Can you use a frame rate limiter with Vysnc?

zod96

Suspected BAD TRADER
Joined
Feb 23, 2008
Messages
2,709
Can you use a frame rate limiter with vysnc enabled?
 

CaptNumbNutz

Fully [H]
Joined
Apr 11, 2007
Messages
23,126
Vsync effectively limits the framerate to the refresh rate of the monitor already.
 

demowhc

Limp Gawd
Joined
Apr 30, 2011
Messages
274
Yes you can use a frame limiter with vsync, it actually improves mouse response too.
 

Samwiches

Weaksauce
Joined
Feb 5, 2012
Messages
90
If your framerate is not exactly at your refresh rate then you are not getting anything from vsync..
 

evilsofa

[H]F Junkie
Joined
Jan 1, 2007
Messages
10,078
I'm going to go out on a limb here and assume most of the posts in this thread are about Skyrim. I'll quote the Tweakguide about Skyrim and Vsync alternatives:

The two main alternatives for VSync in Skyrim are:
1. Leave VSync enabled in Skyrim (i.e. the default), and enable Triple Buffering instead using the D3DOverrider utility which comes with the free RivaTuner software. This will improve performance and can reduce mouse lag without resulting in other negative effects. Furthermore, because the FPS cap from VSync remains in place, your vertical mouse axis will not become incredibly sensitive, though it will still become sluggish at low FPS. The only issue is that Triple Buffering uses more Video RAM which may cause problems on some systems, and it may not reduce mouse lag to the same degree as simply disabling VSync. To counter this, check the other mouse-related variables in this section.
2. Disable VSync and use a FPS Limiter Mod to maintain any FPS cap you wish, though 60 or 30 FPS is recommended. This will prevent extreme vertical mouse sensitivity and physics issues, and will remove mouse lag, but there may still be some tearing.

CaptNumbNutz: Vsync increases mouse lag, and cuts your FPS down by half if your video card renders frames at less than your monitor's refresh rate, so it's worth it to do without Vsync if you can find a way to do so and also reduce tearing.

demowhc: Vsync's only benefit is to reduce tearing. Running a frame limiter without Vsync not only eliminates the mouse lag caused by Vsync, but eliminates other problems caused by Vsync while reducing tearing, so I'm not sure why you would want to run both.

Samwiches: what?
 

wonderfield

Supreme [H]ardness
Joined
Dec 11, 2011
Messages
7,396
To clarify: vertical sync eliminates tearing; a frame limiter can only reduce it.
 

cerwinvega

Limp Gawd
Joined
Jun 15, 2004
Messages
374
I've experimented quite a bit with this over the last few months, and for me, the best experience overall has been with Triple Buffered Vsync + FPS limiter (MSI Afterburner) set to refresh rate -1.

In my experience with that configuration, it seems to provide the benefits of Vsync without any perceivable input lag, and for certain games, the steady 59fps is far more enjoyable than wildly fluctuating FPS, even if overall FPS would remain very high (say, 150-300 fps swings in source games, for example).

All that said, it really is a matter of personal preference.
 

Samwiches

Weaksauce
Joined
Feb 5, 2012
Messages
90
If you are at -1FPS from your refresh rate then vsync is eliminated. All you should get is a really slow tear somewhere on your screen. Maybe you don't notice it.

And if your mouse lag is actually gone, that would be another clue.

Anyway, if you want to limit your FPS to (refresh - 1) then just do that and don't try to turn on vsync. What do you see that is different?

Another experiment you can try is limit frame rate to match refresh, leaving vsync off. You probably get a fixed screen tear that maybe scoots around a little bit. That's how it was on CRT's anyway.
 
Last edited:

cannondale06

[H]F Junkie
Joined
Nov 27, 2007
Messages
16,180
If you are at -1FPS from your refresh rate then vsync is eliminated. All you should get is a really slow tear somewhere on your screen. Maybe you don't notice it.

And if your mouse lag is actually gone, that would be another clue.

Anyway, if you want to limit your FPS to (refresh - 1) then just do that and don't try to turn on vsync. What do you see that is different?

Another experiment you can try is limit frame rate to match refresh, leaving vsync off. You probably get a fixed screen tear that maybe scoots around a little bit. That's how it was on CRT's anyway.
the amount of tearing can vary wildly from game to game. all different types of tearing can occur though any time that vsync is off no matter what the framerate is.
 

boxleitnerb

Limp Gawd
Joined
Oct 9, 2011
Messages
422
vsync nowadays doesn't mean that the framerate is cut in half. Most if not all games either have triple buffering enabled or something similar.

