1 Limitations of the Study............................................................................................ 4 1.2 Significance of the Study.......................................................................................... 5 2 REVIEW OF LITERATURE ...........................................................................................6 2.1 Background ............................................................................................................... 6 2.2 Daylighting Tools ..................................................................................................... 7 2.3 Computer Graphics ................................................................................................... 9 2.4 Case Studies ............................................................................................................ 11 2.4.1 Daylighting Decisions...................................................................................... 11 2.4.2 Software Comparisons ..................................................................................... 12 2.4.3 Atrium Study.................................................................................................... 13 2.4.4 Lightscapes Daylighting Accuracy Test......................................................... 13 2.5 Purpose of this Study .............................................................................................. 14 3 MATERIALS AND METHODS....................................................................................16 3.1 Pilot Study............................................................................................................... 17 3.1.1 Introduction...................................................................................................... 17 3.1.2 Methods............................................................................................................ 19 3.1.3 Results and Discussion..................................................................................... 25 3.1.3.1 Interior Space ............................................................................................ 26 3.1.3.2 Exterior-Interior Space.............................................................................. 28
3.1.4 Discussion ........................................................................................................ 30 3.1.5 Conclusion ....................................................................................................... 31 3.2 Actual Research ...................................................................................................... 32 3.2.1 Introduction...................................................................................................... 32 3.2.2 Methods............................................................................................................ 33 4 RESULTS AND DISCUSSION .....................................................................................37 4.1 Results ..................................................................................................................... 37 4.1.1 Interior Space ................................................................................................... 39 4.1.2 Exterior-Interior Space..................................................................................... 43 4.2 Discussion ............................................................................................................... 47 5 SUMMARY AND CONCLUSION............................................................................... 50 5.1 Conclusion .............................................................................................................. 50 5.2 Recommendations ................................................................................................... 52 5.2.1 Preparing The Model (.dwg):........................................................................... 52 5.2.2 Lighting Preparation and Solution files (.lp - .ls): ........................................... 52 5.3 Further Research ..................................................................................................... 53 APPENDIX A ACTUAL RESEARCH DATA......................................................................................54 A.1 Interior Space ......................................................................................................... 54 A.1.1 Tables and Graphs........................................................................................... 54 A.1.2 Statistical Analysis.......................................................................................... 58 A.2 Exterior-Interior Space........................................................................................... 59 A.2.1 Tables and Graphs........................................................................................... 59 A.2.2 Statistical Analysis.......................................................................................... 63 B PILOT STUDY DATA ..................................................................................................64 B.1 Interior Space ......................................................................................................... 64 B.1.1 Simple CGM ................................................................................................... 64 B.1.2 Moderate CGM................................................................................................ 67 B.1.3 Detailed CGM ................................................................................................. 70 B.1.4 Figures............................................................................................................. 73 B.2 Exterior-Interior Space........................................................................................... 75 B.2.1 Simple CGM ................................................................................................... 75 B.2.2 Moderate CGM................................................................................................ 78 B.2.3 Detailed CGM ................................................................................................. 81 B.2.4 Figures............................................................................................................. 84
vi
LIST OF REFERENCES ...................................................................................................86 BIOGRAPHICAL SKETCH .............................................................................................88
vii
LIST OF TABLES
Table
Page
3.1: The Information that was used for generating the Light simulation in Lightscape........25 3.2: The Total Average of all the Data collected from The Simple Model using the Interior Space Solution Compared to the Actual Space.......................................31 3.3: The Total Average of all the Data collected from all CGMs using the Interior Space Solution Compared to the Actual Space. .............................................................31 4.1: The Information that was used for generating the Light simulation in Lightscape........37 A.1: Daylight illuminance measurements recorded from the CGM using an Interior Space Solution, compared to the Actual Space measurements. .....................................54 A.1 Continued ........................................................................................................................55 A.2: The total average of measurements from the CGMs using an Interior solution at the three different time frames...................................................................................55 A.3: Results of the Regression Analysis................................................................................58 A.4: Daylight illuminance measurements recorded from the CGM using an ExteriorInterior Space Solution, compared to the Actual Space measurements...............59 A.4 Continued ........................................................................................................................60 A.5: The total average of measurements from the CGMs using an Exterior-Interior solution at the three different time frames. ..........................................................60 A.6: Results of the Regression Analysis................................................................................63 B.1: Daylight illuminance measurements recorded from the Simple CGM using an Interior Space Solution, compared to the Actual Space measurements...............64 B.2: The total average of measurements from the Simple CGM using an Interior solution at the three different time frames. ........................................................................64
viii
B.3: Daylight illuminance measurements recorded from the Moderate CGM using an Interior Space Solution, compared to the Actual Space measurements...............67 B.4: The total average of measurements from the Moderate CGM using an Interior solution at the three different time frames. ..........................................................67 B.5: Daylight illuminance measurements recorded from the Detailed CGM using an Interior Space Solution, compared to the Actual Space measurements...............70 B.6: The total average of measurements from the Detailed CGM using an Interior solution at the three different time frames. ..........................................................70 B.7: Daylight illuminance measurements recorded from the Simple CGM using an Exterior-Interior Space Solution, compared to the Actual Space measurements.......................................................................................................75 B.8: The total average of measurements from the Simple CGM using an Exterior-Interior solution at the three different time frames. ..........................................................75 B.9: Daylight illuminance measurements recorded from the Moderate CGM using an Exterior-Interior Space Solution, compared to the Actual Space measurements.......................................................................................................78 B.10: The total average of measurements from the Moderate CGM using an ExteriorInterior solution at the three different time frames. .............................................78 B.11: Daylight illuminance measurements recorded from the Detailed CGM using an Exterior-Interior Space Solution, compared to the Actual Space measurements.......................................................................................................81 B.12: The total average of measurements from the Detailed CGM using an ExteriorInterior solution at the three different time frames. .............................................81
ix
LIST OF FIGURES
Figure
Page
3.1: Research Parts.................................................................................................................16 3.2: Lightscapes Process Parameters. ...................................................................................18 3.3: Lightscapes Quality Wizard. .........................................................................................18 3.4: Lightscapes Daylight Parameter....................................................................................19 3.5: Outline of Study Procedures Actual Space..................................................................20 3.6: Outline of Study Procedures - CGM...............................................................................22 3.7: The Simple Model generated in CAD ............................................................................22 3.8: The Moderate Model generated in CAD ........................................................................22 3.9: The Detailed Model generated in CAD ..........................................................................23 3.10: The Simple Model in Lightscape..................................................................................23 3.11: The Moderate Model in Lightscape..............................................................................23 3.12: The Detailed Model in Lightscape................................................................................24 3.13: First Method Setting a Location ................................................................................24 3.14: Second Method Setting Sun Angles and Intensity.....................................................24 3.15: Plan of the Classroom showing the Measurement Locations in the Space...................26 3.16: Lighting Analysis of the simple model simulated at 10:30 as an Interior space. .........27 3.17: Example of Raytracing on the detailed model at 10:30, set as an Interior Solution. ....27 3.18: The Total Average of all the Data collected from all Models using the Interior Space Solution Compared to the Actual Space....................................................28
3.19: Lighting Analysis of the simple model simulated at 10:30 as an Exterior-Interior Space ....................................................................................................................29 3.20: Example of Raytracing on the detailed model at 10:30, set as an Exterior-Interior Solution. ...............................................................................................................29 3.21: The Total Average of all the Data collected from all Models using the ExteriorInterior Space Solution Compared to the Actual Space.......................................30 3.22: First floor of the atrium space, generated in CAD........................................................33 3.23: Second floor of the atrium space, generated in CAD. ..................................................34 3.24: Third floor of the atrium space generated in CAD. ......................................................34 3.25: Atrium Space in Lightscape..........................................................................................34 3.26: LSRAD command used for Lightscape's Simulation...................................................35 3.27: LSRAY command used for Lightscape Raytracing......................................................36 4.1: First floor plan of the Atrium space, with the location of the first 5 measurement locations (A-E).....................................................................................................38 4.2: Second floor plan of the Atrium space, with the location of the second 5 measurement locations (F-J). ...............................................................................38 4.3: Third floor plan of the Atrium space, with the location of the third 5 measurement locations (K-O). ...................................................................................................39 4.4: The total average of measurements from the CGMs using an Interior solution at the three different time frames...................................................................................40 4.5: Light Analysis for the Atrium Space simulated as an Interior Space at 11:00am. .........40 4.6: Light Analysis for the Atrium Space simulated as an Interior Space at 12:00pm. .........41 4.7: Light Analysis for the Atrium Space simulated as an Interior Space at 1:00pm. ...........41 4.8: Digital image from the Actual Atrium Space-1..............................................................42 4.9: Digital image generated by Lightscapes Raytracing Using Interior Simulation -1.......42 4.10: Digital image from the Actual Atrium Space-2............................................................42 4.11: Digital image generated by Lightscapes Raytracing Using Interior Simulation -2.....42 4.12: Digital image from the Actual Atrium Space-3............................................................43
xi
4.13: Digital image generated by Lightscapes Raytracing Using Interior Simulation -3.....43 4.14: The total average of measurements from the CGMs using an Exterior-Interior solution at the three different time frames. ..........................................................44 4.15: Light Analysis for the Atrium Space simulated as an Exterior-Interior Space at 11:00am. ...............................................................................................................44 4.16: Light Analysis for the Atrium Space simulated as an Exterior-Interior Space at 12:00pm. ..............................................................................................................45 4.17: Light Analysis for the Atrium Space simulated as an Exterior- Interior Space at 1:00pm. ................................................................................................................45 4.18: Digital image from the Actual Atrium Space-1............................................................46 4.19: Digital image generated by Lightscapes Raytracing Using Exterior-Interior Simulation -1........................................................................................................46 4.20: Digital image from the Actual Atrium Space-2............................................................46 4.21: Digital image generated by Lightscapes Raytracing Using Exterior-Interior Simulation -2........................................................................................................46 4.22: Digital image from the Actual Atrium Space-3............................................................47 4.23: Digital image generated by Lightscapes Raytracing Using Exterior-Interior Simulation -3........................................................................................................47 4.24: Best fit line from the regression analysis, showing the final equation at the top of the graph...............................................................................................................48 A.1: The Daylighting illuminance measurements from the CGMs using an Interior solution at 11:00am. .............................................................................................56 A.2: The Daylighting illuminance measurements from the CGMs using an Interior solution at 12:00pm. .............................................................................................56 A.3: The Daylighting illuminance measurements from the CGMs using an Interior solution at 1:00pm. ...............................................................................................57 A.4: The total average of measurements from the CGMs using an Interior solution at the three different time frames...................................................................................57 A.5: Results of the Regression Analysis................................................................................58 A.6: The Daylighting illuminance measurements from the CGMs using an ExteriorInterior solution at 11:00am. ................................................................................61
xii
A.7: The Daylighting illuminance measurements from the CGMs using an ExteriorInterior solution at 12:00pm. ................................................................................61 A.8: The Daylighting illuminance measurements from the CGMs using an ExteriorInterior solution at 1:00pm. ..................................................................................62 A.9: The total average of measurements from the CGMs using an Exterior-Interior solution at the three different time frames. ..........................................................62 A.10: Results of the Regression Analysis..............................................................................63 B.1: The Daylighting illuminance measurements from the Simple CGM using an Interior solution at 10:30am. .............................................................................................65 B.2: The Daylighting illuminance measurements from the Simple CGMs using an Interior solution at 3:30pm. ..................................................................................65 B.3: The Daylighting illuminance measurements from the Simple CGM using an Interior solution at 6:30pm. ...............................................................................................66 B.4: Average of all the Data collected from the Simple Model using the Interior Space Solution Compared to the Actual Space. .............................................................66 B.5: The Daylighting illuminance measurements from the Moderate CGM using an Interior solution at 10:30am. ................................................................................68 B.6: The Daylighting illuminance measurements from the Moderate CGMs using an Interior solution at 3:30pm. ..................................................................................68 B.7: The Daylighting illuminance measurements from the Moderate CGM using an Interior solution at 6:30pm. ..................................................................................69 B.8: Average of all the Data collected from the Moderate Model using the Interior Space Solution Compared to the Actual Space. .............................................................69 B.9: The Daylighting illuminance measurements from the Detailed CGM using an Interior solution at 10:30am. ................................................................................71 B.10: The Daylighting illuminance measurements from the Detailed CGM using an Interior solution at 3:30pm...................................................................................71 B.11: The Daylighting illuminance measurements from the Detailed CGM using an Interior solution at 6:30pm. ..................................................................................72 B.12: Average of all the Data collected from the Detailed Model using the Interior Space Solution Compared to the Actual Space. .............................................................72 B.13: Lighting Analysis of the simple model simulated at 10:30 as an Interior Space.........73
xiii
B.14: Lighting Analysis of the simple model simulated at 3:30 as an Interior Space. ..........73 B.15: Lighting Analysis of the simple model simulated at 6:30 as an Interior Space...........73 B.16: Lighting Analysis of the moderate model simulated at 10:30 as an Interior Space.....73 B.17: Lighting Analysis of the moderate model simulated at 3:30 as an Interior Space.......73 B.18: Lighting Analysis of the moderate model simulated at 6:30 as an Interior Space.......73 B.19: Lighting Analysis of the detailed model simulated at 10:30 as an Interior Space.......74 B.20: Lighting Analysis of the detailed model simulated at 3:30 as an Interior Space.........74 B.21: Lighting Analysis of the detailed model simulated at 6:30 as an Interior Space.........74 B.22: Example of Raytracing on the detailed model at 10:30, set as an Interior Solution. ...74 B.23: Example of Raytracing on the detailed model at 3:30, set as an Interior Solution. .....74 B.24: Example of Raytracing on the detailed model at 6:30, set as an Interior Solution. .....74 B.25: The Daylighting illuminance measurements from the Simple CGM using an Exterior-Interior solution at 10:30am. .................................................................76 B.26: The Daylighting illuminance measurements from the Simple CGMs using an Exterior-Interior solution at 3:30pm. ...................................................................76 B.27: The Daylighting illuminance measurements from the Simple CGM using an Exterior-Interior solution at 6:30pm. ...................................................................77 B.28: Average of all the Data collected from the Simple Model using the Exterior-Interior Space Solution Compared to the Actual Space....................................................77 B.30: The Daylighting illuminance measurements from the Moderate CGMs using an Exterior-Interior solution at 3:30pm. ...................................................................79 B.31: The Daylighting illumi nance measurements from the Moderate CGM using an Exterior-Interior solution at 6:30pm. ...................................................................80 B.31: Average of all the Data collected from the Moderate Model using the ExteriorInterior Space Solution Compared to the Actual Space.......................................80 B.32: The Daylighting illuminance measurements from the Detailed CGM using an Exterior-Interior solution at 10:30am. .................................................................82 B.33: The Daylighting illuminance measurements from the Detailed CGMs using an Exterior-Interior solution at 3:30pm. ...................................................................82
xiv
B.34: The Daylighting illuminance measurements from the Detailed CGM using an Exterior-Interior solution at 6:30pm. ...................................................................83 B.35: Average of all the Data collected from the Detailed Model using the ExteriorInterior Space Solution Compared to the Actual Space.......................................83 B.36: Lighting Analysis of the simple model simulated at 10:30 as an Exterior-Interior Space. ...................................................................................................................84 B.37: Lighting Analysis of the simple model simulated at 3:30 as an Exterior-Interior Space. ...................................................................................................................84 B.38: Lighting Analysis of the simple model simulated at 6:30 as an Exterior-Interior Space. ...................................................................................................................84 B.39: Lighting Analysis of the moderate model simulated at 10:30 as an Exterior-Interior Space. ...................................................................................................................84 B.40: Lighting Analysis of the moderate model simulated at 3:30 as an Exterior-Interior Space. ...................................................................................................................84 B.41: Lighting Analysis of the moderate model simulated at 6:30 as an Exterior-Interior Space. ...................................................................................................................84 B.42: Lighting Analysis of the detailed model simulated at 10:30 as an Exterior-Interior Space. ...................................................................................................................85 B.43: Lighting Analysis of the detailed model simulated at 3:30 as an Exterior-Interior Space. ...................................................................................................................85 B.44: Lighting Analysis of the detailed model simulated at 6:30 as an Exterior-Interior Space. ...................................................................................................................85 B.45: Example of Raytracing on the detailed model at 10:30, set as an Exterior-Interior Solution. ...............................................................................................................85 B.46: Example of Raytracing on the detailed model at 3:30, set as an Exterior-Interior Solution. ...............................................................................................................85 B.47: Example of Raytracing on the detailed model at 6:30, set as an Exterior-Interior Solution. ...............................................................................................................85
xv
Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Interior Design EVALUATING LIGHTSCAPES ACCURACY FOR PREDICTING DAYLIGHTING ILLUMINANCE COMPAERD TO AN ACTUAL SPACE By Mohamed Adel Samy Shalaby May 2002 Chairman: Janine King Major Department: Interior Design. In the last decade, designers have rediscovered daylightings potential for reducing energy consumption. The major obstacle for widespread use of daylight to illuminate buildings is the lack of simple and accurate procedures for predicting interior daylight illuminance the amount of light on a surface. Although there are sophisticated daylight calculation tools that can accurately predict interior daylight illuminance, they are time consuming and difficult to use. Thus, designers are reluctant to use these tools early in the design phases. The purpose of this research is to evaluate the accuracy and application of Lightscape computer modeling software as a preliminary design tool to be used in early design process. In preparation for this research, a pilot study was conducted to test Lightscapes ability to produce daylighting illuminance measurements that are close to measurements recorded from the actual space. This study was conducted in a classroom setting at the University of Florida. This space was day lit by windows on the south faade. The pilot
xvi
study tested Lightscapes accuracy with three computer generated models of varying complexity: 1) simple model minimum details, 2) moderate model adding more details but with no furniture, and 3) a detailed model adding furniture. Comparisons between daylight measurements from the different computer models and measurements from the actual classroom showed that a simple computer generated model minimum detail predicted illuminance measurements more accurately than measurements from the more complex models. The subsequent research analyzed Lightscapes performance in a more complex space, where daylighting was introduced into the space through a skylight. Tests were performed on the Gainesville Regional Utilities atrium space to calculate a correction coefficient factor that could be used in early design processes. One of the factors tested in this study was the Daylight Parameters feature in Lightscape Interior Solution and Exterior-Interior Solution. Comparisons between both computer-generated models and the actual atrium space indicated that the Interior Solution provided daylighting illuminance measurements that were seventeen to thirty five percent 17-35% of the illuminance measurements of the actual atrium. In addition, a regression analysis was performed to provide a correction coefficient factor to better estimate daylighting illuminance measurements in an interior space. Therefore, it was concluded that Lightscape is an accurate daylighting-predicting tool to be used in preliminary design process. Furthermore the correction coefficient factor for the measurements from Lightscape could be utilized to provide better estimates of daylighting illuminance in interior environments.
xvii
CHAPTER 1 INTRODUCTION Increasing productivity has become a compelling economic rationale for supplementing electric lighting with daylighting. Although daylighting provides tremendous opportunities for energy conservation, the total operation and maintenance expenses of daylighting tools represent only six percent of a buildings costs over its entire life cycle [1]. In a study conducted to evaluate the relationship between daylighting design and productivity in different environments, it was discovered that in retail spaces skylights were found to be positively and significantly ninety nine percent statistical certainty correlated to higher sales. Also in schools, it was discovered that students with the most daylighting in their classrooms progressed twenty percent faster on math tests, and twenty six percent faster on reading tests in one year than those with the least [1]. The obstacle of incorporating daylighting in design is the lack of procedures that help designers predict daylight in a simple, useful, and accurate manner. Guglielmetti [2] said that in their office, they have been using Lightscape in all their lighting design work. They use it primarily for calculations, and also they use computer graphics to better understand the visual effect of daylight on buildings they design. There are a multitude of calculations that must be performed in the daily work of designing the lighting of a project. Many of these calculations are performed using a simple calculator, but there is always an area or two that will involve some complex geometry or intense calculations. This is where Lightscape is useful; designers can create models to study both the visual effects of different lighting solutions for any given space
2 and to gather illuminance data. According to Renfro and Guglielmettis study in testing Lightscapes daylight accuracy [3], Lightscape measured illuminance levels of direct sun more accurately than the illuminance levels of indirect light from the north sky. They collected daylighting data from a scale model, and compared the measurements from the scale model to that of a computer-generated model simulated in Lightscape. Other studies tested Lightscapes accuracy in visualizing spaces illuminated by both electric and daylight. Boora Architects [4] compared four of the daylighting design tools, one of which was Lightscape. They came to a conclusion that Lightscape simulation generated simultaneous qualitative and quantitative information for both electric light and daylight. However they cited problems stemming from the time required to learn the software and that the results did not replicate the feeling of being inside the space. On the other hand, both the architectural team and lighting designers saw expanded use of computer simulations in future studies. Maunder [5], a student in Victoria University, compared a number of computersoftware programs for their accuracy in visually representing the built environment. Lightscape was one of the programs in question. Analysis determined that Lightscape ranked second software after Rayfront Lighting Simulation Software, in providing more realistic images of architectural environments. A similar study conducted by Thurnauer [6], compared Lightscapes rendered images with pictures from the actual atrium space. The conclusion was that Lightscape, with careful handling, could be used as a visualizing tool to replicate architectural environments. Most of those studies were conducted to test Lightscapes accuracy as a rendering tool, to replicate architectural and interior environments. However, not enough experiments tested Lightscapes accuracy in
3 measuring daylighting illuminance in interior spaces. The purpose of this research is to analyze the accuracy of Lightscape in predicting and representing illuminance in interior and exterior environments. Understanding the accuracy of this design tool will assist and encourage designers who wish to incorporate daylighting design into their projects. To carry out this research, a pilot study was conducted on a smaller setting. This study was performed to determine what structure and methods should be used in the actual research. In the pilot study, Lightscapes performance accuracy of predicting daylight illuminance was tested in a classroom setting in the College of Design, Construction and Planning building at the University of Florida. This study was conducted during the spring semester April 2001. Based on Renfro and Guglielmettis [3] conclusion that Lightscape is inaccurate with indirect light from the north sky it was decided that the classroom should be located on the south faade of the building. First, daylight illuminance was measured under clear sky condition at five different locations in the classroom. Then, three computer-generated models (CGM) of the classroom were constructed using CAD. The CGMs were built with varying degrees of detail: 1) a simple model, 2) a moderate model, and 3) a detailed model. The CAD files were then imported into Lightscape and daylighting illuminance was measured in the same locations within each model. Comparisons of the illuminance data collected from both the actual classroom and the different CGMs were recorded. The major finding was that a simple computer generated model simulated in Lightscape represented daylighting illuminance more accurately than both the moderate and the detailed models of the classroom.
4 The actual research succeeded the pilot study; it analyzed Lightscapes performance in a more complex space the atrium space at Gainesville Regional Utilities (GRU) downtown office building. Comparison analysis explored the relationship between real daylighting and computer generated daylighting illuminance. The GRU building was designed in the early 1990s to include daylighting as one of the design factors. This research was conducted during the fall semester October 2001. Illuminances were first measured in the enclosed atrium area under clear sky condition at fifteen different locations within the atrium space five measurements within each of the three floors. These measurements were recorded three times during the course of the day. Then, a simple CGM was generated of the atrium space in CAD. The CAD file was imported into Lightscape, and illuminance data was recorded and compared with data from the actual atrium space. The data collected from Lightscape was not consistent within each floor, so the average measurements of each floor were compared from both Lightscapes CGM and the actual atrium spaces. The comparison showed that the average of Lightscapes measurements were thirty percent less than that of the actual atrium space. For example if the measurement in the actual space were one hundred footcandles, it would be seventy foot-candles in Lightscape. Comparison analysis showed that Lightscape is an accurate daylighting-predicting tool, for preliminary design process. Moreover, a correction coefficient was calculated for more accurate daylighting illuminance measurements. 1.1 Limitations of the Study In order to replicate this research, the same sun angles should be used on the computer-generated models in Lightscape; also the same sun light intensity should be
5 used. Changing the sun angles or the sun intensity would change the final results of Lightscapes radiosity and ray tracing calculations in turn changing the final daylighting illuminance measurements in the interior and architectural environment. Both the pilot and the actual studies were conducted at a specific time of the year and measurements were taken at definite time frames during the day. They were both conducted under clear sky conditions. It is recommended that consideration of those limitations will generate results closer to the ones discovered in this study. Another factor to be considered in this study is that the pilot study introduced daylight though windows into the space, while in the actual research a skylight was used to introduce daylight into the interior space. 1.2 Significance of the Study This study provides us with knowledge about Lightscapes daylighting illuminance accuracy as a measuring and calculating tool to be used in preliminary design process. Moreover, it was conducted in an actual architectural space, whereas previous research tested Lightscapes visual performance and its ability to create accurate and realistic representation of the architectural environment from a graphics standpoint [5, 6]. In addition, this research provides a correction coefficient factor that might be useful for designers who want to calculate the amount of daylighting illuminance in an architectural and interior environment they have designed and have modeled in Lightscape.
CHAPTER 2 REVIEW OF LITERATURE 2.1 Background Prior to the invention of artificial lighting, designers were obliged to design their architectural environments in response to sun and daylighting. However, after the invention of artificial light, the concern for daylightings role in the design and function of architectural environments started to wane [7]. Since engineers are primarily concerned with quantitative aspects of lighting interiors rather than the ambiance lighting provides an interior environment [6]. That is because the primary method used for designing the lighting in a space is mathematical calculations, which only determines the amount of light received at a point on a surface. Nevertheless, in the midst of the last century, examples of lights role in shaping architectural forms started to emerge. For example, Le Corbusier [8] said that Architecture is the masterly, correct and magnificent play of masses brought together in light (p.8). With the development of artificial lighting, the illumination levels of interior environments could be more accurately controlled. Shaping form was not the only reason why design professionals started reconsidering daylighting in their designs. With rising concerns for the environmental costs of energy consumption, efficient use of resources has become a primary concern in building professions. Currently, building professionals are reassessing the standard practices of environmental engineering to address concerns regarding energy efficiency, operating costs and environmental quality [9]. The current professional and research literature cites daylighting as a strategy that offers lower energy costs, reduces peak power demand, and increases employee productivity while providing 6
7 a humane quality [10]. Renewed interest in daylighting has increased the design of lighting related architectural features, supplemental electric lighting systems, and lighting control systems that take advantage of daylight as a source of interior illumination [9]. In addition to providing opportunities for energy conservation, daylighting introduced to interior environments may also affect peoples productivity. A recent study conducted on retail and educational facilities by Heschong Group [1] found significant differences in productivity between people shopping and studying in environments that incorporated daylighting and those environments that used only electric light to illuminate the stores and classrooms. In retail spaces it was shown that daylighting from skylights was the second best predictor of sales per store after the number of hours open per week. In schools, it was discovered that students studying in classrooms that had the most daylighting progressed faster than those in classrooms with the least. These studies suggest that daylighting can reduce the amount of energy consumed by the built environment, and it can increase productivity of users interacting with these environments. 2.2 Daylighting Tools Design professionals are interested in using daylighting because it can also animate architectural spaces and promote occupant well being [9]. Daylight, varying in both intensity and direction over time, changing from a point to a diffuse source, challenges the lighting design process [4]. With daylighting as the light source, the building itself becomes the light fixture, and its design must be executed in the same manner that lighting engineers use to design artificial lighting fixtures. To verify whether a daylighting design is fulfilling designers expectations, design professionals typically test daylighting options using different visualization and calculation tools. Analysis
8 methods of daylighting use graphic, numerical or physical model techniques to predict quantity of light, its distribution scheme and in some case the visual appearance of daylighting scheme [9]. Also Erwine [5] stated, For a more realistic understanding of daylights dynamics, the design tool kit should include one or more specialized prediction tools a scale model, computer lighting simulation or full-scale mockup of the proposed design (p.1). In general, there are four categories of tools used to predict daylight distribution in a space. The first category uses calculation procedures formulated to show daylighting illuminance levels for rapid execution in early stages of design process. The most common calculation methods are Daylight Factor Method and the Lumen Method [11]. The second category of daylight prediction tools is called Elaborate Programs e.g. (Superlite, and DOE 2.1C), these simulation programs are capable of predicting light levels in relatively complex spaces under changing sky conditions [9]. Physical models are the third category. They provide information useful for evaluating the visual quality of daylighting. Models are a traditional architectural tool and have been used to visually assess lighting effects since the beginning of the century [12]. In general the main shortcomings of any of these three methods is that they are time consuming and some of them require extensive preparations before gathering daylight measurements [11]. Therefore they are less useful to designers at the early stages of design, but they can be performed to obtain accurate daylighting illuminance in an interior environment. Moreover, they do not provide visual feedback that illustrates how the built environment will be experienced. Over the last decade a fourth category of light measurement and visualization systems have been developed. These new tools use computer-based computations to
9 replicate the lighting design [13]. In the early 1960s thermal engineers develop methods for simulating the radiative heat transfer between surfacesin the mid 1980s, computer graphics researchers began investigating the application of this techniques for simulating light propagation p.5 [14]. Lightscape belongs to this fourth category and is an advanced lighting and visualization application for creating a realistic three-dimensional model or a space. Lightscape provides instant visual feedback about lighting strategies, at any point in the design process. Primarily the advantage for using Lightscape instead of the other methods is that it presents both visual feedback and lighting calculations of the designed space at any time of year or day a user wishes. 2.3 Computer Graphics The growth of computer technologies has lead to many developments in the field of electronic design assistants. For instance, architects and designers are using computers to assist in drawing, constructing and analyzing their designs. Three-dimensional capabilities of todays standard CAD packages are improving with each new release. Also, the advancement of computer graphics (CG) technology and the development of advanced rendering software are providing major assistance to designers in visualizing and analyzing their designs. Improvement of CG software provides invaluable service for most architectural and design professionals who use realistic images of the built environment, interactive animations, daylight studies and artificial lighting simulations to test design ideas within their design process. The goal behind the improvement of CG is to convey ideas about a design from the designer to the client, and sometimes for the designer to better understand their designs. Gallardo [15] says that the architectural CG industry with respect to lighting has grown a lot in recognition of CG as a viable tool for lighting analysis and evolution. Lighting manufacturers now have IES data online along
10 with three-dimensional luminaire blocks for three-dimensional rendering applications. Many manufacturers are embracing the growing technology of the architectural CG industry [3]. This indicates that overall interest in lighting design and CGMs has grown considerably in the last decade. Most designers and design companies are now using new visualization tools to guide them in their designs decisions. How a software renders a computer model is closely related to how it is created. Lightscapes Technological Overview [14] explains that a three-dimensional model contains geometric data defined in relationship to a three-dimensional Cartesian coordinate system. The image on a computer monitor is made up of large numbers of illuminated dots called pixels. The task in creating a computer graphics image of a geometric model is to determine the color of each pixel on the screen screen space based on the model information and specific viewpoint. The color of any specific point on a surface in a model is determined by mathematical computations that use physical material properties of that surface and the light that illuminates it as primary variables. Lightscape uses two general shading algorithms local and global illumination, to determine how model surfaces reflect and transmit light. Local illumination uses information about light arriving at a surface, to calculate the intensity, spectral character color, and distribution of the light leaving that surface. Global illumination uses two algorithms. One is called Ray Tracing, which traces rays backwards from each pixel on the screen into the three-dimensional model. The other is Radiosity, which calculates the intensity for discrete points in the environment.
11 2.4 Case Studies Several case studies demonstrate the impact of daylighting in designing architectural environments has been an increasingly popular topic in studies since the 1990s. 2.4.1 Daylighting Decisions BOORA Architects [4] in Portland compared four daylighting design tools to compare lighting decisions in a school design a scale model, computer lighting simulation or full-scale mockup of the design solution. Also a building energy simulation was employed to evaluate impact of daylight on building energy use. The study was conducted to determine which tools most accurately resemble the built environment. The researchers generated the design solutions in each of the design tools. Lightscape simulation program was enlisted for its radiosity algorithm to calculate light levels and generate photo-realistic images appropriate for client presentations. The second lighting tool used was a scale model; a ( in. = 1ft) scale model was constructed to study the same classroom wing as the Lightscape simulation. Light levels were recorded as daylight factors ratios of interior exterior illumination and qualitative effects were documented photographically. The third tool used was a full-scale mock-up of the classroom space. The mockup proved to be the compelling tool that crystallized the design approach for architect, lighting designer, and client. The lighting designer commented, Scaled models and computer models help immensely; to experience a full scale mockup however, is to experience the actual space(p.3) [4]. Finally a building energy simulation was conducted. From their analysis, the building energy simulation gave valuable feedback on energy costs and balances but gave no qualitative feel to the space.
12 Comparisons showed that the Lightscape simulation generated simultaneous qualitative and quantitative information for both electric light and daylight. The scale model judged by this design team to be the best overall daylighting tool, but it gave no electric lighting information and quantitative measurements. The full-scale mockup it was the most expensive, suffered quantitative inaccuracies, but it was priceless in quickly bringing the group to a design consensus and was pivotal in garnering the confidence of the client(p.4) [4]. In their conclusion, the best tool should be determined using type of project, time and financial constrains as primary criteria. The architectural and lighting designers agreed that once people could easily use computer simulations software the use of CGMs would increase. 2.4.2 Software Comparisons Maunder [5], a graduate student at Victoria University of Wellington, New Zealand, conducted a study to test the following four different lighting simulation programs; Inspirer, Rayfront, Lightscape and ArchiCAD. Maunder tested each of the four programs with simple geometry. The programs were then ranked in the order of precision in representing the physic of light how the lighting appears in the space. The ranking was: 1) Inspirer, 2) Rayfront, 3) Lightscape, and 4) ArchiCAD. Although, Inspirer and Lightscape are similar in many ways; the major difference is the operating interface. Inspirer is a DOS executable program, and Lightscape is a Windows-based program. Rayfront produced some unwanted complications that slowed down the rendering process. However, in concluding this study the researcher decided that these programs could be used to replicate the built environment, and that they could present photo realistic images of the designed space.
13 2.4.3 Atrium Study Another study tested Lightscapes accuracy when representing the built environment [6]. This study was conducted at Miami University in the atrium space in the Alumni Hall. The purpose of this test was to identify advantages and limitations of Lightscape. Comparison analysis was performed between images from the actual atrium space and images simulated in Lightscape. It was concluded, One of the most difficult problems to overcome in Lightscape was the ability to produce semi-reflective surfaces. Since Lightscape considers all surfaces to be perfect Lambertian diffuse surfaces for Radiosity processing, the specularity of a surface has no effect on the Radiosity solution (p.3) [6]. The other problems exist in how surfaces appeared to be either perfectly diffuse or perfectly reflective, and the changes in contrast of the image before and after Raytracing. However, Lightscape rendering processes created an accurate representation of the space, because it allows the color of materials in the space to influence the quality of lighting in a space. This differs from traditional physical daylight analysis models where the surfaces are often produced of all white, and material quality does not influence results. Traditionally changing the color of surfaces in a physical model would be a time consuming task, but in Lightscape, a material can be altered in seconds and a new solution generated. Another advantage of Lightscape is the ability to switch luminaries in a space. Overall Lightscape has proven to be accurate enough in these results to be used for design (p.1) [6]. 2.4.4 Lightscapes Daylighting Accuracy Test Renfro and Guglielmetti [3] conducted a study to test Lightscapes accuracy in measuring daylighting. Their study compared data collected from a physical scale model and from a CGM in an open site under clear sky conditions. Results of this research
14 showed that the radiosity algorithm used in Lightscape to compute levels of light intensity is extremely sensitive to proper modeling procedure(p.3) [3]. Another shortcoming in Lightscape is the apparatus used for modifying sky conditions, currently it is a slider that ranges sky conditions between zero (0) representing a cloudless sky, to one (1) representing full overcast. This in turn, allows only two distinct options either clear or full overcast sky conditions. That is the reason why the tests were conducted under clear sky conditions. It was also proven that testing Lightscape with direct sun was very accurate while indirect light from the north sky was significantly different. Finally coming to a conclusion that Lightscape is accurate enough for qualitative lighting analysis, provided care is taken in the modeling stage, however a correction factor may be devised to adjust illumination data for this method of calculation(p.1) [3]. 2.5 Purpose of This Study Although these studies give us valuable information about Lightscapes performance and accuracy, several questions are still unanswered. Most of the studies discussed above were performed to either test the impact of daylighting on the users, or test a group of computer simulation programs regarding their ability to convey a realistic image of the built environment. The only study that tested Lightscapes accuracy as a daylighting design tool was Renfro and Guglielmettis study. However, their study was conducted using a simple physical model, to compare the illuminance data collected from the physical model with that of a CGM. Using daylighting illuminance measurements recorded from an actual space compared to daylighting illuminance measurements from computer-generated models simulated in Lightscape. Therefore, the purpose of this research is to evaluate the
15 accuracy and application of Lightscape computer modeling software as a preliminary design tool to be used in early design process.
CHAPTER 3 MATERIALS AND METHODS Recent research conducted by Renfro and Guglielmetti [3] recommends that more research be conducted to test whether it is possible to derive a correction coefficient factor that would improve Lightscapes daylighting illuminance measurements. Thus, the overall purpose of this study is to test Lightscapes daylighting accuracy, and to mathematically calculate a correction coefficient factor. Moreover, this research performed the tests on an actual architectural environment, instead of a physical model to be compared with the CGMs from Lightscape. In preparation for this research, a pilot study was conducted on a small and more controlled environment. This pilot study was performed to test lighting measurement methods and to determine the structure for the proceeding research (Figure 3.1). Findings from the pilot study helped generate an overall procedure for conducting the research. Pilot Study: DCP Building Classroom at UF Research Parts Actual Research: GRU Building Atrium Space
Figure 3.1: Research Parts.
16
17 3.1 Pilot Study 3.1.1 Introduction The pilot study was conducted using a classroom in the College of Design, Construction and Planning building at the University of Florida. The classroom chosen for this study was located on the south faade of the building. Daylight was introduced from a window located on the south faade of the classroom. The pilot study was conducted during the 1st third of the year April 2001. The purpose of the study was to test the effect of two factors on the accuracy of Lightscape. The two factors are: 1) Lightscapes Daylight Parameter with two levels: a) Interior Space. b) Exterior-Interior Space. 2) Model Complexity, which has three levels: a) Simple model. b) Moderate model. c) Detailed model. Testing two major factors resulted in a six factor-level combination. Measurements from all of those factor-level combinations were recorded and comparison analysis with measurements from the actual space tested the accuracy of Lightscapes data. Process Parameters is a feature built into Lightscape that allows the user to instruct the software about what type of space it will simulate (Figure 3.2). In the Process Parameters there is the Quality Wizard, which runs the appropriate calculations and determines the spacing of the surface mesh units, and other information used for the model simulation. In the Quality Wizard (Figure 3.3), there are five different level selections that determine the quality of the simulation required, scaling from one: lower
18 quality requires less time and memory, to five: higher quality requires more time and memory. Level three the default was used for this researchs purposes. Once the quality level is chosen, it is required to identify whether daylighting is introduced into the space or not this is the first factor in the study which is called Daylight Parameter. There are three selections in the Daylight Parameter, the first sets the model to be an interior space, second sets the model to be an exterior space and the third selection sets the model to be an exterior-interior space (Figure 3.4). This study tested two selections from the Daylight Parameters a) interior space, and b) exterior-interior space. These two parameters differ in regard to what part of the model they use to gather the data for the lighting calculation. The interior space uses only the interior surfaces whereas the interior-exterior uses a combination of interior and exterior surfaces.
Figure 3.2: Lightscapes Process Parameters.
Figure 3.3: Lightscapes Quality Wizard.
19
Figure 3.4: Lightscapes Daylight Parameter. The second factor of this study is the level of complexity of the CGMs. The pilot study tested three levels for the models complexity; a) A simple model basically a box representing the space with glass on the wall where the windows are located, this glass layer covers the whole wall with no frames for the windows except at the connections with the sidewalls. b) A moderate model similar to the simple model, but frames to the windows replicating the actual space are added. c) A detailed model where furniture has been added to the space to represent the furniture in the actual space, the modeled desks and chairs are similar to those of the actual classroom, and were generated from Lightscapes library of furniture. 3.1.2 Methods The researcher began the pilot study by measuring the actual daylighting levels in the classroom space (Figure 3.5). Using a light meter and turning off the electrical lights in the classroom, illuminance data was recorded at five different locations at a height of three feet approximately desktop height. Measurements at these locations were recorded at three different times of the day 10:30 am, 3:30 pm, and 6:30 pm. These times were chosen to give a range of daylight illuminance measurements throughout the
20 day. Three-daylighting illuminance measurements were repeated in each location over a period of sixty seconds. An average of the three readings was used. Daylight measurement was recorded outside the building to establish the daylighting illuminance to be used in the CGM. Interior finish materials were also recorded for use in the CGM. Measure Interior Illuminance Levels Measure Exterior Daylight Levels Record Interior Materials Figure 3.5: Outline of Study Procedures Actual Space. In order to compare daylight data from the actual classroom with that of the digital model, the classroom was recreated in a CGM (Figure 3.6). CAD and construction drawings from the University of Florida planning department were used to create three three-dimensional models of the space simple, moderate, and detailed (Figures 3.7, 3.8, and 3.9). The three models were then exported to Lightscape (Figures 3.10, 3.11, and 3.12). Interior finish materials were specified in Lightscape using the material library provided with the software. A light meter was modeled using a 1x , in order to most accurately replicate light readings in the original space. These light meters were placed at the same measurement locations as in the actual classroom. They were also set to be nonoccluding and non-reflecting surfaces. Daylight was introduced to the space using Lightscapes daylight feature. The measurements recorded from outside the building plus information from sundesign.com [16] provides sun angles at a specific time, were used to determine the daylight settings.
Actual Space
21 There are two methods to add daylight in a model using Lightscapes daylight feature. The first method determines the sun angles by setting a location and time for the simulation (Figure 3.13). The number of cities that have latitude and longitude information stored in Lightscapes database limits the amount of choices this method offered. To over come this shortcoming, Lightscape allows users to input their own latitude and longitude information. Initial testing using this method provided inaccurate results. The reason for the inaccuracy could be the lack of control over the amount of sunlight intensity at the time chosen to perform the simulation. The second method, which is the direct control method, uses sun angles and sun intensity provided by the user. The second method was adopted for this study because it allows more direct control over sunlight intensity and location (Figure 3.14). Setting daylight levels in Lightscape involves inputting the angles and sun illuminance information. Using level three in the Quality Wizard, and setting the Daylight Parameters, the sets of CGMs were prepared for the initial simulation. The models were run through one hundred iterations. The number of surfaces processed in the simulation determined the number of iterations. The number of surfaces processed in the simulation, reaching approximately ninety eight percent of energy distributed in the model. The CGMs daylight illuminance measurements were then taken in the same locations as in the actual space, using the average foot-candles provided by Lightscapes lighting analysis feature and the data was recorded on to a spreadsheet.
22
Construct CGM in CAD Import CGM into Lightscape Computer Generated Model (CGM) Specify Materials in Lightscape Set Daylighting Levels Measure Illuminance Figure 3.6: Outline of Study Procedures - CGM
Figure 3.7: The Simple Model generated in CAD
Figure 3.8: The Moderate Model generated in CAD
23
Figure 3.9: The Detailed Model generated in CAD
Figure 3.10: The Simple Model in Lightscape
Figure 3.11: The Moderate Model in Lightscape
24
Figure 3.12: The Detailed Model in Lightscape
Figure 3.13: First Method Setting a Location
Figure 3.14: Second Method Setting Sun Angles and Intensity
25 Similarly, this whole process was repeated a number of times to create different simulations for: a) Daylight Parameters interior and exterior-interior solutions, b) different model types simple, moderate, and detailed, and c) different times of the day 10:30 am, 3:30 pm, and 6:30 pm. Analysis compared the collected data from both conditions, and tested the relationship between the data recorded from the actual classroom and data recorded from the CGM. 3.1.3 Results and Discussion Illuminance data were collected from the actual classroom and from the different CGMs at five different locations A-E (Figure 3.15). Information in Table 3.1 was used to generate light simulations in Lightscape, using the sun angles calculated and the sunlight illuminance measured outside the building. Daylighting illuminance data was then recorded on to a spreadsheet. All data collected are presented in Appendix B.
Table 3.1: The Information that was used for generating the Light simulation in Lightscape Azimuth (light) 342 212 180
Time Altitude 10:30 AM 42.48 3:30 PM 54.78 6:30 PM 17.2 Date 4/10/01 Altitude from Horizon Azimuth from South Latitude is 29.64 degrees North Longitude is 82.39 degrees West
Azimuth 72.36 -58.43 -89.7
Sun (fc) 9000 fc 8500 fc 3000 fc
26
Figure 3.15: Plan of the Classroom showing the Measurement Locations in the Space. 3.1.3.1 Interior Space Figure 3.16 shows an example of the Lighting Analysis built in feature in Lightscape that shows the illuminance level distribution in the space, of the simple model simulated at 10:30 am. As shown in the Figure 3.16 the area around the classroom is in dark blue which illustrates that there is no lighting illuminance on these areas, while inside the classroom its apparent that the lighting levels vary from the front of the classroom where there might be direct sunlight red portion of the diagram, then the color starts to fade to blue moving towards the end of the classroom. The reason behind this drastic difference is that this model is treated as an Interior Space, so Lightscape discards any daylighting illuminance outside the model. Figure 3.17 shows an example of Raytracing the detailed model at 10:30, which is the visual representation of lighting illuminance levels in a space. Figure 3.18 shows the total average of measurements from the three CGMs using an Interior solution. The averages were calculated by dividing all
27 measurements at the specific time frames for the three CGMs and the actual space. It is clear in the Graph that the behavior of the lines is similar between the three CGMs and the actual space. However, the simple model showed closer values to the actual space than the moderate and detailed models.
Figure 3.16: Lighting Analysis of the simple model simulated at 10:30 as an Interior space.
Figure 3.17: Example of Raytracing on the detailed model at 10:30, set as an Interior Solution.
28
Total Average Interior
60
50
40 Averages (fc)
30
20
18 17
19 19 15 12
17
10 6 4 0 Simple Actual Space Moderate Detailed 10:30 18 17 6 4 3:30 19 19 7 4 Time Frame 7 4 6:30 12 15 3 2 3 2 17 17 5 3 5 3 Total
Figure 3.18: The Total Average of all the Data collected from all Models using the Interior Space Solution Compared to the Actual Space. 3.1.3.2 Exterior-Interior Space Figure 3.19 shows an example of the Lighting Analysis of the simple model simulated at 10:30 am. The area around the classroom is in bright red in this situation indicating that there is lighting illuminance on these areas, and inside the classroom its apparent that there is higher illuminance levels than in the case of the Interior Solution. The explanation of this case is exactly the opposite than what was seen in the Interior Space Solution; the reason is that Lightscape treats this model as an exterior building, applying daylighting illuminance outside and inside the space. Figure 3.20 shows an example of Raytracing the detailed model at 10:30, which is the visual representation of lighting illuminance levels in a space. Figure 3.21 shows the total average of measurements from the three CGMs using an Exterior-Interior solution. The averages
29 were calculated using the same method used in the Interior solution. In this graph the detailed model showed closer illuminance measurements to that of the actual space, while the simple and moderate models had values almost twice as the actual space. Still, the line behavior in from the CGMs was close to that of the actual space.
Figure 3.19: Lighting Analysis of the simple model simulated at 10:30 as an ExteriorInterior Space
Figure 3.20: Example of Raytracing on the detailed model at 10:30, set as an ExteriorInterior Solution.
30
Averages of Exterior-Interior Models
60 53 50 47 40 Average (fc) 30 20 17 10 0 Simple Actual Space Moderate Detailed 31 28 22 19 26 19 17 42
29 23
16 15 10
10:30 47 17 29 23
3:30 53 19 31 22 Time Frame
6:30 28 15 16 10
Total 42 17 26 19
Figure 3.21: The Total Average of all the Data collected from all Models using the Exterior-Interior Space Solution Compared to the Actual Space. 3.1.4 Discussion This pilot study confirmed the results of Renfro and Guglielmettis study [3]; that found inaccuracy in Lightscapes prediction of indirect light levels. Lightscape does not account for ground reflection and building obstruction and this reduces the accuracy of predicting daylight illuminance. On the contrary, their analysis showed that Lightscapes measurements were within ten percent (-10%) of the actual setting in direct light. In this pilot study Lightscapes measurement of the Interior Solution ranged from thirty percent more (30%) to ninety percent less (-90%) than the actual space. The Exterior-Interior solution measurements differences ranged from two hundred and fifty percent more (250%) to sixty eight percent less (-68%) than the actual model measurements. However, when each models five location measurements were averaged at the three different time frames 10:30, 3:30 and 6:30, the difference between the average from the actual
31 classroom and the CGMs became smaller (Table 3.2). Furthermore, when all the measurement from each model, were averaged the results of the simple model set as an Interior Space (Table 3.3), it indicated that the simple model simulated as an Interior Space had better opportunities in resembling the actual space.
Table3.2: The Total Average of all the Data collected from The Simple Model using the Interior Space Solution Compared to the Actual Space. Average 10:30 3:30 6:30 Total Lightscape 18 19 12 17 Actual Space 17 19 15 17
Table 3.3: The Total Average of all the Data collected from all CGMs using the Interior Space Solution Compared to the Actual Space.
3.1.5 Conclusion
Location Simple Moderate Detailed Actual Space
Average (fc) 17 5 3 17
After conducting the pilot study, it was discovered that measurements from Lightscape were not one hundred percent (100%) accurate. However, the averages of daylighting illuminance measurements from each model provided smaller differences. Comparing all factor level combinations within the pilot study, showed that the simple CGM simulated as an Interior Space in Lightscape, provided daylighting illuminance measurements closer to those from the actual classroom. Concluding the pilot study with that: Simplified computer generated models simulated as an Interior Space in Lightscape
32 can more closely predict the daylighting illuminance measurements of the actual space, than the measurements from moderate or detailed computer generated models. Moreover, it was evident from the data that there was a correlation between illuminance measurements from the computer generated models and the actual space. 3.2 Actual Research 3.2.1 Introduction The actual research was conducted on a more complicated environment the atrium space at Gainesville Regional Utilities (GRU) office building. The building was originally designed with daylighting as one of the design factors. This research was conducted during the 3rd third of the year November 2001. Daylight was introduced into the space using a skylight covering the whole atrium space. Utilizing the findings from the pilot study, a simple CGM was used to represent the atrium space. To get a larger sample of daylighting illuminance, measurements were taken three times during the day 11:00 am, 12:00 pm and 1:00 pm, at five different locations within each floor. However, the factor tested in this research is: Daylighting Parameters with two levels: a) interior space, and b) exterior-interior space. Still using level three in Lightscapes Quality Wizard, but this time using another feature of Lightscape, which is LSRAD a DOS-based program in Lightscape which provides more control to how long the user wants to run the simulations. LSRAD has proven to be faster than Lightscapes regular simulation. Although both of them run the simulations in the same way, but probably the DOS-based program runs the simulation in the background, resulting in faster computations.
33 3.2.2 Methods Similarly the same procedures represented in Figures 3.5 and 3.6 were followed in the actual research. Measurements of daylighting illuminance were taken at fifteen different locations within the building five measurements within each floor. Another measurement was added from outside of the building to be used in Lightscapes simulation. All fifteen measurements and the one outside the building were taken three times during the day at each location; illuminance data was then inputted into a spreadsheet. Also the materials in the atrium space were recorded using digital images that were used in the material finishes in the CGM, and were also used to compare visual differences between the actual atrium space and the CGM rendered by Lightscape. A three-dimensional simple CGM was constructed in CAD using construction documents provided by the GRU (Figures 3.22, 3.23, and 3.24). Then the CGM was imported into Lightscape, materials were assigned to the surfaces in the atrium space, and glass surfaces were assigned as windows in Lightscape surface processing (Figure 3.25). Setting daylighting in Lightscape was performed in the same manner as in the pilot study; the angles were set using sundesign.com [16], and the sunlight intensity was added according to the daylight illuminance measurement taken outside the building.
Figure 3.22: First floor of the atrium space, generated in CAD.
34
Figure 3.23: Second floor of the atrium space, generated in CAD.
Figure 3.24: Third floor of the atrium space generated in CAD.
Figure 3.25: Atrium Space in Lightscape
35 Now the CGM was set to run the initial simulation, as mentioned earlier LSRAD was used (Figure 3.26) the simulation was run to two hundred and fifty iterations which was almost ninety eight percent of the total energy distribution, for both factor levels Daylight Parameters, a) interior and b) exterior-interior. This was also performed for the different times during the day, although it is not considered as a factor in this research only used to reach a larger sample. LSRAD [options] file name. C:\bin\lsrad v term 250 o lightint.ls lightint.lp C:\bin\lsrad v term 250 o lightext.ls lightext.lp LSRAD Is the command line for Radiosity -v Extra information is printed after each iteration -term Radiosity processing will stop after n times 250 -o Output solution file is specified. File name The lighting preparation file name filename.lp Figure 3.26: LSRAD command used for Lightscape's Simulation Then Radiosity was performed using LSRAY (Figure 3.27); this command was also used to test Lightscapes visual ability. Measurements were recorded at the same location as in the actual atrium space from both CGMs. Regression analysis was used for comparisons of both the CGMs and the actual spaces. The analysis compared the collected data from both CGMs interior space and exterior-interior space, and compared the relationship between the data recorded from the actual atrium space with data recorded from both CGM solutions.
36 LSRAY [options] solution_file image file C:\bin\lsray aa 3 vf 3.vw sh soft x 1350 y 900 v lightint.ls lightint.jpg C:\bin\lsray aa 3 vf 3.vw sh soft x 1350 y 900 v lightext.ls lightext.jpg LSRAY Is the command line for Raytracing -aa Antialiasing factor, Higher factor results in higher image quality, but takes more computation time. -vf Loads specific view file. -sh Recompute shadows from sun and other light sources. -soft Compute soft shadows. -x Image width. -y Image height. -v Extra information is printed after each iteration Figure 3.27: LSRAY command used for Lightscape Raytracing
CHAPTER 4 RESULTS AND DISCUSSION 4.1 Results The procedures developed in the pilot study were adopted to perform the actual research on the atrium space in the GRU building. The only addition was the calculation of a correction coefficient factor for Lightscapes illuminance measurements. This test was performed on the CGM of the atrium space with two different Daylight Parameters a) Interior Space, and b) Exterior-Interior Space solutions. Each solution was simulated three times 11:00 am, 12:00 pm and 1:00 pm, resulting in a total of six simulations of the CGM. Illuminance data was collected from the actual atrium space and from the different CGM solutions at fifteen locations (Figure 4.1, 4.2, and 4.3). Using information recorded from the site and from sundesign.com [16] (Table 4.1). All Figures and Graphs of the collected data are available in Appendix A.
Table 4.1: The Information that was used for generating the Light simulation in Lightscape Time Altitude 11:00AM 41.22 12:00PM 44.4 1:00PM 43.12 Date 11/5/01 Altitude from Horizon Azimuth from South Latitude is 29.64 degrees North Longitude is 82.39 degrees West Azimuth 23.67 4.44 -15.51 Azimuth (light) 293.67 274.44 254.49 Sun (fc) 10000 10000 10000
37
38
Figure 4.1: First floor plan of the Atrium space, with the location of the first 5 measurement locations (A-E).
Figure 4.2: Second floor plan of the Atrium space, with the location of the second 5 measurement locations (F-J).
39
Figure 4.3: Third floor plan of the Atrium space, with the location of the third 5 measurement locations (K-O). 4.1.1 Interior Space Figure 4.4 presents the total average of measurements from the CGMs using an Interior solution at the three different time frames 11:00 am, 12:00 pm, and 1:00pm, compared to the actual measurements collected from the atrium space. As shown in the graph the lines are very close except at points K, L and M, and this is due to the presence of direct sunlight. Figure 4.5 through 4.7 shows the Lighting analysis of the CGM simulated at the three time frames. It is apparent that the left side of the image has more red to it, since it is introduced to direct sunlight, also it is clear how the red starts to fade away when time changes showing that the sun is moving away from that section. Figure 4.8 through 4.13 presents a comparison of Lightscapes Raytracing images and pictures from the actual space.
40
Daylight Measurements of Interior Solution (All Time Frames)
7000
6000
5000 Measurements (fc)
4000
3000
2000
1000
0 Lightscape
A 82
B 103 143
C 116 157
D 108 175
E 145 213
F 212 347
G 316 403
H 357 480
I 271 357
O 362 640
204 3619 3705 3676 286 300 5967 6267 6067 527
Actual Space 113
Locations
Figure 4.4: The total average of measurements from the CGMs using an Interior solution at the three different time frames.
Figure 4.5: Light Analysis for the Atrium Space simulated as an Interior Space at 11:00am.
41
Figure 4.6: Light Analysis for the Atrium Space simulated as an Interior Space at 12:00pm.
Figure 4.7: Light Analysis for the Atrium Space simulated as an Interior Space at 1:00pm.
42
Figure 4.8: Digital image from the Actual Atrium Space-1
Figure 4.9: Digital image generated by Lightscapes Raytracing Using Interior Simulation -1.
Figure 4.11: Digital image generated by Lightscapes Raytracing Using Interior Simulation -2.
43
Figure 4.13: Digital image generated by Lightscapes Raytracing Using Interior Simulation -3
4.1.2 Exterior-Interior Space Figure 4.14 presents the same information as in Figure 4.4, but these measurements are for the Exterior-Interior solution. It is apparent that the behavior of the data is the same as in the Interior solution, where points K, L, and M are further apart from the actual measurements. Figure 4.15 4.17 shows the Lighting analysis of the CGM simulated at the three time frames. Still the left side has more red to it than the whole model, however the overall illuminance level is lower than that of the Interior solution. Figure 4.18 through 4.23 presents a comparison of Lightscapes Raytracing images and pictures from the actual space.
44
Daylight Measurements of Exterior-Interior Solution (All Time Frames)
7000 6000
5000 Measurements (fc) 4000 3000 2000 1000
0 Lightscape
A 65
B 80 143
C 89 157
D 72
E 72
F 93
G 127
H 148 480
I 125 357
O 216 640
119 3508 3569 3572 197 300 5967 6267 6067 527
Actual Space 113
175 213
347 403
Locations
Figure 4.14: The total average of measurements from the CGMs using an ExteriorInterior solution at the three different time frames.
Figure 4.15: Light Analysis for the Atrium Space simulated as an Exterior-Interior Space at 11:00am.
45
Figure 4.16: Light Analysis for the Atrium Space simulated as an Exterior-Interior Space at 12:00pm.
Figure 4.17: Light Analysis for the Atrium Space simulated as an Exterior- Interior Space at 1:00pm.
46
Figure 4.19: Digital image generated by Lightscapes Raytracing Using ExteriorInterior Simulation -1.
47
4.2 Discussion The differences in daylighting illuminance measurement between Lightscapes CGM and the actual atrium space ranged from seventeen percent (17%) and thirty five percent (35%) to forty nine percent (49%) and seventy percent (70%) of the measurements from the actual atrium space, in cases of Interior Space and ExteriorInterior Space solutions respectively. Proving what was discovered from the pilot study that a simple model simulated as an Interior Space provides closer measurements to the
48 actual space. According to Kaufman and Howard [17] in their illuminance values for general types of activates in interiors, the range of illuminance measurements for any specific tasks was always in a ratio of 1:2. Example: simple orientation for short temporary visits is five to ten footcandles of general lighting throughout the space. This was almost the same range provided from Lightscapes illuminance data recorded when compared to the actual space. In order to calculate a correction coefficient factor to assist in predicting daylighting illuminance in a space in the preliminary design phase, a regression analysis was performed. The results showed that there is a linear relationship between daylight illuminance measurements from Lightscapes simulation and measurements from the actual space (Figure 4.24).
Figure: 4.24: Best fit line from the regression analysis, showing the final equation at the top of the graph.
49 Since any daylighting illuminance measurements recorded from either Lightscapes models or the actual space is a value greater than zero, it was suggested that the intercept should be removed from the equation because it lies on the negative side of the Y-axis. Therefore to approximately present illuminance measurements that resemble measurements from the actual space, the X-coefficient was used as a correction coefficient factor for the daylight illuminance measurements recorded from Lightscape.
CHAPTER 5 SUMMARY AND CONCLUSION 5.1 Conclusion Electrical lighting is considered one of the largest energy consumers in buildings. Therefore, daylighting was proposed as an energy conservation strategy [11]. However, the obstacle in utilizing daylighting is a simple predicting method that would offer preliminarily calculations of daylight illuminance and in turn would facilitate energy conservation. Lightscape was used in this research to test one of the daylighting design tools for daylight prediction, in comparison with an actual space. To perform this test, the pilot study was conducted on a classroom at the University of Florida, followed by the actual research preformed on the GRU building. The pilot study was conducted to test and structure the testing procedures, plus to test two factors that could affect daylighting illuminance measurements from Lightscape. The factors were Daylight Parameters and Model Complexity. Concluding the pilot study with the following statement: Simplified computer generated models simulated as an Interior Space in Lightscape can more closely predict daylighting illuminance measurements of the actual space, than both moderate and detailed models. Advancing the pilot study to the next step, the actual research was conducted on a more complicated interior environment than the pilot study classroom. Utilizing findings from the pilot study, tests were performed on the Gainesville Regional Utilities atrium space to calculate a correction coefficient factor that could be used in early design processes. However, in this research only two features of the Daylight Parameter factor
50
51 was tested , a) Interior Solution, and b) Exterior-Interior Solution. Comparing both CGMs with the actual atrium space, it was proven that the Interior Solution first level in the Daylight Parameters provided similar daylighting illuminance measurements to the actual atrium space measurements. On the other hand, the problem with the Interior Solution is the lack of any daylighting measurements outside the CGM. The ExteriorInterior Solution provided fairly accurate daylighting measurements outside the model. Therefore, it was discovered that simulating a model in Lightscape using the Interior Solution Daylight Parameter, provides interior illuminance measurements closer to those of the actual space, and provides no exterior illuminance measurements. Simulating a model using Exterior-Interior Solution Daylight Parameters provides exterior illuminance measurements that are closer to measurements of the actual spaces exterior light conditions. However, the Exterior-Interior parameters interior measurements are not very close to the actual spaces interior illuminance measurements. To calculate the correction coefficient factor for Lightscapes illuminance measurements, a regression analysis was performed. The analysis of multiple conditions showed that there is a linear relationship between illuminance measurements from Lightscapes model and the actual space (see Figure 4.24). Concluding the actual research with: Lightscape is an accurate daylightingpredicting tool. When using a simple computer generated model to replicate the interior environment, a correction coefficient factor of 1.67 is recommended to better estimate the actual daylighting illuminance in the interior space under consideration.
52 5.2 Recommendations Recommendations are as follows: 5.2.1 Preparing The Model (.dwg): When using CAD packages to construct three-dimensional models to be used in Lightscape: Use basic three-dimensional modeling techniques in CAD, it is also recommended to minimize the amount of surfaces in generating the CGM. After the model is completed, it is required to explode all elements of the CGM in an axonometric view. In case Architectural desktop is used, the CGM has to be exploded twice. However it is advised to use basic three-dimensional techniques instead of Architectural desktop tools, since it reduces time when simulated in Lightscape. 5.2.2 Lighting Preparation and Solution files (.lp - .ls): When processing surfaces in the preparation files: Surfaces that are not going to be used in the simulation or are not going to appear in portraying the interior environment, should either be set as nonoccluding and non-reflecting or deleted surfaces as: planes between floors in a multi-story space, roofing incase of interior simulationsetc. It is recommended to use the DOS-based program in Lightscape LSRAD, to calculate the radiosity in the model, since it saves time when running the simulation.
53 In using raytracing or rendering the model in Lightscape, it is recommended to also use the DOS-based programs LSRAY, and LSRENDER. 5.3 Further Research Conclusions from this research suggest that performing similar tests on different architectural environments using Lightscape would refine our knowledge about the accuracy and application of the correction coefficient factor. Also further tests should be conducted to compare Lightscape daylighting illuminance information with mathematical calculations to verify Lightscapes position among other daylight-predicting tools. Conducting such tests would provide the design community with simple and accurate methods for predicting and calculating daylighting measurements for their designs. Having useful and accurate daylight prediction tools will help encourage the design profession to integrate daylighting in interior lighting solutions. Furthermore, the improved and increased use of daylight in lighting design will also to reduce energy consumption and may improve user productivity.
APPENDIX A ACTUAL RESEARCH DATA A.1 Interior Space A.1.1 Tables and Graphs
Table A.1: Daylight illuminance measurements recorded from the CGM using an Interior Space Solution, compared to the Actual Space measurements.
Time 11:00 11:00 11:00 11:00 11:00 11:00 11:00 11:00 11:00 11:00 11:00 11:00 11:00 11:00 11:00 12:00 12:00 12:00 12:00 12:00 12:00 12:00 12:00 12:00 Locations A B C D E F G H I J K L M N O A B C D E F G H I Lightscape 82 103 116 109 145 188 296 350 259 196 3559 3597 3558 293 382 81 102 115 107 144 218 320 354 272 Actual Space 130 160 180 175 200 300 380 480 320 300 5900 6000 5800 580 720 110 140 150 170 200 340 410 480 350 Deviation -37% -36% -36% -38% -28% -37% -22% -27% -19% -35% -40% -40% -39% -49% -47% -27% -23% -37% -28% -36% -22% -26% -22% -32%
54
55 Table A.1 Continued

