Use case for Vsync? I never seem to notice frame tearing in-game and every time I do turn Vsync on I can feel input lag.
i remember noticing a difference the first time i tested it in a game like 15 years ago then never noticed a difference ever again when i tried itsome kinda reverse vaporware
With modern G-sync setups, you should have an FPS cap that is ~0.25-0.3ms frametime for your refresh rate, for my 240hz monitor I put a global fps cap of 226FPS via the Nvidia app (or Nvidia control panel). Then I have low latency mode set to "Ultra", and V-sync set to on (again, both in either the Nvidia app, or Nvidia control panel).Then in game you want to make sure V-sync is turned off.
>>107829082>Use case for Vsync?Deprecated legacy tech.
I was told to set the computer's max fps to 237 on my 240 hz monitor in NVIDIA control panel for gsync to work best. Is that trueThey also say disable vsync setting in game for gsync to actually work right?
>>107829082It's more noticeable the lower the framerate. But it doesn't matter, even if you do notice screen tearing, v-sync is way worse than noticing screen tearing once in a while. The performance hit and mouse lag ain't worth it.
>>107829436See >>107829250They stopped using the -3FPS rule awhile ago, now it's 0.25ms-0.3ms, which for 240hz means ~226FPS.
>>107829250Depends on the game or game engine. If you don't care about tinkering for each game, your setup is probably the best. But some games have input lag and are better with in-game cap. Some have said even in-game V-Sync but I haven't found a game that benefits.
>>107829458Vsync doesn't have a performance hit. You're just not wasting computing resources on frames you don't see.
>>107829436https://blurbusters.com/gsync/gsync101-input-lag-tests-and-settings/14/
I use it in games that my pc runs really fast and I don't on games it doesn't. My thinking is there is no need to generate more frames than my monitor can produce. Should I just use FPS cap instead? Only a handful of times have I seen tearing but with those handful vsync has fixed it.
>>107829813If you play in windowed mode then there's no screen tearing because dwm. If you play in exclusive fullscreen with game vsync and driver vsync disabled then I guarantee there will be screen tearing when there's motion.
>>107829801VRR + cap
>>107829826I don't have vrr I have an older monitor.Probably gonna start using FPS cap now though.
>>107829851Depending on your monitor, you might be able to "softmod" it to support VRR using CRU.Has worked for me several times.
It's less noticeable on games with low contrast, like 3D gamesIt's also less noticeable on higher frameratesIf your monitor suffers from VRR flickering heavily, but you have a really high refresh rate, you might actually want to just play with vsync off and cap the framerate, just deal with the tearing you will notice less than said flickeringIf you're playing something where you can tolerate extra input lag, you could use vsync.Where tearing is more present and more annoying is with 2D games with simplistic graphics. Specially in emulation where most of these platforms run at something close to 60fps.
>>107829082to stop screen tearing. and if your running at 120hz that input lag isnt detectable, whatsoever.
>>107831311Most people can tell input lag at 120fps with vsync though, specially if it's triple buffered.
>>107829874you mean picrel?https://kreier.github.io/cru/
>>107831347Yes. I don't know if it's driver related but usually I've had to V rate under Edit too that it says for DisplayPort, even if I used HDMI and FreeSync range data block.
>>107831408You're actually having worse latency compared to limiting FPS.
>>107831408>>107831433to add to that, nothing is stopping you from doing that on windows too
>>107831408You can do that on Windows too.>>107831433Mailbox would give worse frametimes in general compared to manual limiting.
>>107829082Deprecated, we have 1080 Hz monitors now
>>107831433>You're actually having worse latency compared to limiting FPS.>>107831444>Mailbox would give worse frametimes in general compared to manual limiting.explain
>>107831331Triple buffering adds no latency because drivers do their draw call buffering in the background regardless.
>>107829458Vsync doesn't have a performance impact in any modern game. It only ever had a performance impact in poorly written games that sequentially ran input sampling, game update, and render, on the same thread.
