What are your experiences with the SDK 1.0

I hope they would prefer not returning structs, and instead do this:

ovr_GetFovTextureSize(session, eye, fov, 1.0, &size);

So that their API bindings would be less effective at finding bugs from FFI libraries. :D But I doubt they do that either.

"cybereality" wrote:
there are a number of features only available with the consumer product.

As if there was ability to wait left to be taken away.

There is a fundamental flaw with this generation of 'VR' and until it is fixed it will persist.

In-visor eye tracking (much like TrackIR) is necessary in order to alleviate many of the basic problems with virtual reality.

Without consciously thinking about it, humans focus on only small parts of their entire field of vision, the rest is blurry or not in focus. Without real-time eye tracking VR will always give the user sensory overload or motion sickness. If the developers were to incorporate this, along with an appropriate, robust, focus and depth software along with it, they would revolutionize the industry.

This would also be less CPU/GPU intensive as only small parts of the entire field of vision would have to be fully rendered. Undoubtedly the challenge would be the seamless shift from out-of-focus to in-focus and how to incorporate it into a system to save processor/memory resources. Even if initially it taxes resources instead of saving them, it will still alleviate the problems users have with disconnection from the experience and motion sickness/depth perception. Just sticking a 3d image two inches in front of your eyes is useless without eye tracking hardware/software.

On the off-chance that someone from Oculus sees this, your welcome.

It's called Foveated Rendering, and yes they know about it. They have given speeches about it at the oculus connect.

UPDATE: In fact, here is a video from a Oculus Connect 1 Keynote by Michael Abrash talking about it. Jump to 27m 25s.

"ch0c" wrote:
Without real-time eye tracking VR will always give the user sensory overload or motion sickness.

You seem to imply that the fact that the non-foveal part of the image is rendered in perfect focus is a cause for nausea. Do you have a source for this?

"ch0c" wrote:
Just sticking a 3d image two inches in front of your eyes is useless without eye tracking hardware/software.

No offense but it sounds like you haven't tried proper VR or not with a machine powerful enough to sustain the current state of affairs which is that the entire view is in focus.

When we get foveated rendering and can render non-foveal region at lower resolution it will allow much more complex graphics but it won't magically improve something that is already running at full speed today. Also, your eyes will still be focusing on the 3D image two inches in front of your eyes.