Time 12:00 12:00 12:00 12:00 12:00 12:00 13:00 13:00 13:00 13:00 13:00 13:00 13:00 13:00 13:00 13:00 13:00 13:00 13:00 13:00 13:00 Locations J K L M N O A B C D E F G H I J K L M N O Lightscape 204 3744 3787 3761 283 358 82 103 116 108 145 230 332 366 282 213 3554 3730 3710 283 345 Actual Space 300 6000 6400 6000 500 680 100 130 140 180 240 400 420 480 400 300 6000 6400 6400 500 520 Deviation -38% -41% -37% -43% -47% -18% -17% -40% -40% -43% -21% -24% -30% -29% -41% -42% -42% -43% -34% -43% -34%
Table A.2: The total average of measurements from the CGMs using an Interior solution at the three different time frames.
Locations A B C D E F G H I J K L M N O Lightscape 82 103 116 108 145 212 316 357 271 204 3619 3705 3676 286 362 Actual Space 113 143 157 175 213 347 403 480 357 300 5967 6267 6067 527 640
56
Daylight Measurements of Interior Solution (11:00am)
7000
6000
4000
3000
2000
1000
0 Lightscape
A 82
B 103 160
C 116 180
D 109 175
E 145 200
F 188 300
G 296 380
H 350 480
I 259 320
O 382 720
196 3559 3597 3558 293 300 5900 6000 5800 580
Actual Space 130
Locations
Figure A.1: The Daylighting illuminance measurements from the CGMs using an Interior solution at 11:00am.
Daylight Measurements of Interior Solution (12:00pm)
7000
6000
4000
3000
2000
1000
0 Lightscape Actual Space
A 81 110
B 102 140
C 115 150
D 107 170
E 144 200
F 218 340
G 320 410
H 354 480 Locations
I 272 350
O 358 680
204 3744 3787 3761 283 300 6000 6400 6000 500
Figure A.2: The Daylighting illuminance measurements from the CGMs using an Interior solution at 12:00pm.
57
Daylight Measurements of Interior Solution (1:00pm)
7000
6000
4000
3000
2000
1000
A 82 100
B 103 130
C 116 140
D 108 180
E 145 240
F 230 400
G 332 420
H 366 480 Locations
I 282 400
O 345 520
213 3554 3730 3710 283 300 6000 6400 6400 500
Figure A.3: The Daylighting illuminance measurements from the CGMs using an Interior solution at 1:00pm.
Daylight Measurements of Interior Solution (All Time Frames)
7000
6000
4000
3000
2000
1000
A 82 113
B 103 143
C 116 157
D 108 175
E 145 213
F 212 347
G 316 403
H 357 480 Locations
I 271 357
O 362 640
204 3619 3705 3676 286 300 5967 6267 6067 527
Figure A.4: The total average of measurements from the CGMs using an Interior solution at the three different time frames.
58 A.1.2 Statistical Analysis
Figure A.5: Results of the Regression Analysis.
Table A.3: Results of the Regression Analysis. Simple linear regression results: Independent variable: Lightscape Dependent variable: Actual Sample size: 45 Correlation coefficient: 0.9993 Estimate of sigma: 91.29659 Parameter Intercept Lightscape Parameter Intercept Lightscape Estimate -35.390076 1.6729156 Pval 0.0175 0 Std. Err. 16.25328 0.00982819 DF 43 43 Tstat -2.1774113 170.21605
59 A.2 Exterior-Interior Space A.2.1 Tables and Graphs
Table A.4: Daylight illuminance measurements recorded from the CGM using an Exterior-Interior Space Solution, compared to the Actual Space measurements.
Time 11:00 11:00 11:00 11:00 11:00 11:00 11:00 11:00 11:00 11:00 11:00 11:00 11:00 11:00 11:00 12:00 12:00 12:00 12:00 12:00 12:00 12:00 12:00 12:00 12:00 12:00 12:00 12:00 12:00 12:00 13:00 Locations A B C D E F G H I J K L M N O A B C D E F G H I J K L M N O A Lightscape 64 80 90 72 72 93 127 148 127 119 3401 3455 3462 199 213 65 81 88 72 74 95 127 150 126 120 3595 3656 3662 195 218 65 Actual Space 130 160 180 175 200 300 380 480 320 300 5900 6000 5800 580 720 110 140 150 170 200 340 410 480 350 300 6000 6400 6000 500 680 100 Deviation -51% -50% -50% -59% -64% -69% -67% -69% -60% -60% -42% -42% -40% -66% -70% -41% -42% -41% -58% -63% -72% -69% -69% -64% -60% -40% -43% -39% -61% -68% -35%
60 Table A.4 Continued

