Introduction

Contract

Project

Credits

Resume

 

CAVE Conversion Process

Guidelines

Examples

Conclusions

 

3D Studio – CAVE Design Guidelines

___________________________________________________________________________________________________________________________

 

This page outlines some basic guidelines for transformation of 3D studio *.3DS files into the CAVE  format. For an effective VE conversion, the following is a list of issues that you should remember when producing a model in 3D studio:

Material mapping (UVW mapping), compatible format for texture maps, scaling of image coordinates

Lighting attributes in 3D studio, Lights supported by the CAVE

Scaling of models from 3D studio to CAVE

            Translation of Coordinates from 3D studio into the CAVE coordinate system

Translation of 3D studio color into the CAVE

Optimize command in 3D studio, Techniques for Polygon Optimization

Transfer of Pre-defined shapes from 3D studio to the CAVE

Transfer of layers from 3D studio

Translation of Nurbs, splines and nurb surfaces from 3D studio to the CAVE

 

Texture conversion:

 

Textures on a 3d model are the most important attribute to creating a realistic effect. In order to view your model with proper textures as modeled in 3D studio, in the CAVE, you will need to do the following:

Texture Do’s and Don’t’s in 3D studio

 

1.       Sizing of Texture maps:

 The Texture files used in 3D studio have to be sized to be powers of 2 in order to be read into the CAVE. That is a texture file either has to be a square or 2:1

such as (256 pixels x 256 pixels ; 256 pixels x 512 pixels, etc.)

So, if you are using a 3D studio texture file and it is not in that format, you need to copy the texture file (the texture file used for bitmapping) and load into a software like Adobe photoshop. Resize the image to be power of 2. Then save in JPEG format. Then you will have to make a new material in 3D studio and reassign the mapping coordinates to the object.

See Example 1

  1. Format for texture files:

The CAVE requires either JPEG (JPG) or GIF format but then these files must also be converted to Silicon Graphics RGB (red-green-blue format). Even one file in TGA, CEL or other unacceptable format will cause the model to be rejected by the CAVE computer. Generally we use JPEG format since these are recognized and are the smallest files.

 

See Example 1

 

Converting 3D studio texture files to RGB

 

The JPEG texture files must be converted to RGB but still have the original texture maps in JPEG format in 3D studio. Nugraf is an extremely powerful translation software. It has silicon graphics RGB Bitmap Support. Nugraf V.2.1 has the ability to read and write SGI RGB format (used by the open flight convertors). When a .3DS file is exported as an openflight file, a pop-up menu comes up-

Go to the Bitmaps tag and check on “convert all bitmap files to .rgb’ 

 

 

 

 

 

This automatically converts all your bitmaps to. rgb format which is required in the CAVE.

CAUTION:  When you have the above option checked, Nugraf creates Bitmap.rgb and Bitmap.rgb.attr (texture attribute file ) files. Make sure to put these in the same directory as the .flt file on the SGI computer (CAVE simulator).

 

Incase you are not using Nugraf as a convertor:

 

To convert JPEG texture maps to RGB, we use a program written by SGI and you will need to perform the following on an SGI workstation:

 

Download the following script and place it in your path settings. This script will automatically convert all your textures to .rgb files.

 

jpgTorgb ( A script written by David Weatherly in PERL to convert all jpegs at once)

 

Here is its usage:

 

*  To use the script make sure it is copied into the same directory as your textures    on the SGI.

*  Make sure to change permissions of the file after getting it, type:

 "chmod +x  jpgTorgb "

 

Type in the following:

 

 

 

If you are not using jpgTorgb, do the following :

 

Convert your texture files from *.jpg to *.rgb using imgcopy:

"imgcopy oldfile.jpg newfile.rgb" .

For further information on imgcopy, type “man imgcopy”

 

Notes on using imgcopy:

 

  1. Nugraf supports the UVW texture mapping that 3d studio produces and translates this into the CAVE through the open flight format. This is not possible when bringing models in the CAVE using the 3dsToIv conversion program. This convertor does not support UVW mapping, so textures in the model appear to be stretched across the surface when viewed in the CAVE.

