V-Sync Input Lag

[EGomz3]

Does anyone know how much input lag is introduced by turning on V-Sync in the Nvidia Control Panel? I'm debatting whether to just use V-Sync or spend the extra cash for a no tearing AND fast response time experience via G-Sync. 

If there isn't an exact number of ms of lag introduced, a range of values could help too.

 

 

[Mark Kaine]

in my experience turning on vsync ingame is almost always preferable over doing it in NV settings regarding input lag.

 

As for a exact number idk, something between 10 and 100ms maybe. I didn't really measure it specifically I just know nvidia vsync is extremely laggy I never use it therefore (may depend on the generation of card, though)

[LOST TALE]

Take the refreshtime and divide it by two. That would give you the average input lag added by vsync when it occurs.

 

So 60hz = 1000ms/60= 16.67 ms = 8.88ms input lag added on average

 

If your gpu's rendering time stays below 16.67 ms then input lag added by vsync should be 0ms on a 60hz display.

 

The biggest issue isn't the input lag but the variability in it. That's what causes animations to appear to stutter. Vsync input lag when it's working properly is very unnoticeable to most people. serious input lag comes from bugs in the rendering pipeline.

 

Still. I didn't notice it until I got freesync to compare lol

[Mark Kaine]

1 hour ago, LOST TALE said:

Take the refreshtime and divide it by two. That would give you the average input lag added by vsync when it occurs.

 

So 60hz = 1000ms/60= 16.67 ms = 8.88ms input lag added on average

 

If your gpu's rendering time stays below 16.67 ms then input lag added by vsync should be 0ms on a 60hz display.

 

The biggest issue isn't the input lag but the variability in it. That's what causes animations to appear to stutter. Vsync input lag when it's working properly is very unnoticeable to most people. serious input lag comes from bugs in the rendering pipeline.

 

Still. I didn't notice it until I got freesync to compare lol

input lag isn't a bug, and there's always some input lag, usually way  more than people think too.*

 

there's also indeed several sources which add up, and that's what people notice (everything under 100 ms overall response time aka "input lag"  is usually perceived as "ok" especially when people are playing on TVs)

 

 

https://en.m.wikipedia.org/wiki/Input_lag

 

 

*

[Satan_Prometheus]

There are multiple types of V-sync and the amount of added lag will vary from game to game. As the previous poster noted, there will always be some lag added from other sources as well, including the game itself.

 

This video from Digital Foundry is extremely helpful when it comes to understanding different types of V-sync:

 

 

This Blurbusters article contains some lag measurements when it comes to V-sync vs. various G-sync implementations:

 

https://blurbusters.com/gsync/gsync101-input-lag-tests-and-settings/

 

If you're trying to decide whether to buy a variable refresh rate (Freesync/G-sync) monitor or a non-VRR monitor for gaming, YES IT IS WORTH IT to buy a VRR display. It's about so much more than just input lag, it's also about being able to completely smooth over the inconsistent frame pacing that is essentially required by V-sync if you don't have VRR. It's a massive boost to how "smooth" your games will feel.

[Alex Atkin UK]

1 hour ago, Satan_Prometheus said:

There are multiple types of V-sync and the amount of added lag will vary from game to game. As the previous poster noted, there will always be some lag added from other sources as well, including the game itself.

 

This video from Digital Foundry is extremely helpful when it comes to understanding different types of V-sync:

 

 

This Blurbusters article contains some lag measurements when it comes to V-sync vs. various G-sync implementations:

 

https://blurbusters.com/gsync/gsync101-input-lag-tests-and-settings/

 

If you're trying to decide whether to buy a variable refresh rate (Freesync/G-sync) monitor or a non-VRR monitor for gaming, YES IT IS WORTH IT to buy a VRR display. It's about so much more than just input lag, it's also about being able to completely smooth over the inconsistent frame pacing that is essentially required by V-sync if you don't have VRR. It's a massive boost to how "smooth" your games will feel.

I was reading something recently that Freesync has more lag than G-Sync (on a monitor with actual G-Sync hardware).