Time 13:00 13:00 13:00 13:00 13:00 13:00 13:00 13:00 13:00 13:00 13:00 13:00 13:00 13:00 Locations B C D E F G H I J K L M N O Lightscape 78 88 72 71 90 126 147 121 117 3527 3595 3592 197 218 Actual Space 130 140 180 240 400 420 480 400 300 6000 6400 6400 500 520 Deviation -40% -37% -60% -70% -78% -70% -69% -70% -61% -41% -44% -44% -61% -58%
Table A.5: The total average of measurements from the CGMs using an Exterior-Interior solution at the three different time frames.
Locations A B C D E F G H I J K L M N O Lightscape 65 80 89 72 72 93 127 148 125 119 3508 3569 3572 197 216 Actual Space 113 143 157 175 213 347 403 480 357 300 5967 6267 6067 527 640
61
Daylight Measurements of Exterior-Interior Solution (11:00am)
7000
6000
4000
3000
2000
1000
A 64 130
B 80 160
C 90 180
D 72 175
E 72 200
F 93 300
G 127 380
H 148 480
I 127 320
O 213 720
119 3401 3455 3462 199 300 5900 6000 5800 580
Locations
Figure A.6: The Daylighting illuminance measurements from the CGMs using an Exterior-Interior solution at 11:00am.
Daylight Measurements of Exterior-Interior Solution (12:00pm)
7000
6000
4000
3000
2000
1000
A 65 110
B 81 140
C 88 150
D 72 170
E 74 200
F 95 340
G 127 410
H 150 480 Locations
I 126 350
O 218 680
120 3595 3656 3662 195 300 6000 6400 6000 500
Figure A.7: The Daylighting illuminance measurements from the CGMs using an Exterior-Interior solution at 12:00pm.
62
Daylight Measurements of Exterior-Interior Solution (1:00pm)
7000
6000
4000
3000
2000
1000
A 65 100
B 78 130
C 88 140
D 72 180
E 71 240
F 90 400
G 126 420
H 147 480 Locations
I 121 400
O 218 520
117 3527 3595 3592 197 300 6000 6400 6400 500
Figure A.8: The Daylighting illuminance measurements from the CGMs using an Exterior-Interior solution at 1:00pm.
Daylight Measurements of Exterior-Interior Solution (All Time Frames)
7000
6000
4000
3000
2000
1000
A 65 113
B 80 143
C 89 157
D 72 175
E 72 213
F 93 347
G 127 403
H 148 480 Locations
I 125 357
O 216 640
119 3508 3569 3572 197 300 5967 6267 6067 527
Figure A.9: The total average of measurements from the CGMs using an ExteriorInterior solution at the three different time frames.
63 A.2.2 Statistical Analysis
Figure A.10: Results of the Regression Analysis.
Table A.6: Results of the Regression Analysis.