The limiter set to Hz-1 (when using vsync) almost eliminates input lag while you can maintain a tearing-free image. With this method I have never seen any tearing at all. Why should vsync be eliminated when using a limiter? If you have vsync on and have 38fps for instance, you also have no tearing.
 

cerwinvega

Limp Gawd
Joined
Jun 15, 2004
Messages
374
If you are at -1FPS from your refresh rate then vsync is eliminated. All you should get is a really slow tear somewhere on your screen. Maybe you don't notice it.

That doesn't make any sense, no offense. FPS can drop below refresh rate easily in a performance-challenged situation, it wouldn't lose Vsync.

Anyway, if you want to limit your FPS to (refresh - 1) then just do that and don't try to turn on vsync. What do you see that is different?

Well, screen tearing, for one.

Another experiment you can try is limit frame rate to match refresh, leaving vsync off. You probably get a fixed screen tear that maybe scoots around a little bit. That's how it was on CRT's anyway.

I think you need to actually try it before speculating any further about how you think it should work, and I don't mean that disrespectfully. You might be surprised by the results.
 

demowhc

Limp Gawd
Joined
Apr 30, 2011
Messages
274
Have you tested and confirmed this, or is this something you feel to be the case?

Try it yourself in any source engine game, set fps_max 0 in console, enable vsync with solid 60fps, feel mouse. Then set fps_max 60 or 61, 62 etc with vsync still enabled and feel the mouse again.

Same works using the fps limiter in inspector in almost any game, BC2 for example with vsync & frame cap feels far better.
 

MavericK

Zero Cool
Joined
Sep 2, 2004
Messages
31,329
I think you need to actually try it before speculating any further about how you think it should work, and I don't mean that disrespectfully. You might be surprised by the results.

What he described is what I do and his result is accurate, for me at least. With FPS limited to 60 (same as refresh) I get a screen tear that is generally within 1-2 inches of the same spot every time. Still pretty annoying. I've tried turning it down to 59 or 58 but it still seems to occur. How much does refresh rate fluctuate?

And I've tried using VSync + Triple Buffering forced via DXOverride + framerate limiter at 60, and there is still considerable mouse input lag. Maybe it varies with the game, but it makes FPS games unplayable for me still.
 

wonderfield

Supreme [H]ardness
Joined
Dec 11, 2011
Messages
7,396
Try it yourself in any source engine game, set fps_max 0 in console, enable vsync with solid 60fps, feel mouse. Then set fps_max 60 or 61, 62 etc with vsync still enabled and feel the mouse again.
That's not what I'd call a rigid scientific test.

The reason I asked is because the technique doesn't really make any sense from a technical perspective. Given the way frame limiters work, and how vertical sync works (in much the same way as an internal frame limiter works), it just doesn't make sense that setting a frame rate limiter to some arbitrary number above the refresh rate will have any effect on how perceptible input lag is.
 

cerwinvega

Limp Gawd
Joined
Jun 15, 2004
Messages
374
Well, key for me was setting VSync + Triple Buffering and the limiter at 1 frame below refresh rate, not above, i.e., 59fps limit for 60hz refresh.
 

MavericK

Zero Cool
Joined
Sep 2, 2004
Messages
31,329
I feel like I have tried that with no real result. What game(s) are you testing, out of curiosity?
 

cerwinvega

Limp Gawd
Joined
Jun 15, 2004
Messages
374
BF3, L4D2, Skyrim, Borderlands, Payday: The Heist, etc.

Some games it does not work with, Dead Island being one example, but that game felt so bad that I didn't really care to try and see why.
 

cannondale06

[H]F Junkie
Joined
Nov 27, 2007
Messages
16,180
Dead Island felt so bad that I gave up and started using the 360 controller for it. same goes for Alpha Protocol.
 

Samwiches

Weaksauce
Joined
Feb 5, 2012
Messages
90
That doesn't make any sense, no offense. FPS can drop below refresh rate easily in a performance-challenged situation, it wouldn't lose Vsync.



Well, screen tearing, for one.



I think you need to actually try it before speculating any further about how you think it should work, and I don't mean that disrespectfully. You might be surprised by the results.

Well, okay I'll try it. But this is not speculating. I've been a picky FPS player since Quakeworld.