>>107829823>If you play in windowed mode then there's no screen tearing because dwm.Multi-Planar Overlays (MPO) called. They want you to know you're wrong.MPO gives each window its own hardware-overlay where compositing them bypasses the DWM and is directly done on the graphics card. This also means each individual window gets access to the real vsync signal. Aaa--nd each individual window can experience tearing, if it chooses not to use vsync.This is also infinite amounts of fun with old games that deny you the ability to enable vsync when choosing borderless or windowed rendering modes. Because then you have to force it through the Nvidia or AMD control panel software. And sometimes that isn't quite compatible, because some games e.g. use different frame pacing logic when their game-internal vsync option is enabled or disabled. So you get a choice between tearing or stutter-fest.What helps with that is that on modern versions of Windows, both 10 and 11, next to MPO there's also a 'game optimizations' feature. What this does is turn D3D requests for exclusive full screen rendering context, into non-exclusive fullscreen. So you can configure those older games to render in 'full screen' instead - which allows you to enable their vsync option - while in actuality that particular Windows 'game optimizations' feature will force them to render borderless / non-exclusive. (And because of MPO, vsync will still work.)
>>107831775 (cont.)Ofcourse, this isn't a perfect solution either, because MPO is not supported on all DXGI blit modes (yey!).So for games that use an unsupported blit mode, configuring them to render exclusive fullscreen (aka true fullscreen) while the game optimizations feature is enabled, will actually have them render as non-exclusive fullscreen (aka borderless) and without MPO, vsync won't work.The game optimizations feature is a GLOBAL feature.So enjoy toggling that fucker on and off on a game by game basis; as need arises.And ofcourse- it's hidden deep inside the modern Settings app, something like 6 or so clicks deep.Because Microsoft cannot get ANYTHING right.
>>107829082VR headsets have vsync (and no VRR) per default. It can be annoying to have a 90hz headset running on 60fps and therefore running on 45fps with the occasional 'jerk of 90fps'
Doom Eternal with vsync on my 100Hz monitor looks like ASS, like I've put the settings to the lowest and a really low resolution. I thought I was losing my mind. The moment I disabled vsync everything went to looking sharp and crisp (and also got like 300fps) and never saw any tearing. So, no, it's not worth it to use vsync unless maybe your hardware is complete shit.
>>107829769>>107831751V-Sync has a impact on latency.
vsync is a relic of 20 years ago and should not be turned on in any modern setup
>>107831775>>107831794>Ofcourse, this isn't a perfect solution eitherIt doesn't have to be, because v-sync is and outdated and anyone doing gaming is expected to have a VRR capable display.
>>107832170>is and outdatedan*
>>107832128Perfectly responsive on my Linux machine.
>>107829082>Use case for Vsync?Lower power consumptionBut I rather use a fps cap of 59-63. Depends on game
>>107829082VRR LITERALLY solved all issues related to this, but then OLEDs came and suffered from VRR flicker so OLEDfags are now trying to convince everyone that VRR "isn't that important just turn it off"
>>107829458Yeah you're absolutely wrong about that. If the frame goes out of sync it's VERY obvious and easy to see. You're lucky if it never goes out of sync and doesn't stay that way for more than a second, you have to RNG for the frames to be synced to the actual refreshes.Perhaps some games that you're playing are way too blurry for you to notice it, but if you have clear graphics with distinguishable visual elements on screen you can see it right away and it's very distracting.>>107832128Lots of things have an impact on latency including low frame rate, very few of them have meaningful impact. DLSS framegen has much bigger impact and you still see people use that.
>>107829113I believe modern hardware and software is better at not tearing like fuck. Back in the GTX 9800 and Q6600 era I used to force VSync and triple buffering to not have the games look like ass while maintaining low latencies. Around the GTX 1000 series this became a non-issue.