Simple linear regression results: Independent variable: Lightscape Dependent variable: Actual Sample size: 45 Correlation coefficient 0.9988 Estimate of sigma: 117.61958 Parameter Intercept Lightscape Estimate 123.49425 1.6848619 Std. Err. 20.309605 0.012758559 DF 43 43 Tstat 6.0805836 132.05739
Parameter Intercept Lightscape
Pval 0 0
APPENDIX B PILOT STUDY DATA B.1 Interior Space B.1.1 Simple CGM
Table B.1: Daylight illuminance measurements recorded from the Simple CGM using an Interior Space Solution, compared to the Actual Space measurements.
Time First (10:30 AM) Locations A B C D E A B C D E A B C D E Lightscape 26 27 18 10 10 26 27 21 11 11 16 17 13 8 8 Actual Space 20 21 17 15 14 22 24 18 15 14 15 16 15 14 14 Deviation 30% 29% 6% -33% -29% 19% 14% 14% -27% -21% 7% 6% -13% -43% -43%
Second (3:30 PM)
Third (6:30 PM)
Table B.2: The total average of measurements from the Simple CGM using an Interior solution at the three different time frames.
Average 10:30 3:30 6:30 Total Lightscape 18 19 12 17 Actual Space 17 19 15 17
64
65
Simple Model Averages (Exterior-Interior)
80 70 60 Measurement (fc) 53 50 47 40 30 20 10 0 Lightscape Actual Space 42
28 17 19 15
17
10:30 47 17
3:30 53 19 Time Frame
6:30 28 15
Total 42 17
Figure B.1: The Daylighting illuminance measurements from the Simple CGM using an Interior solution at 10:30am.
Simple Model 3:30 (Interior)
80 70 60 Measurement (fc) 50 40 30 20 10 0 A Lightscape Actual Space Lightscape Average Actual Space Average 26 22 19 19 B 27 24 19 19 C 21 18 19 19 Location D 11 15 19 19 E 11 14 19 19 19 19 19 19 19
Figure B.2: The Daylighting illuminance measurements from the Simple CGMs using an Interior solution at 3:30pm.
66
Simple Model 6:30 (Interior)
80 70 60 Measurement (fc) 50 40 30 20 10 0 A Lightscape Actual Space Lightscape Average Actual Space Average 16 15 12 15 B 17 16 12 15 C 13 15 12 15 Location D 8 14 12 15 E 8 14 12 15 15 12 15 12 15 12 15 12 15 12
Figure B.3: The Daylighting illuminance measurements from the Simple CGM using an Interior solution at 6:30pm.
Simple Model Averages (Interior)
80 70 60 Measurement (fc) 50 40 30 20 10 0 Lightscape Actual Space
18 17
19 19 15 12
17
10:30 18 17
6:30 12 15
Total 17 17
Figure B.4: Average of all the Data collected from the Simple Model using the Interior Space Solution Compared to the Actual Space.
67 B.1.2 Moderate CGM
Table B.3: Daylight illuminance measurements recorded from the Moderate CGM using an Interior Space Solution, compared to the Actual Space measurements.
Time First (10:30 AM) Locations A B C D E A B C D E A B C D E Lightscape 9 10 3 3 3 10 12 4 4 4 6 5 2 2 2 Actual Space 20 21 17 15 14 22 24 18 15 14 15 16 15 14 14 Deviation -55% -52% -82% -80% -79% -55% -50% -78% -73% -71% -60% -69% -87% -86% -86%
Second (3:30 PM)
Third (6:30 PM)
Table B.4: The total average of measurements from the Moderate CGM using an Interior solution at the three different time frames.
68
Moderate Model 10:30 (Interior)
80 70 60 Measurement (fc) 50 40 30 20 17 10 6 0 A Lightscape Actual Space Lightscape Average Actual Space Average 9 20 6 17 B 10 21 6 17 C 3 17 6 17 Location D 3 15 6 17 E 3 14 6 17 6 6 6 6 17 17 17 17
Figure B.5: The Daylighting illuminance measurements from the Moderate CGM using an Interior solution at 10:30am.
Figure B.6: The Daylighting illuminance measurements from the Moderate CGMs using an Interior solution at 3:30pm.
69
80 70 60 Measurement (fc) 50 40 30 20 15 10 0 A Lightscape Actual Space Lightscape Average Actual Space Average 6 15 3 15 3 B 5 16 3 15 3 C 2 15 3 15 Location 3 D 2 14 3 15 3 E 2 14 3 15 3 15 15 15 15
Figure B.7: The Daylighting illuminance measurements from the Moderate CGM using an Interior solution at 6:30pm.
Moderate Model Averages (Interior)
80 70 60 Measurement (fc) 50 40 30 20 10 6 0 Lightscape Actual Space 10:30 6 17 7 3 3:30 7 19 Time Frame 6:30 3 15 5 Total 5 17
17
19 15
17
Figure B.8: Average of all the Data collected from the Moderate Model using the Interior Space Solution Compared to the Actual Space.
70 B.1.3 Detailed CGM
Table B.5: Daylight illuminance measurements recorded from the Detailed CGM using an Interior Space Solution, compared to the Actual Space measurements.
Time First (10:30 AM) Locations A B C D E A B C D E A B C D E Lightscape 7 4 2 3 2 7 4 3 3 2 4 2 2 2 1 Actual Space 20 21 17 15 14 22 24 18 15 14 15 16 15 14 14 Deviation -68% -81% -88% -80% -86% -68% -83% -86% -80% -88% -73% -88% -90% -89% -93%
Second (3:30 PM)
Third (6:30 PM)
Table B.6: The total average of measurements from the Detailed CGM using an Interior solution at the three different time frames.
71
Detailed Model 10:30 (Interior)
Figure B.9: The Daylighting illuminance measurements from the Detailed CGM using an Interior solution at 10:30am.
80 70 60 Measurement (fc) 50 40 30 20 10 4 0 A Lightscape Actual Space Lightscape Average Actual Space Average 7 22 4 19 B 4 24 4 19 C 3 18 4 19 Location D 3 15 4 19 E 2 14 4 19 4 4 4 4 19 19 19 19 19
Figure B.10: The Daylighting illuminance measurements from the Detailed CGM using an Interior solution at 3:30pm.
72
80 70 60 Measurement (fc) 50 40 30 20 15 10 0 A Lightscape Actual Space Lightscape Average Actual Space Average 4 15 2 15 2 B 2 16 2 15 2 C 2 15 2 15 Location 2 D 2 14 2 15 2 E 1 14 2 15 2 15 15 15 15
Figure B.11: The Daylighting illuminance measurements from the Detailed CGM using an Interior solution at 6:30pm.
Detailed Model Averages (Interior)
80 70 60 Measurement (fc) 50 40 30 20 10 4 0 Lightscape Actual Space 10:30 4 17 4 3:30 4 19 Time Frame 2 6:30 2 15 3 17 3 Total
17
19 15
17
Figure B.12: Average of all the Data collected from the Detailed Model using the Interior Space Solution Compared to the Actual Space.
73 B.1.4 Figures
Figure B.13: Lighting Analysis of the simple model simulated at 10:30 as an Interior Space.
Figure B.16: Lighting Analysis of the moderate model simulated at 10:30 as an Interior Space.
FigureB.17: Lighting Analysis of the moderate model simulated at 3:30 as an Interior Space.
Figure B.18: Lighting Analysis of the moderate model simulated at 6:30 as an Interior Space.
74
Figure B.19: Lighting Analysis of the detailed model simulated at 10:30 as an Interior Space.
Figure B.22: Example of Raytracing on the detailed model at 10:30, set as an Interior Solution.
FigureB.23: Example of Raytracing on the detailed model at 3:30, set as an Interior Solution.
Figure B.24: Example of Raytracing on the detailed model at 6:30, set as an Interior Solution.
75
B.2 Exterior-Interior Space B.2.1 Simple CGM
Table B.7: Daylight illuminance measurements recorded from the Simple CGM using an Exterior-Interior Space Solution, compared to the Actual Space measurements.
Time First (10:30 AM) Locations A B C D E A B C D E A B C D E Lightscape 56 70 45 26 36 77 61 51 48 28 38 26 26 32 16 Actual Space 20 21 17 15 14 22 24 18 15 14 15 16 15 14 14 Deviation 180% 233% 165% 73% 157% 250% 154% 183% 220% 100% 153% 63% 73% 129% 14%
Second (3:30 PM)
Third (6:30 PM)
Table B.8: The total average of measurements from the Simple CGM using an ExteriorInterior solution at the three different time frames.
76
Simple Model 10:30 (Exterior-Interior) 80 70 Measurement (fc) 60 50 40 30 20 10 0 Lightscape Actual Space Lightscape Average Actual Space Average A 56 15 47 17 B 70 16 47 17 C 45 15 47 17 Location D 26 14 47 17 E 36 14 47 17 17 17 17 17 17 47 47 47 47 47
Figure B.25: The Daylighting illuminance measurements from the Simple CGM using an Exterior-Interior solution at 10:30am.
Simple Model 3:30 (Exterior-Interior)
80 70 60 Measurement (fc) 50 40 30 20 10 0 Lightscape Actual Space Lightscape Average Actual Space Average 19 19 19 19 19 53 53 53 53 53
A 77 22 53 19
B 61 24 53 19
C 51 18 53 19 Location
D 48 15 53 19
E 28 14 53 19
Figure B.26: The Daylighting illuminance measurements from the Simple CGMs using an Exterior-Interior solution at 3:30pm.
77
Simple Model 6:30 (Exterior-Interior)
Figure B.27: The Daylighting illuminance measurements from the Simple CGM using an Exterior-Interior solution at 6:30pm.
80 70 60 Measurement (fc) 53 50 47 40 30 20 10 0 Lightscape Actual Space 42
28 17 19 15
17
10:30 47 17
6:30 28 15
Total 42 17
Figure B.28: Average of all the Data collected from the Simple Model using the ExteriorInterior Space Solution Compared to the Actual Space.
78 B.2.2 Moderate CGM
Table B.9: Daylight illuminance measurements recorded from the Moderate CGM using an Exterior-Interior Space Solution, compared to the Actual Space measurements.
Time First (10:30 AM) Locations A B C D E A B C D E A B C D E Lightscape 28 41 24 22 31 45 28 27 39 17 21 12 11 19 18 Actual Space 20 21 17 15 14 22 24 18 15 14 15 16 15 14 14 Deviation 40% 95% 41% 47% 121% 105% 17% 50% 160% 21% 40% -25% -27% 36% 29%
Second (3:30 PM)
Third (6:30 PM)
Table B.10: The total average of measurements from the Moderate CGM using an Exterior-Interior solution at the three different time frames.
79
Moderate Model 10:30 (Exterior-Interior)
Figure B.29: The Daylighting illuminance measurements from the Moderate CGM using an Exterior-Interior solution at 10:30am.
80 70 60 Measurement (fc) 50 40 30 20 10 0 Lightscape Actual Space Lightscape Average Actual Space Average 31 19 31 19 31 19 31 19 31 19
A 45 22 31 19
B 28 24 31 19
C 27 18 31 19 Location
D 39 15 31 19
E 17 14 31 19
Figure B.30: The Daylighting illuminance measurements from the Moderate CGMs using an Exterior-Interior solution at 3:30pm.
80
Figure B.31: The Daylighting illuminance measurements from the Moderate CGM using an Exterior-Interior solution at 6:30pm.
Moderate Model Averages (Exterior-Interior)
80 70 60 50 40 30 20 10 0 Lightscape Actual Space 31 26 17 19 16 15 17
Measurement
29
10:30 29 17
6:30 16 15
Total 26 17
Figure B.31: Average of all the Data collected from the Moderate Model using the Exterior-Interior Space Solution Compared to the Actual Space.
81 B.2.3 Detailed CGM
Table B.11: Daylight illuminance measurements recorded from the Detailed CGM using an Exterior-Interior Space Solution, compared to the Actual Space measurements.
Time First (10:30 AM) Locations A B C D E A B C D E A B C D E Lightscape 24 28 21 22 21 33 16 20 31 11 17 5 10 16 5 Actual Space 20 21 17 15 14 22 24 18 15 14 15 16 15 14 14 Deviation 20% 33% 24% 47% 50% 50% -33% 11% 107% -21% 13% -69% -37% 14% -68%
Second (3:30 PM)
Third (6:30 PM)
Table B.12: The total average of measurements from the Detailed CGM using an Exterior-Interior solution at the three different time frames.
82
Detailed Model 10:30 (Exterior-Interior)
Figure B.32: The Daylighting illuminance measurements from the Detailed CGM using an Exterior-Interior solution at 10:30am.
Figure B.33: The Daylighting illuminance measurements from the Detailed CGMs using an Exterior-Interior solution at 3:30pm.
83
Figure B.34: The Daylighting illuminance measurements from the Detailed CGM using an Exterior-Interior solution at 6:30pm.
80 70 60 Measurement (fc) 53 50 47 40 30 20 10 0 10:30 Lightscape Actual Space 47 17 3:30 53 19 Time Frame 6:30 28 15 Total 42 17 42
28 17 19 15 17
Figure B.35: Average of all the Data collected from the Detailed Model using the Exterior-Interior Space Solution Compared to the Actual Space.
84 B.2.4 Figures
Figure B.36: Lighting Analysis of the simple model simulated at 10:30 as an Exterior-Interior Space.
Figure B.39: Lighting Analysis of the moderate model simulated at 10:30 as an Exterior-Interior Space.
FigureB.40: Lighting Analysis of the moderate model simulated at 3:30 as an Exterior-Interior Space.
Figure B.41: Lighting Analysis of the moderate model simulated at 6:30 as an Exterior-Interior Space.
85
Figure B.42: Lighting Analysis of the detailed model simulated at 10:30 as an Exterior-Interior Space.
Figure B.45: Example of Raytracing on the detailed model at 10:30, set as an ExteriorInterior Solution.
FigureB.46: Example of Raytracing on the detailed model at 3:30, set as an Exterior-Interior Solution.
Figure B.47: Example of Raytracing on the detailed model at 6:30, set as an ExteriorInterior Solution.
LIST OF REFERENCES [1] Erwine, B., and Heschong, L. (2000, March). Daylighting: Healthy, Wealthy and Wise, LightForum.com. Retrieved on 10/16/2001 from: http://lightforum.com/design/ALM0409.html [2] Guglielmetti, R. (2001, August). Interview with Rob Guglielmetti of Renfro Design Group, INC., CG Architect.com. Retrieved on 10/18/2001 from: http://www.cgarchitect.com/upclose/article1_RG.asp [3] Renfro, R. and Guglielmetti R. (1999), Lightscape Daylighting Accuracy Test, Renfro Design Group. [4] Erwine, B. (2000, April). Daylighting Decisions Tools for an Informed Approach, LightForum.com. Retrieved on 10/16/2001 from: http://www.lightforum.com/design/daylight1.html [5] Maunder, R. (2000). Research 401, Victoria University of Wellington New Zealand, Retrieved on 10/21/2001 from: http://www.arch.vuw.ac.nz/summer_research/ [6] Thurnauer, M. (2001). Atrium Case Study, University of Miami, Retrieved on 10/18/2001 from: http://www.muc.muohio.edu/%7Estudio/mark/html/Atrium.htm [7] Lam, W. (1977). Perception of Lighting as Form Givers for Architecture, McGraw Hill Company, New York, NY. [8] Le Corbusier (1974), Towards a New Architecture Praeger Publishers, New York, NY. [9] Benton, C. C. (1989). Daylight in the US: Trends in Technology and Design, 2nd European Conference on Architecture in Paris, Paris, France. [10] Selkowitz, S. and Griffith, J. (March 1986), Effective Daylighting in Buildings, Lighting Design and Application, Vol. 16, No. 3. p. 34-37. [11] Sun, R. (1986). Evaluating the Use of a 1:12 Scale Model in Predicting Daylight Distribution in a Room, Master Thesis, University of Florida. [12] Hirons, F. (November 1920). The use of Scale Models, Pencil Points, Vol.1, No. 6. p. 5-8.
86
87 [13] Ward, G. and Rubinstein, F. (1988). A New Technique for Computer Simulation of Illuminated Spaces, Journal of IES, Vol. 17, No.1. p. 80-91 [14] Lightscape Technology Overview (2001). Retrieved on 10/28/2000 from: http://www.lightscape.com/overview/body_technology.asp [15] Gallardo, A. (2001, August). Interview with Arnold Gallardo, Visual Content Creator. Retrieved on 10/20/2001 from: http://www.cgarchitect.com/upclose/article1_AG.asp [16] Sun Angles (2000, January), Sun Design.com. Retrieved on 11/10/2000 from: http://www.sundesign.com/sunangles/index.html [17] Kaufman, J. and Howard, H. (1981), IES Lighting Handbook Application Volume, IESNA, New York, NY.
BIOGRAPHICAL SKETCH Mohamed Adel Samy Shalaby was born on May 5th, 1975, in Cairo, Egypt. He completed his bachelors degree in architectural engineering from Cairo University, Faculty of Engineering in Cairo, Egypt, on June 1998. He joined the Department of Interior Design at the University of Florida in fall 1999. He worked as a teaching assistant during his masters degree. His research interest includes daylighting design tools, and three dimensional computer programs.
88
Introduction to Gradient Ramp maps, 3D maps and blend materials