Personally I'd say G-Sync is always going be the best option as even as games get heavier and you struggle to keep the frame rate up, it doesn't matter so much as the lag should stay more or less the same.

[Satan_Prometheus]

26 minutes ago, Alex Atkin UK said:

I was reading something recently that Freesync has more lag than G-Sync (on a monitor with actual G-Sync hardware).

Personally I'd say G-Sync is always going be the best option as even as games get heavier and you struggle to keep the frame rate up, it doesn't matter so much as the lag should stay more or less the same.

Battlenonsense tested this a couple times:

 

 

It seems like the actual difference in delay isn't very big. The problems that he mentions may be issues with the monitor itself, not Freesync as an ecosystem as there are no standards that Freesync monitors have to meet.

 

This is an interesting, related video about 'G-sync compatible':

 

 

[LOST TALE]

9 hours ago, Mark Kaine said:

input lag isn't a bug, and there's always some input lag, usually way  more than people think too.*

 

there's also indeed several sources which add up, and that's what people notice (everything under 100 ms overall response time aka "input lag"  is usually perceived as "ok" especially when people are playing on TVs)

 

 

https://en.m.wikipedia.org/wiki/Input_lag

 

 

*

that's crazy considering cpu and gpu frametimes add up to multiples less.

[LOST TALE]

9 hours ago, Satan_Prometheus said:

There are multiple types of V-sync and the amount of added lag will vary from game to game. As the previous poster noted, there will always be some lag added from other sources as well, including the game itself.

 

This video from Digital Foundry is extremely helpful when it comes to understanding different types of V-sync:

 

 

This Blurbusters article contains some lag measurements when it comes to V-sync vs. various G-sync implementations:

 

https://blurbusters.com/gsync/gsync101-input-lag-tests-and-settings/

 

If you're trying to decide whether to buy a variable refresh rate (Freesync/G-sync) monitor or a non-VRR monitor for gaming, YES IT IS WORTH IT to buy a VRR display. It's about so much more than just input lag, it's also about being able to completely smooth over the inconsistent frame pacing that is essentially required by V-sync if you don't have VRR. It's a massive boost to how "smooth" your games will feel.

tbh I feel like 75hz beats 60hz freesync but that's when I got 75fps too.

[Satan_Prometheus]

9 hours ago, LOST TALE said:

that's crazy considering cpu and gpu frametimes add up to multiples less.

Yeah, and that still matters when it comes to visual consistency. 

9 hours ago, LOST TALE said:

tbh I feel like 75hz beats 60hz freesync but that's when I got 75fps too.

Which makes sense, a locked 75 fps will feel better than a locked 60 fps, but the difference between VRR and no VRR is going to be more evident when you fall under the refresh rate target. 55 fps on a 60Hz VRR monitor will feel better than 55 fps on a 75Hz monitor without VRR.

[Mark Kaine]

18 hours ago, LOST TALE said:

that's crazy considering cpu and gpu frametimes add up to multiples less.

yeah theoretically that's negligible in regards to input lag, but uneven frametimes usually introduce some sort of "microstutter" which as far I can tell is much less understood as input lag ( and that's not exactly common knowledge either how it works)

 

Also fun fact: so on top of these average response times (let's be *generous*, 150ms...) there's also TV / monitor lag which can easily give you another 50-75ms (not monitors but for TVs that's pretty common) plus you also get easily another 100-300ms of "network lag" (referred to as "ping" usually) so we're looking at around 0.5-1 second of actual input latency (lag) when playing online games on "average" (of course this will depend on many factors, but since most people are going to be more or less unaware I think that's a majority of people gaming with these sorts of input lag or latency)

¯\_(ツ)_/¯

[LOST TALE]

3 hours ago, Mark Kaine said:

yeah theoretically that's negligible in regards to input lag, but uneven frametimes usually introduce some sort of "microstutter" which as far I can tell is much less understood as input lag ( and that's not exactly common knowledge either how it works)

 

Also fun fact: so on top of these average response times (let's be *generous*, 150ms...) there's also TV / monitor lag which can easily give you another 50-75ms (not monitors but for TVs that's pretty common) plus you also get easily another 100-300ms of "network lag" (referred to as "ping" usually) so we're looking at around 0.5-1 second of actual input latency (lag) when playing online games on "average" (of course this will depend on many factors, but since most people are going to be more or less unaware I think that's a majority of people gaming with these sorts of input lag or latency)

¯\_(ツ)_/¯

I think network lag shouldn't be in the same category. If you die behind a corner, that's just a visual representation. Fact is you were out of cover for too long and were shot. The enemie's ping doesn't change that, it only changes when you become aware of your loss. So I don't think it's mature to be frustrated at being shot behind a wall as long as hacks are not suspected.