>>107832288>DLSS framegen has much bigger impact and you still see people use that.Nobody caring about latency uses FG. Same with V-Sync.
>>107832128Only noticeable if you play a rhythm games.Since GDDR ram is REALLY FUCKING slow on console, they just tell you to deal with the latency lol.
>>107832605ichI bet you don't even play retro games on a CRT
>>107832605I fucking hate this meme. >only important if you play competitive!>only noticeable if you play rhythm gamesNo fucker, I want high FPS, good frame pacing, low latency and a overall great experience playing ANY game.
>>107832605Holy midwit take.Memory latency is measured in nanoseconds. This has no bearing on click-to-photon latency, architecture and software do.V-sync latency is real, but easily mitigated. See Valve's VR latency presentation.
>>107829082vsync is for non real gamers.you play novels because you don't want to read.if you delay a bot you delay yourself so vsync in your type of games makes no difference in the sports of it since the whole world would be delayed by it.
>>107832605>Since GDDR ram is REALLY FUCKING slow on console, they just tell you to deal with the latency lol.PCL, that latency, nothing to do with *DDR
>>107832576Right, most people just don't care about latency because they aren't playing in a CS tournament.
>>107831426i need to research this stuff, thanks anon
>>107832107>Doom Eternal with vsync on my 100Hz monitor looks like ASSThat's because it's frame-pacing to 60, 100 isn't a decent multiple of 60, and VRR on anything sub 120Hz is going to be a massive stutter pain due to how its frame decimation / step-down works and messes in very bad way with software trying to keep a certain frame pacing.
>>107832987It's not, v-sync syncs to vertical refresh. If your vertical refresh is 100, it will v-sync to 100. Nothing to do with 60 or decimation.
>>107832772You're saying people knowledge enough to know about v-sync don't care about latency?
>>107832772See >>107832630
>>107833010I don't know where the most people argument even came from? Nobody was talking about statistics. Why would I care about "most people"?
>>107829082I had screen tearing in TXR without VSync, so it can still be useful.
>>107832576>>107832772>>107833010People knowledge enough to know FG and V-Sync wont use either.
>>107833039That was literally my point. People who tweak and configure their own shit will drive for lowest latency.
>>107833010nta but i care about vsink and i dont kinda care about latency, i played games in awful pcs most of my life, some of them looked at times like a powerpoint, yet i hate tearing
>>107833046uh no why would they spend 10 seconds changing settings to get lower latency it's not a CS tournament :)
>>107833051Same, that's why I want as good as an experience as I can. I'm trying playing Quake 3 or UT99 at 25 FPS as a teen who can't afford better hardware. Awful framerate, awful latency and tearing, because v-sync would make it even worse.
>>107833020Vsync massively, not slightly, but MASSIVELY improves the game experience if you aren't getting perfect sync between refresh and frame rendering. If you somehow get perfect sync without vsync, great for you - that would be very good. If it's not perfect on the other hand, it looks so ridiculously awful that you'd have to be blind to accept it. It's nauseatingly bad when you get a desync right down the middle of your screen. The only reason you'd accept it if you aren't blind is if the tiny latency increase from turning it on can be the difference between winning a professional match or losing it.
>>107833092Yes, that's why VRR exists.
>>107833092people used to play competitive without v-sync since even though you got tearing the latency was much loweralso these days it's irrelevant since freesync/g-sync exist
>>107829082use case of vsync is to limit game rendering more frames than your monitor can draw, tearing wise its pretty useless with higher refresh rate monitors as human can barely notice tearing even if its there then, but with 60hz screens its good even for that.
>>107833092>>107833106Tearing becomes basically a non-issue even without VRR if you have something like a 85Hz screen (talking old CRTs here) and the game runs at 210 FPS but you get better latency compared to V-Sync ON.