Gradient ramp The Gradient Ramp is a 2D map similar to the Gradient map, however, the Gradient ramp can be shaded more than 3 and there are several different types of effects that this map can produce . The Gradient ramp can produce unbelivable effects when it is mixed with ifself and other maps. It's also easier to use gradient ramp tiling instead of using 2d maps. There is not much explanation required for Gradient ramps, so just take a look at the example pictures below.
Gradient ramp: linear type
Single Linear
Tiling Linear
Example picture of a linear ramp type: This louvre example uses a linear gradient ramp on the diffuse and bump map channels. Gradient ramp: box type
Single Box
Tiling Box
Example picture of the box type: assigned to bump and diffuse map channels
Gradient ramp: radial type
Single Radial
Tiling Radial
Example picture of the radial type: used as an opacity map Gradient ramp: sweep type
Single Sweep
Tiling Sweep
Example picture of the sweep type: using sweep gradient ramp on the specular level map
Gradient ramp: Mapped type The Mapped type is used for adding colors (or sub-maps) on a black and white map. The different values on a black and white map are replaced with the color on the gradient bar as the end result.
Example picture of mapped type masked by a black and white brick map. The shades of black are replaced with blue on the bar and the white is replaced with orange.
The end result of the gradient ramp: mapped type
Example of mapped type, colorful mosaic tile Tips: Basically, I use gradient ramp as a mask map for a 2d mix map or blend material. It's an easier way to change the color of a map by using gradient ramp under a mix map rather than putting the color in gradient ramp itself (unless you have more than 2 color in your map).
Change the colors in the gradient ramp map itself.
Change the colors under mix map is easier
3D maps Noise, dent, celular, specle, splatter and etc. I will skip defining these 3D maps as I assume that you have had some basic experience with them already. We simply use a single 3D map as a bump map, diffuse map etc. The effect of a material that has more than two of these 3D maps with itself or other 2d map ie. gradient ramp, mixed together looks much better.
The example above shows a stone wall done with pure 3d cellular maps in several layers assigned to the diffuse, specular and bump maps. The rest of the materials on other objects are done with gradient ramp with 3D maps.
Another example with all materials created without any bitmaps. The rice is a cellular map and the meat is gradient ramp, and the plate uses noise and cellular maps.
Blend material The Blend material is a mix of two independent materials with different shaders (can also be the same). The border of the two materials are specified by a black and white map that we call a "mask map". The first material in the blend slot refesr to the black color in the mask map and the second material refers to white. Both materials are overlaid in an area of the mask map which contains grey. I have a few examples which are practical to architectural rendering. Example 1 : Creating a grid ceiling with a blend material. From the picture, the material is composed of a grid area and diffuse light (fluorescent) which are specified by a mask map.
1. Mask map
I use a gradient ramp in black and white shade and a gradient type of box with 1x1 mirror tiling. The black in the mask map refers to grid ceiling area which is material 1 and the white is fluorescent area (material 2).
This is how the mask map look like. 2. Material 1
Material 1 (grid ceiling) Go into material 1, which is thegrid ceiling. I use gradient ramp with box type again as a diffuse and bump map.
We also need to adjust the tiling of the grid which is now 3x3 mirror tiling to match with the mask map (1x1 mirror).
If the tiling numbers are diffirent from this sample, when you create a material on your own, the offset distance may also be changed to match with the tiling as well. The example below shows a mismatched position of thetiling ,caused by the wrong
adjustment on offset distance
3. Material 2
Material 2, fluorescent Material 2 is fluorescent that is easy to create. I just change its self illumination parameter to 100 or color it white. If you want more detail on this fluorescent, you may add a gradient ramp map as its self illumination map.
I use 2 layers of gradient ramp to create this self illumination effect.
Grid Ceiling Chart
The ceiling in the image below is done with a blend material.
Example 2: Wet floor Now that we have covered the blend material and using 2d gradient ramps, I hope you understand how to use them in your projects. The following example is for using 3d maps for a blend material that mostly needs no UVW map.
First, let me explain more about the legend of this material and its maps before we go onto the material chart. The wet floor consists of 3 materials, mixed together which are: 1. Dry floor Assign a 3d noise map to the the diffuse and bump channels. The shader is an OrenNayar-blinn type as I want the floor have roughness.
2. Damp floor I just copy the dry material, but change specular level to be brighter, more glossy and make the diffuse color a bit darker as the damp floor contains a bit of water that makes its color darker. I then put in a little raytrace reflection.
3. Wet floor This is the floor underneath the water. I still use the same material but change the specular and glossiness to be more glossy than the water. Remove Bump and add more reflection.
4. Mask map As there are 3 basic materials (dry, damp and wet) there will be one more blend
material that is overlayed on another blend material. I use 2 different dent maps (also 3d map) to specifiy the area of each material.
Mask 1
Mask 2
Below are 3d iamges,wet floor, using blend material
Boolean: Rubber stamp

by Montree T. (easyyong@hotmail.com) - Smoke3dStudio This technique helps to save time spent on modeling and is suitable for models which do not have too much detail. 1. Create a multi-subobject material which material ID 1 = wall and material ID 2 = window. You can use an opacity map or blend material for this material.
2. Draw a box following the window size and convert it to editable poly and assign a material ID on each polygon.
3. UVWmap and assign the material on this box
4. Create a wall and assign material ID to be 1 which is white wall and assign the same material to it.
5. Back to the window box again, edit and copy the polygon as element and it is ready to boolean
6. Boolean and you will see the window material stamp on the wall with the correct UVW coordinates.
Animated Camera Mapping

