Questions Regarding OpenGL and 3D Vision

[AlexV]

Hello MTBS3D,

I would like to begin my first post on MTBS3D by thanking the members of this forum. I have been working on developing some stereoscopic enabled software over the last few months and the information I have found on these forums has been extremely useful. So to everyone who has taken the time to share their expertise on these forums, THANK YOU!

I am a government contractor working on developing the next generation of vision tests for optometrists/ophthalmologists in the government. We ultimately want to build a standalone box, similar to what might be used to test your vision when you get a new drivers license but with more features and instead of using slides illuminated by a back light we are going to use two very small HD screens to present computer generated stimulus. Some of the tests I have been developing are designed to test stereo acuity. In the final product with two screen we of course don’t need to use any single screen technology like Nvidia 3D Vision because we can directly control the images going to both eyes but we also want to be able to release desktop software with versions of the same tests that can be used on a single stereo display.

I currently have some VB.net code working using DirectX(SlimDX) and a GeForce 680 using the NV_STEREO_ SIGNATURE hack but we are thinking of moving to Python/OpenGL and I have a few questions regarding stereo OpenGL programming. Based on what I have read there two ways of getting Nvidia 3D vision working in OpenGL, buy a Quadro Card and use the stereo API built into OpenGL, or use DirectX/OpenGL interlope to trigger the STEREO_SIGNATURE hack as recently described by emblemparade here http://www.mtbs3d.com/phpBB/viewtopic.p ... and+openGL. I think I also saw some reports of the OpenGL stereo API working in fullscreen mode on GTX cards with the new Nvidia 314.07 -.22 drivers but it seems that general consensus on these forums is that it was likely a mistake by Nvidia and will be removed when they notice it.

So I have a few questions:

1) Are there any other known ways to get 3D vision and OpenGL to work on GTX cards (i.e. through NVAPI, or communicating directly to the IR emitter)?

2) In OpenGL's Stereo API can I directly render to the left and right eye buffers independently or can I only set two camara/eye points and render the same scene from two angles.

3) Does the same OpenGL code work for most other types of Stereo3D setups or do you have to do a lot of work to support different stereo technologies.

Thanks again!