[Mark Kaine]

16 minutes ago, LOST TALE said:

I think network lag shouldn't be in the same category. If you die behind a corner, that's just a visual representation. Fact is you were out of cover for too long and were shot. The enemie's ping doesn't change that, it only changes when you become aware of your loss. So I don't think it's mature to be frustrated at being shot behind a wall as long as hacks are not suspected.

well I don't play fps so I rarely get frustrated over them But that doesn't change that it adds to input latency, and also it's not the same for everyone obviously, so some players will always have an advantage or disadvantage regarding this .Same as low input lag monitors (for example) if it's just 20ms or something it can already be a factor ... 

 

my point was more that factually many people play with around a whole second or more of input lag.... that's crazy but also not so bad as it sounds because the brain gets used to this, it becomes a problem if it's fluctuating then it becomes indeed a guessing game and can lead to frustration.

 

 

[LOST TALE]

1 minute ago, Mark Kaine said:

well I don't play fps so I rarely get frustrated over them But that doesn't change that it adds to input latency, and also it's not the same for everyone obviously, so some players will always have an advantage or disadvantage regarding this .Same as low input lag monitors (for example) if it's just 20ms or something it can already be a factor ... 

 

my point was more that factually many people play with around a whole second or more of input lag.... that's crazy but also not so bad as it sounds because the brain gets used to this, it becomes a problem if it's fluctuating then it becomes indeed a guessing game and can lead to frustration.

 

 

In most games, especially fps. the game doesn't wait for network latency to display your actions. So you do not see any of the input lag from it.

 

I don't buy your 1 second. you're inflating numbers. game and monitor latency easily just 70ms and can be as low at 26ms

[Mark Kaine]

54 minutes ago, LOST TALE said:

most games, especially fps. the game doesn't wait for network latency to display your actions. So you do not see any of the input lag from it.

I don't understand what you're saying so it takes  ~300ms from your home to the game server but there is really no delay because..  time travel?

 

 

54 minutes ago, LOST TALE said:

I don't buy your 1 second. you're inflating numbers. game and monitor latency easily just 70ms and can be as low at 26ms

This was kind of a "worst case scenario" even though in reality it can be a lot worse of course because things don't always work perfectly, based on that wikipedia entry which actually has realistic numbers, I just added network latency. Everyone who ever played an online game knows what it is and how bad it can get ...or so I thought.

[LOST TALE]

4 hours ago, Mark Kaine said:

I don't understand what you're saying so it takes  ~300ms from your home to the game server but there is really no delay because..  time travel?

Input lag as experience is the time difference between performing an action and seeing it take place on the screen. Most games, like battlefield 1, perform the action asap client side regardless of ping. In other words, they do not wait for a network response.

[Mark Kaine]

3 hours ago, LOST TALE said:

Input lag as experience is the time difference between performing an action and seeing it take place on the screen. Most games, like battlefield 1, perform the action asap client side regardless of ping. In other words, they do not wait for a network response.

oh right now I get what you meant... yeah u see the action happening on your screen all the while the game is still calculating stuff server side,and may not even received your signal yet, which can lead to things like you're doing a headshot but the target already moved (some time ago) and the headshot doesn't actually count therefore, the target is still standing despite you're getting a clear headshot...

 

I've seen this lol. 

 

That all depends on netcode and how each game handles this stuff individually ,so this differs a lot between games and genres I guess.

 

Though it's still latency, just not display latency, which I didn't specify iirc, I just said overall latency/ lag/ however you wanna call it.