8 ms·
As I understand, the issue is not so much getting the position data, but getting the data quickly (with low latency) to reduce mismatch between the rendered pos
by ealloc 13y ago
As I understand, the issue is not so much getting the position data, but getting the data quickly (with low latency) to reduce mismatch between the rendered position on screen and the user's actual position. The Rift has special 250 Hz position sensors for this purpose, and even that's not fast enough. Webcams are usually only 30Hz.
http://oculusrift-blog.com/john-carmacks-message-of-latency/682/ http://oculusrift-blog.com/john-carmacks-message-of-latency/...
- rocky1138 13y agoFair enough. What about interpolation?
- deletes 13y agoHow many frames of delay do we get then. About 8 to 9. Not really useful with realtime games.
- malandrew 13y agoWhere does most of the delay come from? Is there a way to take a "cockroach" nervous system approach where certain sensors have as direct a connection as possible to the code that updates the view of the user?
- acous 13y agoYou might be interested in the articles Carmack[1] and Michael Abrash[2] have written about VR. [1] http://www.altdevblogaday.com/2013/02/22/latency-mitigation-strategies/ http://www.altdevblogaday.com/2013/02/22/latency-mitigation-... [2] http://blogs.valvesoftware.com/abrash/latency-the-sine-qua-non-of-ar-and-vr/ http://blogs.valvesoftware.com/abrash/latency-the-sine-qua-n...
- malandrew 13y agoHow low is the latency on measuring muscle movements like the Myo does? Is it competitive with the 250 Hz position sensors?