I anaglyphed (Red/Blue) Todd's images, the way Harpo did. If you happen to have a set of anaglyphic glasses, check the images at the bottom of this post. If you do not have anaglyphic glasses, they may be found in many places on-line. I've used
Rainbow Symphony
If you already have parallax images, they can be combined in an image editor (like Photoshop) by dropping one image on the red channel and the other on the green and blue channels. Most anaglyphic glasses have the red lens on the left, but I've run into a few that are reversed. So which image goes into which color channels may be determined by experiment. The same effect can be created in video apps like After Effects with a little "channel math."
Some of Todd's images coincidentally worked with no alteration through
ChromaDepth 3D glasses. ChromaDepth art can use the entire visible spectrum where red is closest and blue is furthest. Scenes with mixed color/depth mapping will still appear dimensional, but may cause some eye strain.
There are many 3D technologies, and all have their pros and cons. Anaglyph may constrain you to monochromatic scenes, while ChromaDepth constrains your color choices in a different manner. Shutter glasses give the best results overall, but few people have the equipment. I have seen full-color movies on the Web that use an anaglyphic approach (sorry, no URL), however simply wearing the glasses will mess up the color a bit.
I also remember reading about a special video card with shutter glasses that derives 3D from the OpenGL info in a game. (Again, no URL. This was a couple of years ago, at least.)
By the way, parallax images for 3D can be derived from 2D images. I've done this in Photoshop by rotoscoping the scene into separate layers, rubberstamping any overlap, then separating the layers into anaglyphic colors. Red/Blue in one offset retreats into the scene, while Blue/Red advances, and monochrome (no offset) forms a mid-ground. The advancing trick should be used sparingly.
One last word on parallax 3D art is that most sources recommend converging the cameras at a given distance, which has also been covered in this thread. I read an article by one artist, I believe he was using
Martin Hash's Animation Master, and he argued for no convergence. His arguments made sense, and his examples proved his point. Our eyes may converge on a particular object for focus, but there is no convergence in the parallax of the images seen by our eyes. Rendering images with convergence will only introduce perspective distortions that make one's eyes hurt. Some apps (like Animation Master) have a built-in anaglyphic function. I prefer to set cameras by hand because the relative scale of objects in the scene and the separation of the cameras may demand it. Lock the cameras together with a null object, and experiment to get the right separation. Resist the temptation to push the 3D effect by separating the cameras further because that only causes eye strain.
[IMG]http://img.photobucket.com/albums/v252/Metryq/Ursula3D_ana.jpg[/IMG][IMG]http://img.photobucket.com/albums/v252/Metryq/streamclose3D_ana.jpg[/IMG][IMG]http://img.photobucket.com/albums/v252/Metryq/lightbox3D_ana.jpg[/IMG][IMG]http://img.photobucket.com/albums/v252/Metryq/DJbox3D_ana.jpg[/IMG][IMG]http://img.photobucket.com/albums/v252/Metryq/balconystream3D_ana.jpg[/IMG]