Does CS2 have an Input Lag problem?
Vložit
- čas přidán 28. 09. 2023
- Input lag / e2e latency tests for CS:GO and CS2. Keep in mind that the results can be very system specific, some may struggle more with latency issues in CS2 than others.
Amit's guide: github.com/amitxv/PC-Tuning
Easy to follow guide: / windows_1011_optimizat...
✅ LethalGamingGear
Discount code (5% OFF): haus5
Referral link: lethalgaminggear.com/?ref=hau...
✅ G-WOLVES
Discount code ($10 OFF): HAUS
Referral link: shop.g-wolves.com/?ref=haus
⚠️DISCORD LINK: / discord
⚠️TWITTER: / hausgames_
Best CS DM Servers for Europeans: brutalcs.nu/servers - Věda a technologie
Note that the sub-tick system achieves the fact that hit registration and the animations are not necessarily in sync. What I'm interested to test, is how the game feels, not the hit registration!
I think reflex would work better with a lower end GPU or an older GPU. Do you have an older GPU you could test with? or test in a part of the map that uses lot of gpu like smokes when it drop FPS like crazy
Hi, from where you took that ms measurments? Performance Overlay from GeForce or these are just LDAT results?
Did you tried with Reflex+Boostg + G-SYNC + V-SYNC (NVCP)?
Thank you for the comparison. Reflex setting is also noticeably worse on my rig and cs feels 'off' in general after years grinding in 128 servers. Putting all the hopes in future patches.
Interesting results. Curious to see your optimization links! (They don't seem to be in your description) Might have to do some tinkering myself.
Added the links!
I'd love to see more games analyzed here! I play a lot of Team Fortress 2 (another source based Valve game) and Overwatch 2, and I can sort of notice a difference. I would love to know the numbers, and whether or not it's just placebo :)
Nice video, It's nice to have some data even if it's just a little sample.
*You can do a delay comparison in different resolutions, for example native resolution 2560x1440 VS 1280x960. I have exactly the same monitor, and I've always had this doubt, because I believe that the native resolution has less input lag, but players always prefer to play in 4:3 resolutions*
Hey Haus, I know this is the wrong video, but are you able to help me discern the differences between the shape of the la-1 and the op1? I tried the op1 in a store and really liked the hump and flat sides, but unfortunately the sides were too thin for my liking, so after seeing you say the la-1 is basically a larger op1 in your op1we review video, are there any differences in the hump curvature and size especially? I'd like it if the sides of the la-1 felt similar to the op1, just wider as well. Great video anyway!
Are you going to retest latency settings? I'm curious if it changed after many updates.
best video to explain this problem at the moment .. thx
If this test is based on your screen feedback then it could be wrong since CS2 animation is tied with the tick, so when you click your mouse, the animation will need to wait until the next tick to appear on your monitor.
The idea of the test is to compare the E2E latency with the gun fire animation to CS:GO. Sub-tick system and hit registration is an entirely different subject.
@@hausgamingthe thing is the deviation will always be 0-16ms on when the muzzle flash appears, compared to when the shot registered. Sometimes u can kill someone and the gun only does the animation 16ms after the guy died
@@stefanpapdi5151 as I said, the sub-tick system aka how hit registration I'd handled and e2e latency are two entirely different subjects :)
CS:GO animations (and shots) are also tied to tickrate. However, in CS2, shots are not tied to the animations, but the animations are *still* tied to ticks. The only real way for valve to solve this is to move to a real event-based system rather than the timestamping system they have implemented right now.
Great video. I've always wondered why you show the lowest/best results at the bottom of your graphs, how come? Oh and I can't find the links you are talking about in the video!
Thanks bro! It's a combination of the graph that I'm using being horrific in terms of inserting results, and me being lazy. Added the links to description just now!
ehhhm did you use the new NVidia driver? There is a special topic for CS2 and Reflex? And I cannot find the link in discription for the optimization stuff you talked about :)=
Links in description, and used the newest Nvidia driver!
