stereo tutorial main page

The following techniques will show how to convert your Bryce2/3D/4 scenes into true 3D Stereo colour images (anaglyphs) . Images that will appear to float inches in front of (or behind) your monitor screen!

In fact the basic technique can be applied to any images that consist of seperate left and right eye views; be they photographs, drawings or computer generated imagery.

These are real 3D stereo images when viewed with red/blue or red/green glasses, they consist of seperately rendered left and right eye (Camera) views of your Bryce scene. The two images are then merged into a single image such that when seen through the glasses the left eye will see only the red image and the right eye the blue image. Because of the offset between the left/right images the brain interpets the 'depth' giving the viewer the impression of a true 3D image.

There are various methods that can be used to create stereo pairs, if you have access to Adobe Photoshop or Paintshop Pro you will get more control over the left/right seperation and blue/red hue but a little more work is involved.

If you don't have access to either of these programs then the third method below works well, is easy to follow and reduces the workload if you have a lot of stereo pairs to produce.

Firstly we will create the seperate images that will form the 'Stereo Pair', I will then give you three options of converting these images into a stereo 'Anaglyph'.

Creating the Left and Right images.

For initial experiments you may find it better to turn off textures and set the 'Document size' to a small value, say 320x200. This will let you make test runs fairly quickly, remember for each stereo image you will have to render the scene TWICE. So don't start with that mega scene file that takes 3hours to render!

With your chosen scene setup in Bryce, render and save the scene as your 'left-eye' view. fig2

Now move ONLY the camera a small ammount to the right, without zooming or tilting.

The best way is to grab the handle as shown in fig1 this ensures the camera will only move left/right and not be raised/lowered/tilted.The ammount to move is dependent on how far the nearest object is to the camera .If you move too far you won't be able to merge the two combined images comfortably and probably give yourself eyestrain in the process. Remember your eyes are only a few inches apart so the Bryce camera only needs to be moved a similar relatively small ammount. Check fig3 to give you some idea of the ammount of movement needed bearing in mind I've used an excessive ammount here to ensure it showed up clearly.

TIP:A handy way to keep track of how much you have moved is to turn on the 'Display Modes' option which shows the last rendered image overlayed with the current scene wireframe (fig3). This makes it simple to compare your current camera position with respect to the last rendered position.

With the camera suitably positioned now render your 'right-eye' view. It can get confusing here as moving the camera to the right causes the near scenery to move to the left ( don't believe it? try holding a pen or similar in front of your monitor, then look through each eye in turn ).When you're satisfied, render then save the scene and rendered image. We are finished with Bryce for now, we have both left and right images it is just a matter of merging them together........

How it's done

Basically all we need to do now is take the Red channel from the Left hand image and use it to replace the Red channel in the Right hand image.

If you're a wizard with Photoshop or Paintshop you may like to skip this part and just dive in and sort it out yourself.
Otherwise, one of the following options should help clarify things.

If you would like to follow the Adobe Photoshop method click here:

If you would prefer to use Paintshop Pro click here:

If you have neither of these and would like a program and method to do the leg work for you then click here:

Finally:

For best results try and avoid strongly saturated colours, especially reds, blues and greens.

It helps to have some foreground objects to enhance the depth perception.

Don't overcompress your stereo images in JPG format, I find it can introduce 'noise' into the depth field which degrades the stereo effect.

Keep experimenting and don't forget to let Joerg and myself know of any stereo images you produce. Joerg is particularly keen to hear from anyone using his program, a small price to pay I would say.