Mathematica has great capabilities to display 3D Graphics rendered to the 2D screen. This should make it perfectly suited to use the Graphics3D system together with an NVIDIA Graphics card, a 3D capable monitor with high frame rate and a 3D shutter glass (e.g. the NVIDIA 3D Vision sets). Now if it comes down to documentation the situation is less clear. I just found a nice Visual Studio Implementation
How to control stereo-frames separately with C#? (NVIDIA 3D shutter glasses)
for getting a simple 3D application (which is actually just testing that you can render two different colors for the left and the right eye), but nothing in the Mathematica documentation or on the web. Wouldn't it be nice to have the full 3D power of Mathematica with built-in support to show the Graphics3D output on a 3D Vision system?
Any idea how to migrate the C# sample code to Mathematica and use the Graphics3D output to generate the stereo image pairs dynamically? Any other idea of implementation is of course welcome too!