Comment by monocasa
7 hours ago
The thing is that I'd expect foveated rendering to increase latency issues, not help them like it does for bandwidth concerns. During a lag spike you're now looking at an extremely down sampled image instead of what in non foveated rendering had been just as high quality.
Now I also wonder if an ML model could also work to help predict fovea location based on screen content and recent eye trackng data. If the eyes are reading a paragraph, you have a pretty good idea where they're going to go next for instance. That way a latency spike that delays eye tracking updates can be hidden too.
My understanding is that the foveated rendering would reduce bandwidth requirements enough that latency spikes become effectively non-existent.
We’ll see in practice - so far all hands-on reviewers said the foveated rendering worked great, with one trying to break it (move eyes quickly left right up down from edge to edge) and not being able to - the foveated rendering always being faster.
I agree latency spikes would be really annoying if they end up being like you suggest.
Enough bandwidth to absolve any latency issues over a wireless connection is not really a thing for a low latency use case like foveated rendering.
What do you do when another device on the main wifi network decides to eat 50ms of time in the channel you use for the eye tracking data return path?
I believe all communication with the dongle is on 6GHz - both the video and the return metadata.
So again, you just make sure the 6GHz band in the room is dedicated to the Frame and its dongle.
The 5GHz is for WiFi.
Pretty funny to me that you're backseat engineering Valve on this one. If it didn't have a net benefit they wouldn't have announced it as a feature yet lmao