Question Tearing vs input lag - is there a better solution?

@[S3]Kitsune
You are right, of course.
It was my mistake to call it input lag.
I meant the delay between pressing the button/ turning the wheel and the displaying of the car starting to move.
(my virtual wheel is disabled)
To the best of my knowledge vsync introduces display lag. Even if the physics and input are computed at 400fps. If I get the images later that they are supposed to be seen, then it will be disorientating to feel something in the wheel and see something else.
But maybe I'm wrong and its only a placebo for me.

 

-

So, if i understand right from this measurements, if u enable vsync you get input lag but no tearing and of u enable triple buffering with vsync, there is almost no input lag and also no screen tearing?

 

Yes, that's correct, but you have to be careful with the term Triple Buffering because sometimes developers call something else Triple Buffering although it works differently than real Triple Buffering.

 

I tried that out and is much better, no lag and no tearing! I didnt turn vsync and triple buffering in game but in nvidia profiler and use riva tuner static server to lock fps on 60

 

@bojan.bogdanov
I don't think it's a bad idea to try that. I've also had some success in some games with that setup.
It's really worth a try.

Altough, one might want to test somehow how big the delay is with vsync and triple buffering on versus both off.
If there isn't any testing bench available you could try going by the feeling: just try 1 setting...have a go... try the other... have a go...decide.

I will also try to test this tonight.

 

@bojan.bogdanov
are you 100% sure that your monitor is set for 60Hz and not 59Hz?
you can see this in device manager

 

yes, 60 Hz

 

I tested in RR and RF2 and when i lock FPS on 60 i have tearing, then with vsync on, tearing was gone but i get input lag that is noticeable, but when i turned triple buffering with vsync, lag was gone, at least i cant notice any, and also no tearing so now is much smoother, much easier to play

 

v-sync and derivatives and the confusion between input and visual lag is boring and talked to death so allow me a very slight derail here since I'm looking only for a yes or no answer (unless you want to develop a bit on the why )

Are S3 considering upping the physics (and drag ffb along) tick rate to at least 500Hz?

Reason I ask is because most wheels do that as minimum and In both AC (originally cutting the Hz in half by default) and pC2, upping the ffb to the highest possible IS noticeable. Subtle details become clearer.

 

@vlad 2 can I just ask, what monitor your using or are you using a TV?

What your describing sounds a lot like what I was getting when I was using my main 49" TV. In that case, it was the TV having an input lag of over 100ms due to the TVs dynamic motion engine, after I figured that out, I found a "game" setting on the TV which reduced the input lag to something closer to a regular monitor.

 

@D.Boon
laptop monitor (MSI GT60 0ND)

 

That's my theory out of the window then, sorry bud.

 

maybe someone with a high speed camera (60fps or more) could make a display lag test in Raceroom. to connect a button to a LED and that to a usb...the usb to the PC and in raceroom to map that button to the brakes. switch to third person view. film the monitor and the LED. push the button. count the frames between LED turning on and brake lights turning on in-game.

• no vsync, no triple buffering, no fps cap
• 60fps cap in RTSS
• 60fps cap in RTSS, vsync and triple buffering on in nvidia

 

@[S3]Kitsune
in order to find what is the best possible setting for achieving the lowest display lag in this game

 

As it turns out, this has already been done.

 

Just so you know, the triple buffering option in Nvidia Control Panel or NvidiaInspector only affects OpenGL applications, AFAIK it does absolutely nothing for DirectX apps. If you want to force triple buffering in DirectX apps, you'll need a tool like D3DOverrider. There's a couple of download links out there on the Net, but I won't be posting them here as they're all on mediafire or googledocs and the like. Just google "d3doverrider" and you should find those links.

I use it in conjunction with vsync in all sims I play and suffer no input lag whatsoever, but on the other hand I don't really notice any difference when I disable it...I have no clue how one could measure if it actually works or not.

 

Steam doesn't start with D3DOverrider launched.

 

Because the global setting was set to medium.
For steam to start it has to be on none on global profile.

 

D3DOverrider does not work (even with the high setting).
Vsync is not enabled in-game.