See Example 2

          

Tips and Limitations :

1. When exporting out of 3D studio, you can export only standard and multi-sub object, and only the following components of the material:

·         Diffuse, ambient, and specular color.

·         One texture map, which must be in the Diffuse channel. Use JPEG or GIF files for your maps, as these are recognized by most browsers and are generally the smallest files. Some browsers, such as the Autodesk browser, recognize other file formats. Multi/sub-object materials will export colors and textures.

            See Example 3

·         Note that texture maps slow down the browser and increase download time. Use them sparingly.

·         Shininess (but not shininess strength)

·         Opacity

 

2.  If there are multiple bitmaps on a surface, only the one on the diffuse channel gets exported during translation.

                See Example 4

3.      It is important to put all your texture files and the model files on one directory    when exporting out of 3D studio and Nugraf.

 

Translation of Lights from 3D studio to the CAVE:

 

 

The following is a list of lights available in 3D studio. When exporting out of 3D studio only target spot, omni and default ambient get exported - All 3 are supported in the CAVE.

 

 

3 D Studio Light

Exported out of 3 d Studio

Supported in CAVE (running Saranav)

Omni Light

Yes

Yes

Free Spot

No ( cannot export

         free spot lights)

 

Free Direct

 No (cannot export   

 directional lights

 

Target Spot

Yes

Yes

Target Direct

No (Cannot export directional lights)

 

Default Ambient Light

Yes

Yes

Sunlight

No (Cannot Export directional lights)

 

 

 

See “ Example 1 - Demonstrates the use of all 3d studio lights

    “ Example 2- No lights used in 3D studio

    “ Example 3-  Only an omni light used in 3D studio

    “ Example 4-Only target spot light used in 3d studio.

 

Default lighting in saranav

 

Saranav the navigation tool for the CAVE uses a default headlight and ambient light source.

The headlight source is fixed to the head of the CAVE user. By default this is a sunlight source.

The ambient component of the light source can be manipulated adjusting the r,g,b values:

-A [mbient] <r> <g> <b>

default: r = 0.4 , g= 0.4 , b= 0.3 (yellowish)

 

The lighting in savanav is not a very good source for depicting accurate light in a VE for architectural development.

Previously Dennis Cafiero has worked with the source code of savanav and removed the default  lighting. So if you use the lighting that is developed in 3D studio or Multigen  you will be able to view your model as is in the CAVE.

His version of Saranav with no lighting can be downloaded here:

saranolit

Put this in the same directory as the model file on the SGI.

 

 

 Scaling of models from 3D studio to CAVE:

 

See “ Example 1-

       “ Example 2-

            saranav  – S .0833 filename.flt.

      .0833 (1/12) is the scaling factor that converts inches to feet.

      See “ Example 3-

 

           saranav –S (0.833 x Number) filename.flt

          (where if number is =1: No change, if number = 0.3 : then 30% change, etc.)          

 

About the scale function:

-s[   ] <scale>

set the scale of the dataset using 3.281 for a dataset with meter units and 0.0833 for a dataset with inch units

default: 1.0 ( for a dataset with foot units)

 

 

 Aligning 3D studio images in the CAVE

 

The VT CAVE is a 10' cube having its origin (0.0, 0.0, 0.0) at the center of the cube.

There is an exact translation of coordinates from 3D studio to the CAVE.

If an object is made with its center at the origin (0,0,0) in 3D studio, this object will be

brought up at the origin (0,0,0) in the CAVE upon conversion.

 

See “ Example 1-

 

Now if you want to move the center of the CAVE by x, y, z in relation to the object center, type:

saranav –i x y z filename.flt

See “ Example 2-

 

BUT, If you want to change the position of the object itself in the CAVE, this will have to be done in 3D studio.

Say from (0,0,0) in 3D studio, you can move the object center to (50,50,50).This will put the object at (50,50,50) in the CAVE.

 

 

Use of color in 3D studio

 

See “ Example 1-

 

 

To see color in the exported model, it will have to be assigned to the object as a material of type standard, which has the desired color.