@@hausgaming ahh shit... I thought you prob. used the older one.. ok sad story... the games subtick is so broken you ruin your whole aim with CS2 atm....
As other people have said, when you measure latency using a mouse click, the delayed animations in CS2 make it appear that the latency is worse than it is. One way that people get around this is to run the test using mouse movement rather than mouse click. One post on reddit showed that CS2 latency was better than CSGO latency when capping framerate (at least on his setup) , but CSGO latency was better with uncapped FPS, due to the higher framerate on CSGO as compared to CS2.
I'm aware of the difference between hit registration by the sub-tick system and the animations, the animations are not delayed compared to CS:GO however, the sub-tick system can just handle hit registration at faster rate!
Can you share what guides you used for win 11 optimization?
Done, links are in description!
Hi nice latency test. i wonder what is latency with hags on vs off. With reflex enable and disable?
Hey yo Haus, taking the opportunity here, and giving a video suggestion if you like.. please, could you compare NVIDIA vs AMD GPUs in CS2 ? Does it make sense to buy a NVIDIA GPU to support Nvidia Reflex if people complain about being input lag ?
Cheers from 🇧🇷
nice video bbut your forget the link in description for the tweak :D
Done!
Those windows optimization results are very surprising. Definitely going to have to look more into that.
It's just too bad that you have to download so much crap to get an optimized windows :/
just disable threaded optimization in nvidia settings. unfortunatelly there is no way how to disable it on amd card.
what sens or cm/360 do you run normally in cs?
Link ?
I think you forgot the links, I'm interested on those. Also as someone else mentioned, testing with mouse movement will be better to test input lag in CS2.
Added them, and the gun fire is actually exactly what I want to test.
I mean i am playing on a 60Hz monitor anda GTX970 from 2012 and getting around 200fps, but the input delay is something else. When i look left and right i can feel my fov dragging behind. This wasn't the case in csgo on the same setup.
I would like to know why playing on Valve servers feels like playing with 150ms ping even when I have stable 200fps, 240hz, 99.8% smooth frametime and 3-5 ping! I always feels huge delay and because of that it's really hard to burst and spray.
Do you know antyhing about that issue which have most of the players?
I just downloaded cs2 for the first time playing but left the right or forward movement is weird when I press any movement button my character just keep moving on that way even I stopped pressing button.
this guys inventory is clean asf
Links for e2e latency?
Bro, I'm in love with your accent
I can’t put my finger on it, but mouse input feels weird when I’m moving and preaiming angles
can you test valorant with reflex on vs off?
where are the links for the optimizations?
what he say
Nvidia Graphic Driver Version? 537.42 has "Game Ready Driver for Counter-Strike 2 with NVIDIA Reflex." I don't know if that will help with latency
Tested with the newest driver indeed!
did you test this with the latest nvidia driver at the time?
537.42 states "This is also the Game Ready Driver for Counter-Strike 2 with NVIDIA Reflex." if not, can you try the current latest driver 545.84 and retest?
Considering how new this game is, it's a non-issue. But yeah compared to the old CS, it's bound to look bad. Still early days so have to wait for future updates, optimisation and hardware to bend it to their knees. Just hope some gameplay is in sue.
Can you do this to valorant to now I wanne know that reflex is the problem
Every single shot that i hit i feel like i shouldn't of hit it. And every shot i miss i feel like i should have hit it
share your video setting bro
Yknow, I keep seeing that number "26" behind you and I keep thinking that its Mahrez. What does it mean?
I'm a Chelsea fan, it's a John Terry displate!
It probably isn’t the game itself but rather the servers. A few weeks and it should be fixed
Did you use same fixed fps? You probably get higher fps in cs go which should reduce input latency.
Same fixed FPS, reached easily with both games! But the conclusion of the video is anyway, that there isn't really any real input delay between CS2 and CS:GO 64 tick!
Was you using native resolution?
Yes!
CS2 has the advantage of subtick which will technically negate some of the delay if we disregard the firing animation. Your shot will land sooner but you'll see it on your screen with a similar delay to CS:GO 64 tick.