>>107833009> itt: anon not understanding the difference between vsync and frame-pacingFrame pacing is something different run along the side of rendering, and attempts to keep everything, , game logic included, tuned to a stable output rate to prevent lag spikes. It's basically a dynamic rendering cap, preventing both the over-rendering in contexts where v-sync is disabled (or you'd get crazy situations such as the Starcraft 2 title screen melting GPUs) while also using historical frame times to drive establishing a frame cap that leads to smooth output, rather than oscillating between 60+ FPS output and e.g. 30 FPS output on 1% outliers. Because THAT's the exact jerkiness you notice and are annoyed by.
How do you get low latency with V-sync on Loonix with Vulkan?OpenGL is very snappy if I add "block_on_depleted_buffers" to driconf comparatively.
>>107833092I'm the anon who posted >>107832630.This is relevant to wanting a good experience these days, how? We don't need v-sync anymore and low latency makes the experience better in every scenario.
>>107829082Generally depends how close to your monitor's refresh rate the frame rate is.If you're blasting well past the refresh rate, then tearing is less likely to show up.If you're hovering around the refresh rate, then tearing really shows up. In that case you won't notice much difference in latency by enabling vsync, but if you're hitting 200fps with vsync off, turning it on and limiting to 60fps will make latency really obvious.
>>107833132Except game logic doesn't run at 60 ticks for ages anymore.
>>107833132Yeah, but this isn't 2010 anymore.
>>107833150A lot of games freak the fuck out if you go above 60fps, though.It's often what speedrunners abuse for tricks.
>>107833168Maybe if you're Bethesda and it's 2015.
>>107833168see >>107833155
>>107833128Frame rate doesn't mean anything to tearing. Tearing is a desync between the refresh rate and frame rate. You either get a desync or you don't. You can get it at high frame rate/refresh rate, you can get it at low. It's just a matter of luck whether the frame pacing is different between the display and framebuffer.Like I was testing a game for graphics settings yesterday and I could do 120 FPS without any desync... then suddenly out of nowhere when I moved my screen fast it desynced and got real nasty.>>107833142The fuck are you talking about? The technology hasn't changed. You still need some way to sync framebuffer to monitor refresh. Vsync, gsync, freesync, you need something. If you have nothing, you get tearing.
>>107833181>The fuck are you talking about? The technology hasn't changed. You still need some way to sync framebuffer to monitor refresh. Vsync, gsync, freesync, you need something. If you have nothing, you get tearing.You're not syncing framebuffer to monitor refresh though, you're syncing monitor refresh to the framebuffer. Adaptive sync is how it should have been from the start.
>>107833181>Frame rate doesn't mean anything to tearing.once your fps hits double or triple your max refresh rate tearing becomes harder and harder to notice
>>107832107This shouldn't really happen, I can't explain it.All vsync does is hold off writing the next frame to the frame buffer until the monitor has finished reading the previous frame, so there's no reason why you should get a lower quality image from it.
>>107833177Not being 2010 doesn't change that.
>>107833208It's VERY noticeable on double, I can't do 3x that's way too high so I don't know about that.
>>107833214User error problably, the id Tech engine Doom 2016/Eternal and Dark Ages run on are very well optimized and hard to shit up even on low end hardware
>>107833221much less noticeable than fps being close above or below refresh
>>107829082>Use case for Vsync? I never seem to notice frame tearing in-game and every time I do turn Vsync on I can feel input lag.i'm programming a graphics renderer for opengl and to get my slop to run reliably on windows machines, i use vsync so you can have a solid framerate, whatever the monitor is set to, otherwise it's jerky city and random splits in the screen. it's pretty useful.>>107832128not seeing it
>>107833150A lot of console ports that use frame-pacing that strives for 59.9 FPS, even if they don't have game logic hard-tied to a 60FPS vsync.And console ports are what 90% of the market is, like it or not.
>>107833247>opengl>in 2026why?
>>107833250Many, but a faction compared to decade or two ago.
>>107833250You keep saying frame-pacing but that's not really what frame-pacing is.
>>107833250>>107833260I get what you mean tho