So if my average maintained framerate is around 85 FPS with regular dips, what refresh should I start with? (120Hz display, many other modes available)

Also have plenty of games that do either hundreds of FPS with no slowdown ever, as well as some CoD games which are all capped at 90FPS to start with.
 

cerwinvega

Limp Gawd
Joined
Jun 15, 2004
Messages
374
Can't say for sure @120Hz since I still run at 60, where it seems the key is just making sure you maintain a framerate that never exceeds refresh rate -1, so 59fps for me, maybe 118-119 for you ?

Triple Buffering is also important so that you aren't stuck with multiples of refresh rate.

That said, I guess if you are regularly below refresh rate, it wouldn't matter.

EDIT: I guess for games that are always up in the stratosphere with FPS, maybe CS and the like, it wouldn't matter there either.
 
Last edited:

MavericK

Zero Cool
Joined
Sep 2, 2004
Messages
31,329
Well, I tried Vsync + Framerate limiter at 59 and it actually seemed to make a difference. I also made sure to force Triple Buffering specifically for the application (instead of Global which seems to not always work?) and it seemed to decrease the mouse lag significantly.

I could have sworn I tried this already, but maybe when I tried it before Triple Buffering wasn't being forced for some reason? Or maybe just the 59 FPS limit made a difference. Anyway, me likey. :cool:
 

cannondale06

[H]F Junkie
Joined
Nov 27, 2007
Messages
16,180
Well, I tried Vsync + Framerate limiter at 59 and it actually seemed to make a difference. I also made sure to force Triple Buffering specifically for the application (instead of Global which seems to not always work?) and it seemed to decrease the mouse lag significantly.

I could have sworn I tried this already, but maybe when I tried it before Triple Buffering wasn't being forced for some reason? Or maybe just the 59 FPS limit made a difference. Anyway, me likey. :cool:
where are you using triple buffering from?
 

MavericK

Zero Cool
Joined
Sep 2, 2004
Messages
31,329
D3DOverrider. Not sure how else you'd do it with nVidia cards (since apparently the CP + D3D doesn't work).
 

demowhc

Limp Gawd
Joined
Apr 30, 2011
Messages
274
That's not what I'd call a rigid scientific test.

The reason I asked is because the technique doesn't really make any sense from a technical perspective. Given the way frame limiters work, and how vertical sync works (in much the same way as an internal frame limiter works), it just doesn't make sense that setting a frame rate limiter to some arbitrary number above the refresh rate will have any effect on how perceptible input lag is.

I understand what you are saying but what can I tell you it works, as mentioned below you can try 59 at 60hz too but I find its not as smooth (slight jitters when strafing for example). I was skeptical too, but tried it and it does indeed work. BC2 for example shows vast improvement as do most source engine games. Portal 2 with vsync was horrid until setting max fps to 59/60/61/62. I also set frames to render ahead to 1 or 0 in driver control panel.

Well, key for me was setting VSync + Triple Buffering and the limiter at 1 frame below refresh rate, not above, i.e., 59fps limit for 60hz refresh.
 

Galvin

2[H]4U
Joined
Jan 22, 2002
Messages
2,697
Has anyone figured how using a fps limiter + vsync actually reduces input lag. We know it works, but how does it work. I've tried googl'ing and never was able to find an answer.
 

Forceman

[H]F Junkie
Joined
Aug 7, 2005
Messages
9,243
Has anyone figured how using a fps limiter + vsync actually reduces input lag. We know it works, but how does it work. I've tried googl'ing and never was able to find an answer.

Perhaps it is because the frame rate limiter is overriding the Vsync setting, and you are actually running without vsync. Intellectually, I can't see how those two things could work at the same time - if the frame rate limiter says to limit the speed to 59 FPS, and Vsync is trying to force 60 FPS, which one wins?
 

Galvin

2[H]4U
Joined
Jan 22, 2002
Messages
2,697
Perhaps it is because the frame rate limiter is overriding the Vsync setting, and you are actually running without vsync. Intellectually, I can't see how those two things could work at the same time - if the frame rate limiter says to limit the speed to 59 FPS, and Vsync is trying to force 60 FPS, which one wins?