            Go to:  Main toolbar > Material editor

                        Tools menu > Material editor

Highlight a material patch and give it a desired color by adjusting the values of the ambient, diffuse and specular components, in the material editor window.

Assign this material to the object.

See " example2"

 

 

 

 

 

 

These standard materials, which are created by adjusting the ambient, diffuse and specular values need to be resized to the powers of 2 or converted to JPEGS.

 

Polygon management

Current hardware and software technology imposes limitations on what you can do with 3D modeling and animation if you intend it to be displayed and interacted with in the CAVE. Because you want your work to look as good as possible and also to display and move as fast as possible, you need to be aware of these limitations and be prepared to work with them.

Tips and Limitations:

 

A complete design with 5000-10,000 polygons is manageable for most computers. The LOD helper lets you manage the polygon count of the objects in design, If the model is very large the animation will be jerky caused by a delay in computer processing  You can get an idea of how many polygons are in 3D studio by model by examining the data on faces. In 3D studio Viz or Max, under

File>summary information

Examine the number of faces. A face is roughly equivalent to a polygon

 

 

 

Use of predefined shapes in 3D studio:

 

There are predefined shapes in 3D studio viz under the Create-Geometry panel

 

 

All shapes get exported from 3D studio viz as .3DS format including - all standard primitives, extended primitives, doors, windows, stairs and AEC extended objects.

 

I have found that all these solid objects will translate into the CAVE properly. All shapes and surfaces get exported out of 3D studio as a series of triangulated polygons.When the CAVE deals with shapes, it treats a shape as a series of triangles. Each triangle is considered a polygon. So when creating a smooth surface it is important to remember that the smoother the surface the more polygons that area required. A large polygon count slows down simulation.

 

Tips and Limitations:

 

Object vertices should be less than 64K in order to export out of 3D studio viz in .3Ds format.

This problem was faced when trying to export the inbuilt foliage-trees, available in 3D studio. In order to keep the vertice count low- reduce the density of the foilage.

 

Use of Layers in 3D studio:

 

One of the most useful tools in 3D studio is the use of layers. Objects can be turned on and off and selected ‘by layer’ for rendering and modeling. This can be applied to converting files to CAVE format as well.

When exporting your file as .3DS file, it acknowledges any layers that you may have. So any layers you have turned off when exporting as .3DS file, will not show up in the CAVE ,and vice-versa, all layers turned on when the .3DS file is exported will be shown in the CAVE.

 

Ideally you would want to be able to bring  a drawing file into the CAVE and turn layers on and off at will. This would allow you the greatest freedom of design and visualization. However, at this time , the CAVE cannot do this. You can navigate through your drawing but you cannot turn layers on and off.

 

Please Note:

You have to create a separate .3DS file for every variation of layers you want to see in the CAVE.

 

 

Nurbs, splines and Nurb surfaces:

 

Nurbs are not yet supported in virtual environments and the CA VE. So when you use the Nurbs command, when the model is converted, the Nurbs are turned into a series of triangulated polygons. This creates a larger polygon count slowing down the simulation.

 

A Nurb surface gets effectively converted and translated into the CAVE as a series of triangulated polygons.

See " example1 "

 

Tips and Limitations:

 

            see " example 2 "

 

      see " example 3 "

 

Translation of 3D studio Text to the CAVE:

 

Any text written in 3D studio is treated as closed nurb curves and splines. This gets successfully exported out of 3D studio as a  series of triangulated polygons, and shows up in the CAVE

See " example 1 "

 

Maintaining Clean Geometry

 

It is especially important to have a clean geometry for the CAVE.

Do not overlap any objects. Use Boolean geometry or make objects adjacent. In the CAVE the computer first renders the first object and then the second very rapidly, so it appears to flash. This is especially disturbing for the viewer. This does not appear on the personal computer in still frame rendering or animations, so one tends to overlook this aspect.

 

Use of ‘Copy’ command in 3D studio

 

When copying an object in 3D studio it is important that you create a copy and not an instance. An instance referencing the original object but changes only its location in the three dimensional space. The convertors for VE’s do not support the use of instances in 3D studio. If a model is converted using instances only the original object will show up in the model when viewed in the CAVE