Blur Busters Forums

Hello, https://blurbusters.com/howto-low-lag-vsync-on/ With this guide, you can achieve low lag input lag with -0.01 of the refresh rate. How about 1/2 1/3 and so on multiples? For example my screen refreshes at 119.982 after 40+ seconds of stability with refresh rate test. If I want to use 1/3 Vsync to lock a game to 40 FPS, what should be my framecap? 39.990? What would be the ideal approach in such a situation?

Hi, I have similar question on this forum and sadly nobody can answer me. But I made some progress to testing it and it looks like you can't divide your screen refresh rate for no tearing and low input lag.

For example my screen refresh rate is 239.713Hz and minimal fps in game I want to run is for example 100fps. So I divided 239.713 / 3 and set framelimit in RTSS for this value and yes, input lag is low but screen tearing is still visible.

So I downloaded Nvidia Pendelum Demo and set the framerate to the moving line to achieve the most non-visible tearing below 100fps. It turned out that for my 239.713Hz monitor the best option for framelimit is 93fps.

It follows that it is better to test it "by eye" than to believe the calculations.

But there is another question. What V-sync does if you turn on framelimit for above example for 93fps on 240Hz display panel. Nobody can answer this question

yamaci1775 wrote: ↑

28 Nov 2022, 17:32

Hello, https://blurbusters.com/howto-low-lag-vsync-on/ With this guide, you can achieve low lag input lag with -0.01 of the refresh rate. How about 1/2 1/3 and so on multiples? For example my screen refreshes at 119.982 after 40+ seconds of stability with refresh rate test. If I want to use 1/3 Vsync to lock a game to 40 FPS, what should be my framecap? 39.990? What would be the ideal approach in such a situation?

To my knowledge, the 1/2, 1/3, and 1/4 Refresh Rate V-SYNC settings available in Nvidia Inspector basically apply V-SYNC at a division of the current max refresh rate, hence 1/4 Refresh Rate V-SYNC at 240Hz is an effective 60HZ V-SYNC, which means, at 240Hz, each frame is repeated 4 times, so frame 1 = 1, 1, 1, 1, frame 2 = 2, 2, 2, 2, and so on.

It's basically identical to standard 60+ FPS V-SYNC at 60Hz, just at a 240Hz scanout, hence it's just evenly repeating each frame 4 times in a row within the same period, resulting in standard V-SYNC behavior, latency and all.

As for low lag V-SYNC methods, I haven't tried them with the divisional V-SYNC options, so can't speak on that directly.

CorvusCorax wrote: ↑

02 Dec 2022, 06:30

For example my screen refresh rate is 239.713Hz and minimal fps in game I want to run is for example 100fps. So I divided 239.713 / 3 and set framelimit in RTSS for this value and yes, input lag is low but screen tearing is still visible.

You can't remove tearing without a syncing method, no matter what your framerate/refresh rate ratio is set to.

As for input lag, if V-SYNC is off, input lag at the display-level is already the lowest possible, regardless of your current framerate/refresh rate ratio. Input lag only becomes a factor in this respect when you enable a syncing method.

If, however, you're saying you're experiencing tearing with a syncing method enabled, then there's simply something wrong with your configuration, since the majority of syncing methods can't tear if set up properly.

jorimt wrote: ↑

02 Dec 2022, 11:01

CorvusCorax wrote: ↑

02 Dec 2022, 06:30

For example my screen refresh rate is 239.713Hz and minimal fps in game I want to run is for example 100fps. So I divided 239.713 / 3 and set framelimit in RTSS for this value and yes, input lag is low but screen tearing is still visible.

You can't remove tearing without a syncing method, no matter what your framerate/refresh rate ratio is set to.

As for input lag, if V-SYNC is off, input lag at the display-level is already the lowest possible, regardless of your current framerate/refresh rate ratio. Input lag only becomes a factor in this respect when you enable a syncing method.

If, however, you're saying you're experiencing tearing with a syncing method enabled, then there's simply something wrong with your configuration, since the majority of syncing methods can't tear if set up properly.

I know that I can't eliminate tearing without sync, but as I wrote above you can make it less visible. For my case it is a 93fps for fixed 240Hz display panel without V-Sync. No idea why 93fps works BTW

But... I started doing some experiments with above settings and I eventually turned V-Sync on in nvcp (with framelimit in RTSS set at 93fps) on fixed 240Hz. It turned out that it works too, I have constant 93fps, no tearing, smooth frametimes on fixed 240Hz display panel and I don't feel any difference in input lag. I wonder how this works, why V-Sync behave different with framelimit. Any ideas?

CorvusCorax wrote: ↑

03 Dec 2022, 08:42

I'm replying in your existing thread that is more on topic here:
viewtopic.php?f=10&t=10700&start=10#p87438