I've already tested this in skyrim and in world of warcraft. And looked for tearing when FPS dropped below 59. Didn't see any. Yeah thats the question is how does it work while keep vsync going below 59fps. I've tested adaptive vsync and saw tearing below 60fps. So vsync + FPS limiter is not the same thing, but vsync is on 100% of the time.
 

cannondale06

[H]F Junkie
Joined
Nov 27, 2007
Messages
16,180
I've already tested this in skyrim and in world of warcraft. And looked for tearing when FPS dropped below 59. Didn't see any. Yeah thats the question is how does it work while keep vsync going below 59fps. I've tested adaptive vsync and saw tearing below 60fps. So vsync + FPS limiter is not the same thing, but vsync is on 100% of the time.
that makes no sense. if the framerate is capped at 59 then of course you will still get tearing with adaptive vsync since vsync will never kick in. if you run regular vsync then there will be no tearing at any framerate but capping the framerate at 59 while doing that will make things a little stuttery. in other words that 59 fps myth is nonsense because all you have to do is pan around slowly and you will see a little hitch at 59 fps with vsync on.
 

Galvin

2[H]4U
Joined
Jan 22, 2002
Messages
2,697
that makes no sense. if the framerate is capped at 59 then of course you will still get tearing with adaptive vsync since vsync will never kick in. if you run regular vsync then there will be no tearing at any framerate but capping the framerate at 59 while doing that will make things a little stuttery. in other words that 59 fps myth is nonsense because all you have to do is pan around slowly and you will see a little hitch at 59 fps with vsync on.

Like I've said before. I've tested adaptive vsync in WoW and saw tearing below 59fps. Now follow me closely here. Now I turn on vsync instead, turn off adaptive vsync so vsync is application controlled. Set fps limiter on to 59. Go into WoW and see no tearing below 60fps. So common sense says that if you can see tearing with adaptive and not with fps limiter + vsync. Then vsync is actually staying on all the time with the fps limiter.

You totally didn't understand what I was saying. I know how adaptive works, but the point is I can see tearing below 60fps and I don't see it with vsync + fps limiter.
 

cannondale06

[H]F Junkie
Joined
Nov 27, 2007
Messages
16,180
Like I've said before. I've tested adaptive vsync in WoW and saw tearing below 59fps. Now follow me closely here. Now I turn on vsync instead, turn off adaptive vsync so vsync is application controlled. Set fps limiter on to 59. Go into WoW and see no tearing below 60fps. So common sense says that if you can see tearing with adaptive and not with fps limiter + vsync. Then vsync is actually staying on all the time with the fps limiter.

You totally didn't understand what I was saying. I know how adaptive works, but the point is I can see tearing below 60fps and I don't see it with vsync + fps limiter.

of course there is no tearing ever if you have real vsync on. you can cap the framerate at whatever you want to but it will not make any difference because no tearing will occur anyway. if you have vsync off or are using adaptive vsync then of course there will be tearing if you have the game capped at 59fps. by itself, capping a game will not stop tearing.
 

DeathPrincess

Fully [H]
Joined
May 15, 2010
Messages
18,205
On the nvidia driver functions for frame rate limiting it has an option for a 58fps limit with the comment "for no inputlag on vsync 60hz" etc.

4j9ibt.png
 

cannondale06

[H]F Junkie
Joined
Nov 27, 2007
Messages
16,180
On the nvidia driver functions for frame rate limiting it has an option for a 58fps limit with the comment "for no inputlag on vsync 60hz" etc.

http://i49.tinypic.com/4j9ibt.png
they can call it that but capping the framerate below refresh rate with vsync on will cause a consistent little hitch when panning around in the games I tested in. removing the below refresh rate cap removed the hitching every time and turning it back on reintroduced it.
 

DeathPrincess

Fully [H]
Joined
May 15, 2010
Messages
18,205
they can call it that but capping the framerate below refresh rate with vsync on will cause a consistent little hitch when panning around in the games I tested in. removing the below refresh rate cap removed the hitching every time and turning it back on reintroduced it.

Did you try it with or without triple buffering? Just interesting that it's listed there like that. There must be something to it, I suppose, though I usually just limit at 60 or 75 and don't find that much lag.
 

Galvin

2[H]4U
Joined
Jan 22, 2002
Messages
2,697
Lag drops off a lot the higher you fps. 120fps and you probably won't notice any lag at all.

In skyrim I did very slow pans left and right looking for tears top to bottom and found none. So i'd be surprised if someone saw tearing with vsync on. If vsync is on its impossible to get a tear unless something is turning it off.

But the mystery is how does it fix input lag since vsync is the cause of it in the first place.
 
Top