Forum Discussion
Idea - How to get fast head tracking with any frame rate
Here's an idea of how to get low latency head tracking with almost any variable frame rate. Render the scene from the player's point of view with no distortion, but to a larger render target with a...
So, am I getting this right, in understanding that this correction will always take place, adjusting for motion between each frame's pre-render... err.. "camera time_instance position lock" (whatever that may really be called), and the time it is actually displayed, even should the HMD receive a bazillion frames every second?
(EDIT: I suppose it may be necessary to point out that the above was in response to what geekmaster said about Iribe's speech and tomf's catching on to it (...assuming that's what he did).
...And now that I actually clicked the link gm provided, I guess I did get it right; For every frame, also if it arrives on time: Mitigate latency by warping the rendered image to account for prediction errors, based on the latest sensor data, compared to the prediction the frame was rendered around -- heuristics on top of heuristics, one might say ;). I am also excited to hear, directly after, in the video, about not wasting processing power on rendering tons of ultimately unused frames in between screen refreshes -- never got why people seem to tie the entire simulation to the rendering that way, rather than doing it real-time (Elder Scrolls crazy physics, anyone?).
)
(EDIT: I suppose it may be necessary to point out that the above was in response to what geekmaster said about Iribe's speech and tomf's catching on to it (...assuming that's what he did).
...And now that I actually clicked the link gm provided, I guess I did get it right; For every frame, also if it arrives on time: Mitigate latency by warping the rendered image to account for prediction errors, based on the latest sensor data, compared to the prediction the frame was rendered around -- heuristics on top of heuristics, one might say ;). I am also excited to hear, directly after, in the video, about not wasting processing power on rendering tons of ultimately unused frames in between screen refreshes -- never got why people seem to tie the entire simulation to the rendering that way, rather than doing it real-time (Elder Scrolls crazy physics, anyone?).
)