Tutorial by: Chris Bullen - Element2 This is a powerful technique to create engaging animations from static stills or context shots. This technique is widely used in cinematic productions to bring life to a static camera shot, or matte painting, by creating the illusion that one is traveling within a 3-D environment when in fact it is only a digitally enhanced 2-D image.
The best way to describe this is to imagine a slide projector with the image you wish to walk through being projected onto some basic geometry. The image will only be shown where the light is projected. This means that there would be no image where there are shadows and any object whose surfaces fall away from you (like a sphere or cylinder) would have texture stretching if you would make any exaggerated lateral movement. Once you finish this tutorial, you'll see the limitations this technique has, but you will also see the tremendous power that this technique holds. Take a look at the completed scene and play with it to see the set up! I've chosen the following royalty-free image to explain.
By examining the shot, we can break down our approach by deconstructing the scene into basic shapes. For the most part, the building can be considered to exist on one plane. We'll start here and refine our solution as we progress. The most dominant element is the portico. This consists of the deck, columns and pediment. If we were actually on site viewing this image, we would be able to take several steps to either the left or right to reveal what lies behind the columns. Unfortunately, we are not given this luxury with a still image. This is where other reference shots or our imagination will take over. We'll need to take this image into a photo manipulation program, like Photoshop, to create a modified texture to use for the building. Essentially, we need to get rid of objects in the foreground. With a little digital elbow grease we can create the following image by cloning out what we dont want. The techniques used below will be explained in another tutorial on Photoshop.
Now were ready to create the scene. You can choose to create the geometry in real world units so as to place other CG elements within the scene and have them at scale, or you can choose the shot to strictly be an artistic representation of the image. Well choose the later in this case since were only examining the technique. Start your 3D software and place the image on a 3D plane making sure to preserve the bitmaps aspect ratio. You will have to make the bitmap visible in the viewports in
order to continue. This is also where you may wish to try to scale the image in real units to create a fairly accurate scene. To do this, make one of the features (like the window) of known dimension an accurate size. For this tutorial were using 3DS MAX but this technique is applicable for any software with camera mapping functionality. You should have something similar to the following.
Next well use the image as a template to create some basic geometry. The cool thing with this technique is that you dont have to be totally accurate to create very convincing results. By switching to the front view, we create the following geometry. Note that the plane object we used to create the basic geometry will be used again for the building in the camera map.
Use common sense when creating your geometry. For example, the building in the image has a set depth for the portico. We dont know the exact measurements from just the image, but we have an idea as to how far the portico may extend. Also be aware of the axis that you may be pulling vertices around in as it is easy to get things out of square quite easily by moving in the XY plane as opposed to just the X axis. The geometry doesnt have to be precise in this stage as it will need to be fudged when we camera match. Now we need to create a camera and match the shot closely, but not necessarily exact. In order to do this, we need to place the image of the building as a viewport background. We can do this in 3DS MAX by selecting an active viewport and pressing alt-b to open the viewport background dialogue. Select File and choose the same image you used for the plane, making sure to select Match Bitmap.
Now we should have an image similar to the following ready to create a camera!
Create a target camera and match the perspective as close as possible by playing around with the camera and target positions as well as the focal length if necessary. Youll notice that things line up close but youll still need to fudge the locations more accurately.
Notice that the perspective is not 100% accurate. Dont let this bother you as it is close enough for this technique. Now look at the patio. The front view showed it to be at the right height but in the camera view, its not high enough. Correct this by moving the box up in the Z axis until its height is right. You may also require the height of the box to be taller. You will also need to adjust the size of the plane object that were using for the house to fill the frame. Notice the columns are out of whack slightly as well. This can be corrected by moving the columns laterally in the X or Y axis and adjusting their radius and taper. Other things may present themselves such as the ceiling of the portico not matching up with the transition point of the ceiling and wall in the image. Again, play around with location and size and shape of the box in 3D space until you get it matched closer. See the following image after we have fudged everything.
The set up is complete and the rest is incredibly simple. We basically have two images to apply as textures to the scene. The only difference is that instead of using traditional UVW maps, well use camera mapping. A little note before we go on. Prior to MAX 7 we had to make sure the geometry was tessellated enough to avoid texture munching. That is we needed to have enough vertices to define how accurately the texture will be mapped onto the geometry. If there werent enough vertices, the texture was shifted and fragmented and didnt appear smooth. This was fixed by applying a tessellation modifier with high settings to create enough vertices. Now in MAX 7 there is a camera map per pixel which allows you to use low poly geometry to camera map the image on. Well use the camera map per pixel (located as a material type) instead of the camera map modifier (which can be applied like any other modifier in MAX). Select a new material in the material editor and click the diffuse channel slot. This will open the Material/Map browser window where you can choose Camera Map per Pixel as the map type.
Now youre presented with the following.
Click on the Camera button and select the camera we created for matching the shot. Then click on the texture button to select the image with the Portico removed. Name this material House and make the material 100% self illuminated. Copy this material to a vacant slot and name this copied material Portico. Now in the Portico material, click the Texture button and choose the original image. Select the plane object and apply the House material. Now select everything else and apply the Portico material to it.
Were almost ready to render the image! The only thing left is to match the rendering output to the original image. In this case the original image was 480x293 pixels. When you render youll get the following which looks remarkably similar to the photo.
To create an animation, clone the camera were using for the camera map as a copy and animate the camera motion moving into the shot. This will give you the parallax and make the image look as though its a real video. Now youll need to see where the limitations of this technique exist. Start by tweaking the dolly animation to move slightly side to side. Youll see that there is a range at which this technique holds its own, but if you deviate too far away youll see some texture stretching and anomalies from the process. For the most part, if this is to be the background of an animated scene, those discrepancies would vanish since theyre not the focus of the video. If you took more time you could remove any of the discrepancies by adding more detail or maps where needed. Try going through this tutorial 2-3 times (not in a row) to get familiar with the process. Then try it on one of your own photos. After this, you should be comfortable enough to do it quite efficiently. The image and scene for this tutorial took only 15 minutes to create (creating the Photoshop images for textures and setting the scene from start to finish) which gives you quite the bang for the buck! DOWNLOAD SAMPLES FILES
Railings in 5 Minutes
by Montree T. (easyyong@hotmail.com) - Smoke3dStudio Imagine that we are drawing a building with a big area and open atriums going though each level and we also need to draw the railings. The basic way is just to create an object and clone them using spacing tool or array, but many times this can be a bit of a headache and confusing. I have come up with a simple and easy method to solve this problem using our basic knowledge of editable poly. 1. Suppose the atrium opening shape looks like the spline in the picture. We begin with drawing a contour line follow the atrium shape. Try to have as few vertices on the spline as possible. In the picture I hve drawn 8 vertices.
2. Select all segmentw on the line and divide. The length of each divided segment is approximately the same as the length of the railing panal or each piece of glass panel. Note: we may use the Normalize spline modifier instead to make sure the length of each segment is the same.
3. Change the vertices properties from smooth to corner for avoiding unwanted segment when we extrude in next step.
4. Extrude the line with out any cap. I make the height of extrusion about 90cm. Copy the object as a spare part and hide the copied object first.
5. Convert to editable poly.
6. Select all polygons and assign them the same ID number. (ID=2). (This is the easiest way to select polygons by ID in the following step)
7. Inset the polygons. The inset amount is suppose to be the distance between the gap in each glass panel.
8. Assign another ID number to the inset polygons.(from the picture, ID = 1)
9. Select polygons by ID. I select No2. ID (that we have assigned). You will see that this is more convenient than if we select the polygon using a window selection.
10. Delete the selected polygons. Now each glass panal stands seperately so we have the gap between each glass panal.
11. Select all polygons and extrude to give it thickness.
12. Cap all of the glass panels in border sub-object.
13. Assign a glass material and render.
14. In this step we are making glass fins. Go back to the 4th step where we have hiden a spare part. Un-hide it and convert to editable mesh or edialble poly. Select edges as shown.
15. Extrude edge in edit mesh is different from editable poly as the extruded edge in edit mesh will become 2 side with the thickness = 0.
16. Delete the polygon between each fin and now we have glass fins standing seperately.
17. I will skip how to make the railing on the glass panal as it is easy. Just use the line, outline it and extrude.
18. I have another example of drawing the railing from staircase. We just adapt from the method above. Drawing a 3D line
19. Segments divided.
20. Extrude, convert to poly.
21. Make gaps and thickness.
22. Render
23. Draw fins.
24. Draw the railing using another spare line, extrude and covert to poly.
25. Select polygon, extrude polygon using local normal type.
26. Done, Now wasn't that easy?
Modelling a Terracotta Roof

I had an idea of how to draw roof tiles on a roof with editable poly, as one of my recent works needed a close up view of the roof and needed to be more detailed than a bitmap. The model is quite light and not very complicated. 1. Create a box then delete the polygon at the top, side and botom. 2 polygons remain as shown in the picture below.
2. Now work in edge sub-object mode and insert a vertex to create one more edge as shown below. Note that the 2 vertices in the picture below must be at the same Xposition.
3. Select 2 edges for dividing in the next step.
4. Using "Connect" create 4 segments between the 2 selected edges. (You can find the connect tool when we use edge sub-object).
5. Select edges and connect them again as shown in the two images below.
6. Select edges and chamfer.
7. Go to polygon sub-object mode and select the polygons created by the chamfer as shown below.
8. Extrude selected polygons.
9. Select edges shown below and connect them with one segment.
10. Select the vertices in between the edges then move down to make the roof wavy.
11. Use a Slice plane to make the horizontal segments in the roof representing each course of tiles.
12. In the polygon sub-object mode, extrude theselected polygons. This is give the roof some thickness.
13. Extrude each row again one after another until reach the top row.
14. This is how the polygons look like when we finished.
15. Assign smoothing group to polygons.
Face UVWmap
by Montree T. (easyyong@hotmail.com) - Smoke3dStudio The following technique will be about the relation between modeling and its UVWmap. Some beginners draw their 3D model by extruding from 2D shape and just use a simple UVWmap such as plane, box or cylinder type, however this may sometimes produce incorrect UVW mapping on surfaces and the result looks incorrect.
A 3D model which is extruded from a 2d shape without any adjustment of polygons on top.
A simple planar UVWmap is assigned to the top.
The mapping coordinates are incorrect after we apply face UVWmap(face) to it. Drawing a model with good alignment of polygons on its surfaces can solve this problem before we assign face UVWmap. There are several ways to model this. 1. Standard primitive models are basically well built in terms of their polygon alignment.
The better way to draw a cylinder is with a standard primitive rather than to draw a 2d circle and extrude.
Some of the standard primitives also have an option of generating mapping coordinates.
Draw an inlay object, starting from standard primitive and delete some of its polygons. 2. Loft object
3. Spline network and patch modeling
An example technique using a spline network. 4. Editable polygons can help to create a proper and clean alignment of faces on the surfaces as well. I found this the easiest way to do this:
An example technique, using editably polygon. Apply UVWmap (face type) and assign a material only after you have a good alignment of polygons on a surface.
CAD Modeling for Cg Viz Applications

