Gysnc with vsync on or off - input lag

[Jonathan Cave]

I watched this video which has left me confised.. slightly...

[video="https://www.youtube.com/watch?v=MzHxhjcE0eQ"][/video]

Should i run my games with vsync on or off to reduce input lag?

According to this research if i run a 4k game like Witcher 3 ultra ~45 FPS i would benefit from reduced input lag with vsync off.

but isnt the point of having gsync that you can turn vysnc off and lt gsync do the work?

Games like cs:go become unplayable with gsync on and vsync off due to the input lag. i've enabled vsync via the control pannel and its improves it alot.

edit: Turning G-Sync on and V-Sync on and setting fps_max to 60 results in buttery smooth gameplay with little to no input lag.

all help appreicated!

 

[bystander]

So one final question from me,

I've established that having Gsync on and Vsync on (as nvidia intended) results in no tearing and no input lag @ 60 fps.

Why does this switch around @ 45 fps in crysis 3 ? linus hasnt yet done another video on this yet and its really bugging me
(images below)

That actually depends on the situation. If the game does not use triple buffering and you do not have multiple GPU's, that is true, but there are games which use triple buffering, either by option or default. You might also use a multi-GPU setup, which behaves like triple buffering (each GPU has a separate buffer along with the front buffer).

 

[Jonathan Cave]

The above tests were completed on a single GPU system. I understand the only setting they changed was they enabled/disabled the vsync optons in the NVCP (global settings) and tripple buffer is disabled by default.

The ONLY difference between the charts below is FPS and vsync on/off and fps.

IT MAKES NO SENSE THAT VSYNC OFF HAS LOWER LATENCY THAN WITH IT ON SPECIFICALLY AT 45 FPS ON THE SAME 144HZ MONITOR.

 

[bystander]

The above tests were completed on a single GPU system. I understand the only setting they changed was they enabled/disabled the vsync optons in the NVCP (global settings) and tripple buffer is disabled by default.

The ONLY difference between the charts below is FPS and vsync on/off and fps.

IT MAKES NO SENSE THAT VSYNC OFF HAS LOWER LATENCY THAN WITH IT ON SPECIFICALLY AT 45 FPS ON THE SAME 144HZ MONITOR.

The higher FPS showed the reverse effect as lower FPS.

There are another factors at play, and that is the engine and drivers. Nvidia does use prerendered frames by default, and the engine may be doing some odd things. Perhaps even the monitor plays a role in it.

Anyways, those charts are definitely inconsistent.

 

[boju]

Signal processing is another factor. ie Acer Pred vs Asus Swift, GtG 4ms vs 1ms. The Acer has faster signalling so the end result is the same response, not bad at all for an IPS.

 

[bystander]

After a 2nd look, I do have a theory. It is the very high FPS charts which do the reverse of what you'd expect, and that may be due to how Nvidia has disabled V-sync at your refresh rate or higher FPS. Originally, Nvidia did not offer that as an option, and perhaps their work around to add that as an option, due to AMD having it as an option, required some coding that just isn't very efficient.