Page 22 of 44
Re: flood's input lag measurements
Posted: 22 Mar 2015, 01:09
by MaximilianKohler
Could you keep the OP updated with the results you've been posting?
Re: flood's input lag measurements
Posted: 22 Mar 2015, 01:15
by spacediver
Re: flood's input lag measurements
Posted: 22 Mar 2015, 01:28
by MaximilianKohler
Yeah, I've seen those, but there are things in the comments of this thread that aren't in the OPs of any of the links.
Such as the mat queue mode results.
Re: flood's input lag measurements
Posted: 22 Mar 2015, 14:09
by spacediver
I think all the data is available in the google doc link
Re: flood's input lag measurements
Posted: 23 Mar 2015, 03:00
by flood
you can consider the esr thread the main thread where i post updates
Re: flood's input lag measurements
Posted: 23 Mar 2015, 19:08
by MaximilianKohler
Here is all data so far:
preliminary conclusions:
default fps_max 100 adds between 0-10ms of input lag (expected behavior) over the lag with uncapped framerate of ~2000fps
raw input doesn't affect cs1.6
Maybe I'm asking too much, but it would be nice if you would summarize each finding in text like that, as interpreting the raw data is a bit of a task.
Re: flood's input lag measurements
Posted: 24 Mar 2015, 03:53
by flood
i will eventually, but at the moment, i don't have many conclusions yet as i want to irresponsibly give a premature conclusion for anything
Re: flood's input lag measurements
Posted: 27 Mar 2015, 03:10
by Sparky
Have you tested FPS_max vs driver based framerate limits yet? I think a GPU driver based framerate limiter could achieve low input lag by spoofing backpressure instead of just making the GPU take longer, but I must admit I've got zero clue how they're actually implemented.
Frame pacing algorithms could also confound the issue.
Re: flood's input lag measurements
Posted: 27 Mar 2015, 19:57
by flood
not yet, i've been busy working on more important research
i think that csgo's fps_max is already good in the sense that it doesn't add any unnecessary input lag on top of that which is expected from lowered framerate. the only way an external cap can be better than this is if it has better timing precision
Re: flood's input lag measurements
Posted: 31 Mar 2015, 15:47
by Sparky
flood wrote:not yet, i've been busy working on more important research
i think that csgo's fps_max is already good in the sense that it doesn't add any unnecessary input lag on top of that which is expected from lowered framerate. the only way an external cap can be better than this is if it has better timing precision
I'm mainly curious about the best option for games with no internal framerate cap, and how much worse it is than an in-game cap.