by William Alexander (wda@asapnet.net) - WDA Design Arched Stairway TUTORIAL Part#1 There has to be a Better Way There are many ways to model circular stairs. The Max platform has this capability as a creation object and it is a good tool, used many times to create the base model for a CAD application. However to get reasonable Architectural detail close up it does become problematic. 90%+ of the time both the ADT (& similar) and Max platforms serve the purpose very well, even for Visuals. This series of tutorials explores the other 10% of the time when close up or very accurate modeling is required in the workflow. All or part can be used in many situations to get an effective level of accuracy. We will be using many techniques common and not so commonly utilized to CAD Model for CG Viz Applications Related CAD & 3D Problems Bending 3D Extrusions Well in ADT and AutoCAD extrusions (solids) do not bend but are modeled bent. Using a 3d polyline as the extrusion path causes the profile to rotate when conforming to a helix form. With ADT the Railing Design Content using a custom profile works well, except alignment of the vertices is offset at profile transistions. This can be a problem in 3D applications looking like blocks poorly glued together. This can be resolved by welding vertices in the Max platform. As you can see below, custom profile added as hand railing.
http://www.cgarchitect.com/resources/tutorials/wda/tutorial4.asp (1 of 17)29/10/2007 20:46:43
Major issues begin to arise with flaring the bottom of the stairway, creating complex stair parts (volutes, up easing goosenecks...), non standard treads with nosing, under tread moldings, the list goes on. ADT's AEC Design Content is a powerful spatial visualization tool, not always the best modeling tool. At the root of all of this is geometry and maintaining the integrity of the parametric numeric data. Simply put stuff just does not bend and twist at the same time in a three-axis situation perfectly. In the real world, laminating strips on / in a jig, then shaping the material creates a complex helix form. Fortunately we do not need to work that hard, however there does not exist a script or plugin to solve all of these problems. Therefore, we will build a stairway using a number of commands and techniques. For the sake of accuracy, realism and understandable workflow we will follow the traditional steps of staircase construction. STAIR LAYOUT Determining Key Measurements We will be creating a staircase with a grand feeling to fit an existing round wall 13'5.5" radius, ending at a balustrade landing. Total rise of the stairway is 9'-0". For simplicity, each common stair tread will span a 6-degree arc segment. At an inside radius of about 9'-0", the walking path distance on the tread will be about 12.5". The entire staircase will span an 84-degree arc segment. Using 15 risers gives us 7.2" of rise per step. The rise and run of the stairs should have a grand and majestic feel when walked and almost certainly requiring a gooseneck, in the railing, to meet the upper newel post ;-}. With the basic carriage, form outlined, the specifics of general stair component centers, and placement is next. The treads will be 1" thick (5/4" nom.) with a 1 1/2" overhang past the riser and stringer, allowing room for a decorative molding underneath. The handrail will have a 3 3/4" width, and the newel posts will have a 4 1/2" square base. Everything all worked out the handrails are to be centered 4" in from their respected edges of the staircase. Reference jpgs, plan with notes and basic elevation. Click on image for a full size printable view in a new browser window
These are just a brief modeling specific reference note sheets to use if you like. Spending a few moments, up front to plan some details, will make modeling much easier. CAD LAYOUT Next we will set up some layout reference lines in AutoCAD. Start by creating a new layer for the construction lines and set to current layer. Top view, at the point 0,0,0 , create a reference center cross. Modeling in the center of space is important for many 3D programs, the farther away from the center, the model's accuracy tends to diminish. Draw a polyline from the center mark + 13'-5.5" at o degrees. This line is then parallel with the front view and the reference-starting ray for the angular measurements. Next, start a polyline from 13'-5.5",0",0" (end pt last line), right click and select arc, right click again and select center from the menu. The center of the arc will be our reference cross, select that center and project the polyline arc past 90 degrees (CCW) and click to end that segment and enter or esc to end the command. With the round wall determined next draw two circles, radius 9'-1" and 13'-4", centered on the reference cross. These are the radiuses for the stair carriage.
Draw a line to represent the first riser, 4'-3" in length, between the inner and outer carriage circles. Select the new line and start the array command. Set the array to "polar", center to 0,0, Method to "Total number of items & angle to fill. The total number is 15, and the angle to fill is 84 degrees. (total number of risers and total stair arc segment). Preview and if it looks right accept. Double check the angles formed by the new lines, they need to be 6.0000 degrees. The overall angle from first rise to last rise should be 84 degrees. The model uses this layout for its construction. If it is not accurate now you will find out later and most likely need to re-model or trouble shoot on the fly. That said, that is the basic footprint of the "common carriage" of the stairway, completed. Layout for the flare and starter step The flare consists of an arc that is tangent at the fifth riser to the common inside stairway radius. The starting step has an arc centered for the volute newel and sized for the general radius of itself. Draw a line from the reference center to the fifth riser line. Draw a 3' radius circle at the reference center. With the new circle selected, right click and choose move. The OSNAP settings need to have "intersection" checked and turned on. Select the intersection point of the line to the fifth riser and the new circle. Move the circle and crosshair to where the newest line meets the inside stair arc and fifth riser. The two arcs now have a common point with the same intersecting line being tangent to both arcs, thus a smooth point of transition.
Draw a circle, 1'0" radius, tangent to the first polyline created and 1'-6" from the inside carriage radius. The placement & size of the circle will be the round of the starting step. Now we have construction lines, that outline the basic footprint, of the staircase. Extruding & Modifying Solids to model from. As with the actual construction of a staircase jigs and forms are the easiest way to make them. Extruding of the basic footprint for a jig that the actual model is made from. Outline the first common tread with a closed polyline using arc segments over the radius, ends overlaying the circles of the staircase footprint. Using shift select grab the mid points of the tread polyline-arc segment and drag them to intersect with the staircase circles.
Create a new layer named "stair_solid" set to current, Switch views to SW isometric. Using the extrude CommandDraw>Solids>Extrude or from the Solids toolbar>extrude button or type extrude in the command line. Select the tread polyline and in the command line type 7.2", hit enter.
Using the ARRAY command select, the new solid and the reference cross as the center. The rest will be identical to the array done to the tread/riser polylines. Turn off the construction line layer Select all but the first tread block>copy with base point>Choose a point on the bottom>Paste>insert to the point directly above the base point.
Continue to stack the tread blocks to form the shape of the stair. You can also extrude faces, copy paste & array, what ever gets the blocks stacked up. Check the total rise and count of risers to make sure everything matches the specific key measurements. Use the union command, select all of the blocks and hit enter.
Flare and Starting step form. Turn on the construction lines layer and make it the current drawing layer. Start by using a closed polyline trace over the tangent circle, snapping to intersection's end points, to make the footprint for the flare only. Adjust the arc segment if needed Create a new layer named "flare solid" and set it to current. Extrude the new polyline to 7.2" Using the copy with base point method, stack the flare solid up 4 times. Use the slice tool, select the second stacked block. Define the slicing plane by clicking on 3 of the points of the second risers face, select the side we want to retain. Repeat this process for the other two flare blocks. Extrude the starter step circle to 7.2" Slice the bottom block on the zy-axis or using points on the extruded circle's center yx-plane. Make one solid by using the Union command, selecting all of the flare blocks & extruded circle for the operation. Do not union with the common stair.
These two solids now make up the "cage" / framework / jig for creation and assembly of the staircase components. We have already in a geometry sense "proven" the measurements, making what can be a lengthy trigonometric calculation into a matter of drawing some lines for reference. Creating The Stringers 3d Face modeling of a Helix Shape Once again looking at face modeling or using solids, however there is a problem. Any time a material is on an incline & bent (points are rotated around an axis) it is difficult to CG model with a specific profile. Take two pieces of paper, fold them in half (very stiff edge), and align them, folded edge to folded edge, in the same plane & on the same incline. Now rotate the pieces at the common center point of the folded edges, keeping them parallel to the ground. The end points no longer touch and worse they are no longer in the same plane. The ADT Design Content Railing's offset vertices are an example of this relationship. It is possible to extrude per tread a rectangle or better curved closed polyline to make the finish stringers for the staircase. Slice the top and bottom to the pitch of the common stair rise & run, slice (miter) rotate and stack them up. However, because of the corner vertices not being in the same plane the top surfaces are not continuous / smooth. Using the slice command and three-point slice plane, is one solution. A lot of
work and twice as many faces as needed. For a smooth look these solids need to be narrow slices of the 2 stringers @ 3 pcs per rise, 3 unseen faces per rise a quick 90 additional faces outright. If imported into a 3D program with a low ACIS tolerance this could be as high as 3000 faces. With that in mind face modeling becomes very efficient. Smoothing in Max, for example, is fairly easy and does not require the shape to have a dense mesh for good results. Face Modeling the Stringers We need to start by adding some additional lines to create snap points ( vertices) for the face mesh. + SW view port, top UCSII view, new layer-stringer-layout, draw 4 polylines or lines, from the inside edges of the first "common step" at the corners, 1" towards the center of the stair. Use the OSNAP function for the proper angle of the line aligning with the 6-degree increment edges. 1" is the thickness of the finish stringer on this side.
Next, select the new line of the first riser on the ground plane. Select copy with base point, select the top corner endpoint of the common stair solid, directly above the line. Paste the copied selection at the bottom corner of the common stair solid that is coplanar with the copied line.
Create a new layer; Flared-Stringer, set to current Using the 3dface command, create polygon and tri based faces. Remember the previous tutorial about normal orientation. When selecting the vertex (endpoint) selections of the individual faces, use CCW for the normal facing (visible) and CW for the normal backfacing (not visible), directions
Turn Off the layer: stringer-Layout Create a block from the newly created faces. Choose the base point as shown in the diagram above, Convert selection to block and name the block "Inside_common_Strg" Use the array command to array the "Inside_common_Strg" block by the same method used previously. Select the first four blocks and erase. The rest of the blocks use the move command and just the base point to place them in the correct z axis alignment
Using basic geometry and trigonometric relationships, with no calculations, we now have a "mesh" object that accurately models the helix form of the inside finish stringer. Only unfortunate part is we still need to check our work. An error within the
accuracy setting whether system measure (inches), precision, selection and the OSNAP can cause the "appearance" of coplanar vertices. Only to find out at the top of the stringer the blocks is off noticeably caused by an unnoticed offset of vertices by .05" multiplied 15 times=.75". For those of that use the beloved Imperial system of measurement it is advisable to use the CAD generic or engineering units of measure. 16ths (.0625) are not as accurate as .0001 increments of any measure. The Flared Stringer Each step segment changes in both dimension and rotation there for each segment is different and the repetition of a block does not work. The length of the edges of the faces are longer then the common stringer, causing the faceting of the smoothed object to be different and more noticeable. Each step segment needs modeling individually and with a "mid" set of vertices to aid smoothing. We are using the jig (stair solid) to define the outside limits. This means the jig needs to be off to fully see the newly modeled stringer in a shading mode. The explanation; you don't want to have to snap to the flare arc center, extend a series of congruent arcs past a snap point that lie with the plane of the riser. Set the current layer to "flare- construction" Make sure the "midpoint" function is selected in the OSNAP settings. In a similar fashion to the previous stringer, extend construction lines into the stair jig solid. This time add a set aligned with the midpoints. You may need a reference mark / points to fit the stringer mesh to the starter stair. Now is a good time, if your modeling needs require it, although to avoid tunnel carpal syndrome it's not covered in this tutorial.
Select the new flare stringer-construction lines. Copy with base point and choose the midpoint of the 5th riser. Paste the copied selection at the mid point first lower edge of the common stringer mesh/block, again paste the selection at the end point directly below the last past point. Now we have the snap points for the flare stair faces.
Create the faces for each step segment, as done previously, on a new layer "flarestringer". The flared portion of the finish stringer is finished, subdivided mesh for better smoothing, and aligned with the common inside stringer mesh.
The Outside Stringer We will create the outside stringer. Methodology will be similar to the creation of the inside stringer. We need to add to the faces to the stair segments, similar to the flared stringer, because a single face projects too far onto the tread. This stringer needs to stop base molding top-bottom and project vertically above the tread nosing. We will use geometry and CAD functions to avoid lengthy calculations similar to a stair builder or master finish carpenter. Set additional angles for the "Polar Settings" >3.oo for the middle face lines & vertices >6.00 for the common riser angle >-1.5 (358.50) for the reference lines to create the base molding stop face. At the bottom of the first riser, on layer-stringer-layout, draw a polyline at 0 degrees 1.5 inches long & one directly above it-top of riser. This is the width of this stringer. Select the bottom polyline and array using the reference cross-center -6 degrees 1
time. This array line is a reference for the common stair pitch for creation of the molding stop cut. Draw 2 3d polylines from the riser top and bottom edge points. Draw 2 additional polylines from the riser endpoints to the mid points of the previous polys. The mid points of the new lines define the edge of the molding stop face cut. Switch the UCSII to "front". Starting from the mid point of the first tread draw a polyline 90 degrees (vertically) 3.2" long. 3.2"= 1/2 of total rise or the rise at middle of a tread. This polyline is the reference edge for the faces of the stringer segment.
Optional- Rotate the new polylines about a world z-axis defined by the vertical edge of the first riser, using the "reference" function of the rotate command. Aligning the mid point (bottom line) to intersect with the outside stair circle / arc. Finish filling in the poly construction lines, using the polar functions to set the correct angles of the lines and lengths to 1.5". Switch the UCSII to front Select the top lines and copy with base point
Right click and select paste. Hover over the base-piont, setting up the paste's reference point, and move 90 degrees polar and 2.5" in distance, hit enter. We now have our reference frame for the faces of the outside stringer.
Switch UCSII to "top" & create a new layer; outside-stringer, set it to current Create the new faces for the stringer step segment, then draw a new block using the lower left top face endpoint as the base point. Finish creating faces for the extended stringer and base molding stop cut. Select the new block and array typical of the previous step segment arrays. Move the blocks into the proper vertical z-axis alignment using the blocks base points.
The finished stringer meshes rendered in viz/max
Parting Thoughts Complex Stair Modeling The stairway used AutoCAD for modeling. Any CAD program that has the ability to create faces in 3D can accomplish this, including Viz/Max & others using this methodology. For example the max platform you would; use the array tool, create faces using the construction lines, if needed create groups instead of blocks, extrude faces for the stair solid and use the 3d snap tool. AutoCAD just has more precision built in up front due to its nature. Viz/max has the capabilities but just not as obvious due to its free form nature and general use. The tutorial is detailed and apparently a very complex series of steps. It is not that complex in practice, compared to some other options for modeling any part of the staircase, this method has greater control, accuracy, and efficiency. With familiarity of the commands and confidence in the geometry, this part of modeling the stair can take 20 minutes maybe even less. The intent of the tutorial is to emphasize and reinforce the basic concepts. The handrail modeling is challenging enough to explain with out having to deal with explanations of the array, specific angles, geometric relationships and so on. The general approach of modeling the handrail uses the same methodology but has a few new factors to consider. Modeling of the handrails using Cad Modeling for 3dviz Applications is next. Cheers WDA
Face Modeling
by William Alexander (wda@asapnet.net) - WDA Design
Why Face Model in CAD Exclusive face modeling in Cad requires knowledge of 3D application/s being used. Many people feel that CAD is the best way to model. In CG arch Viz, CAD is the native information you receive from the client. While others feel, the 3D software modeling is more efficient. ADT has a mixture and compatibility of both CAD and 3D. The intent series of tutorials will be to explore this gray area in between. One must know how to model with faces to effectively clean model in either or both. Therefore, we will start with face modeling basics. Exclusive Face / Polygon Modeling An Ornamental Ceiling We will start by creating a room in ADT, typical of what you might receive from an architect for example. For those metric oriented: Interior wall dimensions based on standard square ceiling grid of 6x6 with a perimeter for an ornamental crown.
Before we, start modeling let us look at face and polygon definitions. A face is typically an area defined by 3 points or vertices. The face will have 3 edges. A polygon is typically an area defined by 2 faces and 4 vertices as a minimum usually 4 sides or boundary type edges. Polygonal Modeling is: primarily the use of quad polys because of the smoothing and editing characteristics, however tri shaped polys are
utilized. There are different ways of viewing & controlling a polygon. The most typical is by the boarder or boundary edges, the polygon may have many vertices and thus face edges contained in the area set as defined by the polygon. This is the way, the Max platform, deals with polygons. A 3dface in ADT can be created as a three-sided face or four-sided polygon. Sounds redundant but depending upon how you will use the 3D software it does make a difference. A face can define a mesh, a combination of faces or once again can be part of a polygon that makes up a polygonal mesh. A mesh and polygonal mesh have different inherent display, smoothing, and editing characteristics making the choice of how you model related to how your model will render. To sum it up, faces are the individual pieces that make up any surface but specifically a mesh surface. Polygonal modeling defines pairs of tri faces primarily as quads (rectangle), then using a tri poly to fill in where needed to close the polygonal mesh. For our scene, we will be oriented towards polygonal modeling. A four sided 3Dface being used in ADT to create the surfaces. Surfaces have only one side in virtual reality; that is to say, the software only understands them, to have any effect on visibility or light from one side. Yes, you can add a two-sided material, but now the software has to calculate both sides and some like Lightscape gives the surface no understanding except opacity (no light calcs). This surface property is the "normal". Changing it is very easy in the Max platform, not so always so easy on the ADT/Cad non-Viz platform. Make sure the UCS II is set to "WORLD" or "TOP" and the view is SW isometric. Next, we will create two faces using the 3dface AutoCAD command. The simple way is to just type in "3dface" in the command line or use the drop down menu DRAW>SURFACES> 3D FACE. Draw one polygonal face 2'x2' in a clock wise (CW) motion and another counter clock wise (CCW). In ADT, the faces are two sided; the normal however is created by the clockwise counter clockwise motion. The CCW motion creates the normal facing the viewer of the SW isometric view and backfacing (away) for the CW direction of creation. The normal is only view dependant when the 3Dface is drawn. After that, it will not change unless edited. It is good practice to always draw in a uniform direction when creating faces for a surface. It is no fun to find a checkerboard of normals when it is time to render.
Now it's time to actually make something. Create; 1 new layer called C-tiles-color 30, 1 called C-crown-color 40 and 1 called CLine-color 24 (Construction Lines). Set the C-Line (construction line) layer to current and draw a line or polyline from one top outside corner of the room diagonally to another. The mid point of this line defines the center of the room and where we start the ceiling grid of 3dfaces. Because we are modeling a ceiling looking down into the room the normals for our surface need to be backfacing from our view and normal to the interior view. So... select the 3dface drawn CW, change the faces layer to C-tiles, right click>copy with base point>select upper left vertex as the base point. Move the curser out over the diagonal line, right click>paste- the insertion point is the upper left corner, move the curser until the mid point is highlighted and left click to paste.
If you would want the center of the tile on the rooms center, you can use a diagonal line or poly to define the center (copy base point) as we did with the center of the room. To finish filling in the ceiling grid you can just paste individual tiles as with the first one. Another method would be to fill in 1/4 square, select those nine tiles and use the copy with base point and paste 1/4 quadrant, then half. The array tool is useful, also. Modeling Slanted Faces The Ornamental Crown Molding A fine copper ceiling just does not look right with out an egg and dart crown or rake molding. Again, we will use the 3dface command to create a polygonal mesh. Make the USC II world or top. We need end points or vertices for reference and to snap to while creating the faces. The Polar & Object-Snap Drafting Settings on, polar set to 90 degrees or less and the Endpoint selected, snap settings. Change the current layer to C-lines, Select the polyline command, and start the poly
at the upper inside corner of the room, 8' on the Z-axis.
After starting the polyline hover the curser over the first endpoint of the ceiling tileface grid, until the "endpoint" is indicated. Then move inline with the poly start point 0 degrees on the X-axis until the polar position is indicated as Polar: <0 degrees, extension; <270. Left click and create a vertex endpoint. You can continue on using the Polar and snap or extend the curser and type 24 (inches default on my system), because the next point we need is 24" down the line. Continue this around the perimeter of the room and click on the starting point, right click and select close.
Now we have snap references for our copper crown faces. The polyline needs to be moved down so the 3dfaces are pitched somewhat like crown molding. In this case, the molding projects 3" out into the ceiling field and down 3" onto the wall field. Select the polyline, you may need to press control and cycle through the possible selections (wall and poly). With the polyline highlighted, hit enter and now it is ready for editing. With the polyline selected change the UCS II to front. This gives us a vertical x-y
working plane. Right click and select move, click on the corner vertices and move the selection down, showing the Polar to be 270 degrees. Type 3 (inches) in the command line and hit enter. The polyline is now at 7'9" on the front y-axis and 7'9" on the World Z axis. With the polyline selected, right click and select properties. The elevation data entry shows 7'9" (world UCS). You can also change the elevation of an object here by highlighting the measurement and typing in a new measurement (z Axis Coordinate).
Now create the crown molding faces. Change the current layer to Ccrown. Use the 3dface command and connect the dots, endpoints. Each face should match a ceiling tile as a rectangle. The exception being the corner faces, as these will be trapezoids. Use the creation direction based on the SW view. They will all be backfacing or the normal is oriented into the room at a 45-degree angle. A few words about mapping... The ceiling tiles can be a single image applied to each polygonal face. The crown molding because of the trapezoid shape will distort the image map to fit it's shape or other ugly things. It would be wise to put each wall's crown molding on a separate layer and label north, east.... Alternatively, create blocks by wall. This allows the use of plane based UVW mapping, rotating the UVW plane (gizmo max) to align with the face and adjustments of fit & repeat. Creating layers or blocks allows each to be properly mapped as opposed to having all four face meshes as one object, plane will project only on to face meshes and 90 degrees and the box does project poorly. When importing using layers to define objects you do not want to weld vertices in max for this type of mapping to individual faces. Although in this case, a tiling map will work just fine for both the crown and ceiling tiles. The final render of ceiling in max. Render Of The Work MAX6 & Mental Ray We are not going in depth, at this time, about importing into 3D software. Quick explanation of the rendering... Imported into max- layers to entity, no weldingceiling tile UVW map face image for bump / displacement only- Crown edited into 4 separate objects, UVW plane image (tiled) mapping for bump / displacement only- MR GI; photon & FG
As you can see, Modeling in AutoCAD can produce the same results as modeling in the max platform. For those of us "Drafting Orientated" we can quickly and inherently model for specific image aspects (dimensions). Or, even jump into PhotoShop and create a bump images for the polygon aspects used. As was done for these materials. Parting Thoughts Not so productive Cad Commands There are a couple of handy commands in AutoCAD that will make your face modeling life hard. Rotate; this rotates the material generation direction, so if you are using a directional map and rotate 90 degrees, the orientation of the image will also rotate 90 degrees. Mirror; Will flip the normals and the material generation direction for mapping left to right, for example. It will always be best to create and copy faces on the same UCS II orientation. There cannot be enough said about being methodical and very organized. Layers are the key to this in CAD. Even if you do not create objects by layers, a thoughtful layer organization will give you one more method of filtering selections, if your 3D software supports AutoCAD layers.
The next tutorial will go into working in both ADT and Viz/Max. With the basic of basic principles behind us, we can work deeper into the AEC difficulties and solutions in 3D applications. Cheers WDA
CAD Modeling for CG Viz Applications