Yeah, the sub-tick system is an entirely different subject, it makes the hit registration work very well in some cases. What I'm mainly interested in how the game feels at a client level!
what is a tick haus I dont know what that even means?
Y E S
yes it does. i7 12700 rtx 3060 32 gb of ram and top notch peripherals, moving is delayed, mouse is delayed.
They won't fix it.
And the hit reg is really incosistent
@@justzen486bro its terrible, even when i get one taps they feel horrible and delayed. Feels like playing on 140 ping
maxim made a video about de-subticking cs2, and u can bind commands like bind mouse1 “+attack;clear” and then shooting is no longer subtick its on the next tick like before, could be great for testing now to see if its lower or higher input latnecy
now try it with a 9600k and 2060super so we have some real world comparison. I pay lots for hardware but upgrading always to the newest is unrealistic
i have big promlem !
Its hard to test right now, on cs2 the ldat data doesnt rly show real latency, its a 0-16ms variable. In csgo shooting was tied to tickrate so both animation and shot happened in the same delayed tick. In cs2 the shooting happens instantly but the muzzle/animation is on the next tick like csgo
I'm aware of this, what I tested is e2e latency and compared it to CS:GO. I'm not interested in the latency between you clicking your button and the shot being registered.
@@hausgaming ah i see, but with how cs2 tickrate is there is no consistent result, even after like 5000 tries it still wouldnt be conslusive i think. Maybe the 1000fps cameras could better measure it than ldat in cs2 but its not like i expect ppl to have 3000€ cameras lol
@@stefanpapdi5151 I mean 1000fps camera can record the same end to end result as LDAT, so in that sense it's not more accurate, both have the same issue between sub-tick and tick. It's just a matter of what you want to measure
@@hausgaming Well im not sure if the wall would show a bullet hole the same moment the gun id shot or if the decals are also tick after shooting like the muzzle flash and gun animation… 3kliksphillip said ldat could be used without shooting by just moving the mouse to a brighter area of the map, but have it re-creatable with some machine istead of per hand
Measuring the e2e latency of animations in CS2 doesn't really answer the question in the title. A more interesting video imo would be measuring the sensor latency.
I am highly against buying/trusting twitter tweaking services as majority of them are poorly informed on tweaking and overclocking. Every configuration is different and properly setting up a system is not done in 30 minutes.
Even a slight mistep in bios or windows can results in dips in fps and a single day is not enough to test stability of overclocks and make sure no error correction is happening. Boot instability is a thing, you might be stable for a single boot and be completely/slightly unstable the next boot.
Being around tweaking discords, I've heard many stories of awful optimizing being done by twitter optimizers.
Amits guide is great, but it is highly important to follow the guide from scratch and not on a existing windows install. Do not combine random tweaks without knowing what each change does.
That being said with a good system CS2 has felt great to me, I've had no issues with stuttering so far and although I cannot confirm it CS2 feels more responsive to me.
It's a direct comparison with how the game feels on client side, when you measure the delay between two comparable effects :). Optimization guides can often have a lot of settings which really do not do much, the whole scene is affected by a lot of placebo. Definitely makes sense to validate what you do to your own needs!
Isn't Reflex supposed to decrease the input lag?
Only if your GPU is near maxed out
@@yoked391 Oh, ok so I assume the GPU is not maxed out in CS2 and thats why input lag is higher?
Reflex only makes a difference when you are GPU limited.
cs2 ak spray is not the same and nobody can convince me otherwise
Capping with RTSS will give the most consistent and smooth frame times
And in game to 999
just make it performace
no, nvidia drivers have problem. disable threaded optimizations and boom.
I get better fps with hyperthreading off
Apparently putting makeup on a pig makes it better, good job Valve! for replacing and downgrading an existing game without anyone's consent. Asus should come and replace my monitor aswell because i was thinking that i needed a new one anyway soon and hopefully
Your accent looks like FInnish if im not wrong
Yeah he is Finnish