by William Alexander (wda@asapnet.net) - WDA Design
Series Introduction Where Will This Go? Where will it go? Well the intent is to explore the gray area between Cad and 3D visualization programs. Exploration presented on a detailed level for editing and clean modeling in both formats. It sounds basic in passing, but presents many winding roads when put into concepts and practice. Segment Introduction The Work Flow The typical Architect & Cad format to CG Viz applications is the norm if used. This methodology usually locks both parties into separate software formats and perceptions of Freudian envy. The exploration of 3D editing here is of AutoCAD - Architectural Desktop and Viz/MAx. The AutoDesk Suite of products may have some advantages, not the discussion here! Never the less, the workflow illustrates the use of many CAD to Viz and combination software. We will explore in this series modeling in AutoCAD - ADT for optimal results in Viz Max. Understanding the how the AEC, Multi-View blocks, blocks, Solids and 3D Face components can be used and manipulated before entering the 3D Viz worlds. Whether you work with a CAD, CG Viz or both types of applications it may be advantages to understand what may be going on in each or other types of applications. Allowing for greater flexibility of accessing components, then being able to quickly, definitively clean and properly order complex CG Viz scenes. This segment will look at some basics related to ADT/CAD and 3D programs. Then move to light application in calculated lighting solutions, Global Illumination (GI). The use of the term GI is used very broadly here as a catchall term. A narrower view for GI is for the Photon Based lighting engines-Vray, Mental Ray, Brazil vs. the Radiosity based of Max and Lightscape. Anyhow, the GI based and even fakiosity lighting has the greatest affect on "clean Modeling". Understanding how these GI engines work is the key to proper and efficient modeling. Starting with making a model and developing it in ADT. The dwg file drawing is of this real world space. In the future, we will be adding some proposed cabinetry and a new 1/4 round window. (Change order -- For illustrative purposes only!)
THE DWG FILE The file is all AEC Doors, glazing, walls and slab. These cause the most frequent number of co-incidental faces. Additional unneeded faces add to rendering times, especially in Lightscape. Not to mention irregular shadows and light bleeding. We will start by looking at the AEC walls. The ADT advantage is flexible & multi-fold. Generation of Construction line type drawings and 3D modeling combined in AEC objects. The CG Viz disadvantage is generation of redundant coincidental faces.
AEC WALLS COMPONENTS ADT uses definable boundaries to create the lines and 3D objects of walls / AEC objects. Walls have many adjustable parameters that control the boundary definitions, in the style of the wall. Most important to our discussion here is the entity display dialog; function relating to materials and elements in Max. A Viz/Max compatible Multi-sub material applied by default to all walls. Viz has a default material/s with individual maps and materials that correspond by color to the boundaries and or components, of most AEC entities. The default multi-sub materials have sub materials ranging of order to 154+ materials. These sub materials can be applied or altered; however, it is usually more practical to create and assign one with just the number needed. To do this you need to assign a new color. The colors in AutoCAD are ordered by numbers; red=1, yellow=2, green=3..... to a defined black at 255. Each number
represents a place or slot in the multi-sub material: therefore red is the first multi-sub material, yellow second..... These numbers also directly correspond to element selection and function of editing in the max platform.
In this diagram, a default or no end cap style is applied. If not needed for line drawings, it is just as easy to edit the vertices of the wall components than having to eliminate narrow faces of a returned drywall for example. The AEC generation of 3D objects creates faces based on corner vertices. The diagram shows with orange and violet shading the co-existing faces that typically generate with normals facing each other in max; "co-incidental faces". Viz and Max have overcome some of the problems with processing these by ignoring them. However, it takes processing time to do this. In Lightscape, Mental Ray and others light energy is still processed, on these surfaces. A large scene can spend hours, needlessly processing useless information. Unneeded faces and geometry should be eliminated. Doing this in the CG Viz application is much easier. By using the entity display properties in ADT, you can make selection of these surfaces simple. The multisub material function is very useful for exporting into Lightscape from max. Both are directly related. ADT ENTITY DISPLAY Adding Usable CG Viz Controls There are many wall styles available in ADT. Using StudX .625 GWB (both sides) style is a good for both CAD aspects and 3D walls with different materials on each side. To make good use of this wall style we will change it's entity display characteristics.
Select a wall and right click in ADT. Select "Edit Wall Style". In the wall style properties select Display Properties. From the view drop down list select "Model". This display mode is what gets exported out. Ignore "Draws in current View: True or False" Select "Wall Style" from property Source List. Check the attached column "yes box" in Wall Style Row. On the bottom portion of this dialog box select "Edit Display Properties". Brings up Entity Properties dialog box. In the components Column review the boundary list for used or needed ones. If the Stud component will not show in the Viz platform turn off visibility, we will leave it on for this example. Ignore the layers. On the Boundary 1 Color swatch click. This brings up a Select Color Box. For Basic Colors 1-9 you can select a color swatch from the Standard Colors. For Colors to represent materials past 9 you can select from the full color pallet. It is usually best just to enter in a number, representing a material slot order, in the Color Text Box. Click ok apply as needed and close all entity display and style windows. Repeat changing colors and adding material number information for the rest of the 3D components. Make notes on color / number assignments for multi-sub materials. Or use the slot order numbers, of existing Multi-sub materials, for assignments in ADT.
The transformation from the first view of the dwg file to the one above takes place. Changing the display option for the model view will only effect that view and not disrupt any other views in ADT. However, it is a good idea to save this file exclusively for exporting/importing into a Max file. AEC WINDOWS & DOORS COMPONENTS AEC Windows and Doors in the "Model" view are not editable as solids in ADT. They explode to AutoCAD native 3dfaces. They are independent vertices, accumulated by the number of faces adjacent to the corner/s. This can actually add vertices, edges and thus data to a viz / max scene. It would be similar to importing dxf or 3ds files. Rarely optimal modeling, with out a lot of work. Importing the dwg file with AEC objects creates one entity-(Viz Block)- with elements by color, as described above. Each element is a solid that can be converted to and or edited as a mesh in Viz / max. Why edit a predefined object as a mesh? Below illustrates how this solid modeling creates non-realistic effects. Even with a good glass material, the reflections of the glass edges are highly visible. The refraction properties are correct; the problem is the face is co-existent with the glazing frame. Not being recessed, in the reflection and or refraction shows the environment, not the darkened pocket of a glazing frame.
The co-existent (co-incidental) faces also are a real problem with all Global Illumination (GI) rendering. Use the glass edge image for an example. Lightscape, Radiosity, Mental Ray (Photo Based) GI all will equally light the glazing frame on both sides of the glass. To overcome this a dense; mesh sampling level, or photon number works to better the lighting and shadow, however at the cost of processing time. CALCULATED ILLUMINATION Radiosity - GI Explanation& Modeling Lightscape is a good place to see a lot of what goes on with Radiosity or mesh based GI. All objects, blocks in Lightscape, become faces that combine to create a surface mesh of an object. Polygons, 4 sided surface or sum of 2 triangulated surfaces, are ideal. Keeping in mind that a square is the optimal shape and narrow long rectangles are problematic.
The images below show the parameter driven receiving meshing that Lightscape uses to apply illumination and shadow (the lack of illumination). It is polygonal in nature. The objects surface mesh will be divided by the receiving mesh, if larger than the parameters set (meshing size). The ceiling is one large polygon, divided into smaller square receiving polygons, based on the meshing parameters. The window mull, right side, shows the interaction between actual mesh size (width) and the receiving mesh divisions (height). By the x,y dimension of any given surface plane. Looking at the walls receiving mesh there is a rectangular division. Further refinement based on a triangular refinement, being complementary to the face of the wall surface as define by the lower window vertex. Continuing refinement allows for a more accurate illumination gradient, where needed. You can see a basic gradient example by the violet highlighted polygon mesh segment. This is Shadow leak, caused by the lack of illumination of the wall, transferring out in the polygon. Decreasing the meshing size will minimize the artifact and in the process dramatically increasing the time to calculate light distribution for this surface. This is a very simplified explanation of a solid and flexible Lightscape program. LIGHTSCAPE EXAMPLE
Back to the glass edges, using the basic Lightscape meshing for radiosity we will look at the window frame & glass. The example on the left demonstrates the meshing and illumination gradient. The glazing material was change to opaque. The radiosity solution calculated, after hiding the glazing panel for rendering. Notice the fairly even gradient from illumination to darkness, not a realistic or accurate illumination. Changing the radiosity, meshing can help. An optimal method is model the frame splitting this surface plane into separate meshes. The vertical edges, vertex to vertex, match the edges of the glazing solid. The example on the right shows the results. Even not being modeled as described, edge to edge, some overlapping occurred. The Lightscape parameters and calculation time were, identical for both. The gradient well lit on the illuminated side and dark where it was not. The surface mesh does have a
positive effect on radiosity illumination, of course when done correctly. MAX RADIOSITY EXAMPLE The Max Platform uses a mesh similar to Lightscape. However, it is a triangle based mesh, similar to faces-tri/s as opposed to polygons-quads of Lightscape. The triangle meshing can form squares & rectangles. More often, it creates a mesh that is more adaptive to irregular surfaces and not so perfect modeling. Overall more straight forward and simpler to control. The example below on the right shows a 1'0" meshing size and mesh. The Model is identical to the Lightscape example. The typical, rendering of the glazing object. Notice the difference of the illumination gradients. The mesh is fixed using a global size and does not refine it's shape to the shadows as Lightscape does, even more reason to use clean modeling techniques. You can very easily override the global mesh size on the object level, allowing adjustments for improving local radiosity solutions.
Looking at the window frame notice the gradient in the left image, above. Similar to the Lightscape example, but it has a curve to the distribution. This is a tri vs. quad meshing difference. Still is not correct. By splitting the face of the window frame surface, we get accurate results, right image. Clean modeling can overcome the shadow leaks and artifacts. You can see these in the highlighted and marked areas. A continuous roof slab causes an extended shadow
into the room, based on the mesh size and layout. Looking closer, the shadows above the window are triangular and directly correspond to the subdividing mesh. A smaller meshing size will help. Regathering will help a great deal. Both will add to the radisoity calculation time. Clean modeling of these surfaces can keep the calculation times close to, one directly out of ADT, with better results. Photon Based GI EXAMPLE The mental Ray rendering engine plug-in for Max6 illustrative of the photon based GI. Again, the same model and set up. The way a photon based lighting calculation is quite different from the mesh based. It is usually mesh independent, an equal opportunity illuminator. Surface properties, materials and rendering engine specific shaders, have the greatest control. With this comes speed and loss of physical correctness. Although very close to being, correct. Samples, sampling size, photon size, the numbers of photons, and energy will render higher quality results. Clean modeling may not be quite as critical, but why store lighting information buried behind visible faces.
Photon based GI applies indirect light by intersecting a photon, with a face, leaving light (energy) and bouncing light energy back into the scene. You can see the disco ball effect in the left image above. The photon parameters are quite poor, but illustrate how the energy is applied. Look at the circled areas; the photon energy
equally applied across the glazing frame at the opaque window panel (hidden for rendering). The arrows show a material property having an effect on the light energy. The high-energy sun/sky photons outside of the window reflected back out and not applied uniformly to the frame. Being caused by a high angle of incidence and bouncing off the reflective glass surface. Usually with a photon, based GI is a final or re-gathering of light energy. As with Mental Ray, final gather by itself or in conjunction with the photon mapping (distribution) of light. The right image has fairly low final gathering sample range. You can see, the additional illumination applied by the final gather process, areas affected by the photon mapping, and sampling effects (smoothing) of the final gather. The arrows point out the material shader effects at the glazing frames and the parentheses show these effects on the exterior soffit. POLYGON MODEL Lighting & Image Map Based Thinking The mesh-based radiosity uses a scalable mesh for light calculations. Why not model with a polygonal mesh to conform to lighting characteristics. Better yet, control the materials. You then have, at the time of modeling, verified polygon dimensions and aspects for editing a bitmap image to fit.
As shown above, an image (bump/displacement MR) mapped to individual polygons. The actual polygons are visible in the wireframe. An obvious problem is the rectangular polys above and below the opening. An advantage is the ability to verify the dimensions and ratio aspect of an edited image to fit in ADT. The wireframe displacement image shows how a MR displacement actually creates a rendering mesh for the "physical" polygon mesh. This kind of displacement being more efficient than having loads of vertices "physically" displaced in max and of a higher quality. WRAPPING IT UP Why Lighting Examples? Understanding how software applies lighting is the key to successful modeling. Using your understanding of the mesh parameters to build a scene, allows for quick accurate and unadjusted results from your render. Further knowledge of how to take advantage of cross platform information and workflow is critical. Having verified aspect ratios, from ADT, for an image map will make editing in Photoshop a one-stop effort. Building in, entity display functionality makes selection; mesh editing and material application seamless when brought into the Max platform. These are some of the fundamental basics for working in a single platform or cross platform modeling. This is the foundation for further tutorials. Not very hands but that will change............
Advanced Lighting Techniques Part 1

by Montree T. (easyyong@hotmail.com) - Smoke3dStudio If you wish to only use standard lights, the following method is necessary to make the scene look more realistic and detailed. In this tutorial I will show several examples to guide you though the process of simulating advanced realistic lighting. The most important thing however is to first make your own observations of lighting in photographs. (Not from any GI computer generated images.) Using standard lights we have:
Omni lights with a spherical shaped beam that emits in all directions,
Spotlights with conical or pyramidal beams,
and direct light (cylinder or cube beams). The mixture of these kinds of lights can produce an abundance of lighting and shadow effects. Note: The following techniques may produce only lighting effect, you may need to create another group of lights to cast the shadows. Two spot lights combined together to create a halogen light effect. Two cone layer effect.
A group of omnis, as shown in the picture below. Some are created to be direct illumination from cope light and some are created for simulating indirect illumination around the area.
1. Tube lights
This is based upon my own observations on a tube lights. The beam of the light looks like the picture below.
I place two free direct lights on the left and right of the cylinder to create the effect on the sides and put an omni in the middle. Use uniform scale to fit the omni's
attenuation onto the cylinder. This creates the effect of the light from the top and bottom part of the cylinder. Change the shadow density to be negative so that better area shadow are created by the light.
2. Ring Light In my previous example, I just mentioned only the far attenuation, but in this case we need to use the near attenuation to create the ring effect. I use and omni and turn on the near attenuation. The area in between the near attenuation has no light so the effect looks like a ring.
The blue circle in the picture above indicates the range of near attenuation.
3. Sphere Light The sphere light is easy to create. We just put an omni inside a sphere object (that is
light source). The light emits from every surface on the sphere as its attenuation form is as same as omni, but its shadow depends on how big the light source is, so we need to adjust the shadow's softness to follow the size of the light source.
4. Environment Light Environment light is light that is scattered around a scene. It's also considered as an area light as the light will effect in every part of the object. A good example for this is skylight in an exterior scene: "Skylight is the light from atmoshere, simulated by the reflectance of sunlight over the sky, cloud and everthing in the atmoshere. As the light is scattered and affected in every direction in atmosphere so we also considered it as a big dome of an area light in the scene in CG." As mentioned in the previous example, first we imagine the area light shape as a 3d object (that we've assumed skylight shape is like a dome) and then place standard lights perpendicular to every surface on the object. In this case I choose freespots for faking the effect by placing them around the imagined dome. The dome light also result a good soft shadow effect. 1. Create a unit freespot light and change the color to light blue and enable overshoot.
2. Array and copy (instance) the light, following a half dome shape. In the example I use lesser numbers of spotlight to reduce the rendering time. (The more lights their
are, the more realistic and detailed on shadows will be.)
3. Copy and mirror the spot light (no shadow casting) below the building. This is for green illumination from the green grass effecting on the buildings.
4. Place a sunlight to complete the scene.
An exterior scene, lit with a skydome lighting technique. Not only is skylight good for faking environment light, itis also suitable for any exterior and stand alone objects in the scene.
Advanced Lighting Techniques Part 2

by Montree T. (easyyong@hotmail.com) - Smoke3dStudio 5. Playing with light multipliers and shadow density 5.1. Negative multiplier parameter of light changes the light to be a shadow. When we create a large scene some object's shadows may not appear correctly and some may be missed. A good way to create an accent shadow on an object is to place a light at the object's shadow position and change its multiplier to be negative.
Placing a negative omni behind the teapot the effect look like a shadow. When the multiplier is negative, the color will be effected as well. For example, a green colored light will appear inverse (become dark red) when the multiplier parameter is negative and white colored light will become black after we change its parameter to be negative. 5.2. Playing with shadow density Now we are playing with positve shadow density. The postitive density (more than 1) causes the shadow color to be come brighter (with the exception of a black color shadow, the positive density will make the shadow appears darker). Sometimes when we want to create a lighting effect behind an object which has a complex shape (ie. text or free from object), the first way is to use many omis, placing them to follow the object's contour.
6. The second bounce and sub-ambient The second important thing besides the ambient (that I've mention in Part 1 of this tutorial), that will make the scene look more interesting, is indirect illumination. In the real world, there are many lights bouncing/scattering around. It's hard to imagine this and apply it to your own scenes started from scratch, unless you copy a photo or use a powerful GI plugin. We don't need to put all illumination in a scene. It's good enough if our eyes only capture a second or third bounce from the main light source ie. illumination from sunlight, illumination from a big light source and etc.
Placing omnis as a direct illumination from a diffuser.
Adding omnis to simulate 2nd light bounce makes the scene look more interesting. Around the 2nd bounce area is also a sub-ambient light in the scene. This means the objects around the omni attenuation will be affected by its color. From the picture above, notice the white chairs and ceiling affected by the light- blue light from the 2nd bounce.
Middle and right: Placing omnis as an illumination from floor to the environment

3DS Max - Tutoriais de Arquitetura

Uploaded by

Copyright:

Available Formats

You might also like

3DS Max - Tutoriais de Arquitetura

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

3DS Max - Tutoriais de Arquitetura

Uploaded by

Copyright:

Available Formats

Welcome to CGarchitect.

http://www.cgarchitect.com/resources/tutorials/misc/tutorial1.asp (1 of 17)29/10/2007 21:13:59

http://www.cgarchitect.com/resources/tutorials/misc/tutorial1.asp (2 of 17)29/10/2007 21:13:59

http://www.cgarchitect.com/resources/tutorials/misc/tutorial1.asp (3 of 17)29/10/2007 21:13:59

http://www.cgarchitect.com/resources/tutorials/misc/tutorial1.asp (4 of 17)29/10/2007 21:13:59

http://www.cgarchitect.com/resources/tutorials/misc/tutorial1.asp (5 of 17)29/10/2007 21:13:59

http://www.cgarchitect.com/resources/tutorials/misc/tutorial1.asp (6 of 17)29/10/2007 21:13:59

7. Switch to left view and repeat the step 5 and 6.

http://www.cgarchitect.com/resources/tutorials/misc/tutorial1.asp (7 of 17)29/10/2007 21:13:59

http://www.cgarchitect.com/resources/tutorials/misc/tutorial1.asp (8 of 17)29/10/2007 21:13:59

http://www.cgarchitect.com/resources/tutorials/misc/tutorial1.asp (9 of 17)29/10/2007 21:13:59

http://www.cgarchitect.com/resources/tutorials/misc/tutorial1.asp (10 of 17)29/10/2007 21:13:59

http://www.cgarchitect.com/resources/tutorials/misc/tutorial1.asp (11 of 17)29/10/2007 21:13:59

http://www.cgarchitect.com/resources/tutorials/misc/tutorial1.asp (12 of 17)29/10/2007 21:13:59

13. The spline should look something like this now:

http://www.cgarchitect.com/resources/tutorials/misc/tutorial1.asp (13 of 17)29/10/2007 21:13:59

http://www.cgarchitect.com/resources/tutorials/misc/tutorial1.asp (14 of 17)29/10/2007 21:13:59

http://www.cgarchitect.com/resources/tutorials/misc/tutorial1.asp (15 of 17)29/10/2007 21:13:59

http://www.cgarchitect.com/resources/tutorials/misc/tutorial1.asp (16 of 17)29/10/2007 21:13:59

http://www.cgarchitect.com/resources/tutorials/misc/tutorial1.asp (17 of 17)29/10/2007 21:13:59

Modeling a Barcelona Chair

1. Create a Plane or Quad Patch.

5. Turn to poly or convert this patch to editable poly.

http://www.cgarchitect.com/resources/tutorials/smoke3d/tutorial7.asp (2 of 5)29/10/2007 21:04:43

http://www.cgarchitect.com/resources/tutorials/smoke3d/tutorial7.asp (3 of 5)29/10/2007 21:04:43

http://www.cgarchitect.com/resources/tutorials/smoke3d/tutorial7.asp (4 of 5)29/10/2007 21:04:43

13. Use an FFD box to add a curve to the chair.

http://www.cgarchitect.com/resources/tutorials/smoke3d/tutorial7.asp (5 of 5)29/10/2007 21:04:43

http://www.cgarchitect.com/resources/tutorials/smoke3d/tutorial19.asp (1 of 6)29/10/2007 20:53:59

http://www.cgarchitect.com/resources/tutorials/smoke3d/tutorial19.asp (2 of 6)29/10/2007 20:53:59

http://www.cgarchitect.com/resources/tutorials/smoke3d/tutorial19.asp (3 of 6)29/10/2007 20:53:59

Normal reflection of flat marble floor

http://www.cgarchitect.com/resources/tutorials/smoke3d/tutorial19.asp (4 of 6)29/10/2007 20:53:59

Normal material in library

http://www.cgarchitect.com/resources/tutorials/smoke3d/tutorial19.asp (5 of 6)29/10/2007 20:53:59

http://www.cgarchitect.com/resources/tutorials/smoke3d/tutorial19.asp (6 of 6)29/10/2007 20:53:59

Material Tips and Tricks Part 3

http://www.cgarchitect.com/resources/tutorials/smoke3d/tutorial17.asp (1 of 10)29/10/2007 20:57:18

http://www.cgarchitect.com/resources/tutorials/smoke3d/tutorial17.asp (2 of 10)29/10/2007 20:57:18

http://www.cgarchitect.com/resources/tutorials/smoke3d/tutorial17.asp (3 of 10)29/10/2007 20:57:18

Shows the parameters and settings of Shadow and light.

http://www.cgarchitect.com/resources/tutorials/smoke3d/tutorial17.asp (4 of 10)29/10/2007 20:57:18

http://www.cgarchitect.com/resources/tutorials/smoke3d/tutorial17.asp (5 of 10)29/10/2007 20:57:18

Glass material legend

http://www.cgarchitect.com/resources/tutorials/smoke3d/tutorial17.asp (6 of 10)29/10/2007 20:57:18

http://www.cgarchitect.com/resources/tutorials/smoke3d/tutorial17.asp (7 of 10)29/10/2007 20:57:18

3. Masking map : splash

http://www.cgarchitect.com/resources/tutorials/smoke3d/tutorial17.asp (8 of 10)29/10/2007 20:57:18

This is roughly how the edge of the diamond looks like.

http://www.cgarchitect.com/resources/tutorials/smoke3d/tutorial17.asp (9 of 10)29/10/2007 20:57:18

Put seveal color of lights around the diamond

http://www.cgarchitect.com/resources/tutorials/smoke3d/tutorial17.asp (10 of 10)29/10/2007 20:57:18

Material Tips and Tricks Part 2

http://www.cgarchitect.com/resources/tutorials/smoke3d/tutorial16.asp (1 of 9)29/10/2007 20:57:51

2. Gradient ramp map tree for self illumination maps.

http://www.cgarchitect.com/resources/tutorials/smoke3d/tutorial16.asp (2 of 9)29/10/2007 20:57:51

2.2. Fluorescent diffuser (light box)

2.3 Fake translucency using falloff map on self illumination map

http://www.cgarchitect.com/resources/tutorials/smoke3d/tutorial16.asp (3 of 9)29/10/2007 20:57:51