Why Gamers are Switching to High DPI

Check prices on Amazon below
Logitech G Pro X Superlight: geni.us/ixYX
Razer Orochi V2: geni.us/iY82ij
Endgame Gear XM1: geni.us/cxXJ
Razer Viper Ultimate: geni.us/jfIEIMv
DPI and sensor lag - what you need to know.
Video gear
Camera: geni.us/cN16f
Primary Lens: geni.us/Mfv0kQO
/ optimumtechyt
/ alisayed3
/ optimum
As an Amazon Associate I earn from qualifying purchases.

[@SimonT54]

My grandson said he is “cracked” after I show him this video. I think that is good. Newly subscriber!!! Thank you!!!

[@torontovlr]

lol, nice to hear that. I guess I'll try it out too

[@Arrradiance]

lmao

[@yosh1111]

why is this so wholesome lmao

[@uz5979]

you’re a cool grandad.

[@mistadude]

Simon T you are cracked

[@MatthewMora]

The amount effort put into this video is crazy. The combination of relevant PC/gaming information and god tier cinematography is unmatched! 👏👏

[@jolness1]

I’m guessing you recently found the channel? All the videos are mega well done, have fun watching the back catalog!

[@jolness1]

@Tube You what does that have to do with anything though? "I can't afford this so it sucks" is your argument?

[@rizqonfadillah4084]

@Tube You lol its just u not us

[@linjix]

@Tube You so u suggest we go and buy china ripoffs then

[@1dylancs]

@Tube You If you’re playing a game competitively / at a high level, it’s a massive help to have the best peripherals and hardware available. It definitely feels expensive but it’s worth it if you care about playing to the best of your ability, which many people do. Going from a top-end mouse to a lower-end mouse, you can definitely feel a difference in some aspects, even if you don’t play video games. There’s a reason that the professional players choose to use certain mice over others.

[@DeSinc]

not sure if it's been mentioned here but some sensors interpolate between samples for their higher DPI giving you a more noisy output that isn't as accurate and drifts (not that much but it's still something) with some sensors actually getting interpolation turned on at as little as anything above 1600 dpi. basically it pays to check your specific sensor and see what its max dpi is before it starts running interpolation and use that. or alternatively, pick what feels best in windows at medium sensitivity (6/11 setting in windows) and call it a day (never change windows sens value, if nothing's changed since the early days then changing that setting introduces even more artifacts that you don't want to deal with). also some mice run better at 500hz and struggle to run 1000hz and give you inconsistent reporting. you can download a mouse polling rate checker and move your mouse around quickly and see if you get 990-1000hz reporting or if it bounces between 638hz 782hz 598hz 802hz 712hz etc. for really consistently fast movement - probably just stick to 500hz if you see that. (if you move your mouse really slowly it can do that too though just be aware) and not to mention some CPUs actually can get slowed down a bit with 1000hz polling during an intensive game. my old CPU used to actually have noticeable skips in gameplay at 1000hz polling rate, just be aware of that

[@Bombipog]

half life man

[@mellified6500]

Do you have any sources for that windows sensitivity bit? WPS shouldn't be affecting your in-game sense at all and I don't see how it would be adding any artifacts when it shouldn't be affecting any games at all, but I'd love to know if it does since I do run lowered sensitivity in WPS.

[@DeSinc]

@@mellified6500 it's to do with skipping pixels. anything that is not 6/11 on the slider is dividing your inputs and missing pixel movements periodically to get the desired movement speed of the cursor on-screen. 6/11 (middle default sensitivity value) is the only one that is 1:1 and does not alter the ratio and doesn't pixel skip. going above this is worse than going below, but both are still not optimal. a setting of 8/11 can result in two pixels being entirely skipped over from time to time. a setting below 6/11 will simply result in inputs being periodically dropped entirely to emulate the lower sensitivity by not moving the mouse as far.

[@DeSinc]

@@mellified6500 further to that I've also heard that because of how the interpolation and dropping of inputs is done at settings other than the middle 1:1 scale you get at 6/11, settings above 6/11 have a minor acceleration effect and below actually have a negative acceleration effect. that is to say the faster you move the mouse at settings below 6/11, the more it undershoots, and the faster you move above 6/11 the more it overshoots, all due to how it divides numbers up in figuring out how much to divide or multiply your input by.

[@DeSinc]

@@mellified6500 sorry for annoying follow up but I just realised, if you use raw input (ACTUAL raw input) then windows sens won't affect it, you're correct. the problem is most games never used to use raw input and I'm pretty sure still don't to this day, and even some games that SAY they use it ACTUALLY don't and are still using dxinput which is altered by windows sens. just had to mention that because I just remembered it

[@BrochMan6]

Awesome video, it's really good to know dpi effects latency that much. I'd like to see you test more mice on this, and kinda come up with a top ten low latency wireless mice or something like that.

[@romaDa]

This is next level tech reviews... builds his own machines, writes his own code, and all that delivered in such a good way without any obnoxious screaming and weird facial expressions

[@Nestho_Des]

Who are you referring to

[@JTMGJTMG]

@@Nestho_Des right, who have you in Mind romaDa

[@callisto1560]

and hes good at games!?!? and hes good LOOKING!?!? what is this sorcery

[@ahmedabdul1]

@@JTMGJTMG lmaoo

[@Velakor9150]

@@Nestho_DesLinus Fake Tips??? 🤔🤔

[@JaredBusch]

I have to say; I've watched several videos on the same topic, but this is BY FAR the first one that was digestible. a.k.a. NOW I understand.

[@peeuw]

what im baffled about is how he's using Nvidia's LDAT software, but if you look up the question "Does a higher DPI give you lower latency" nvidia says it doesn't.

[@randomname2914]

how can you not understand or have known this before any videos? Do you lack brain cells?

[@peeuw]

@@randomname2914 well no since a higher DPI actually doesn't give you lower latency. it's all bullshit

[@peeuw]

@@De4dSp0t yes, i remember 3kliksphilips video debunking this.

[@pino_de_vogel]

@@peeuw it does give you lower latency... i records data more frequently thats all you should need to know, there is no measurement required. even at a fixed 100 hz polling rate 2000 dpi still haves Never data each of the 100 clock cycles its just common sense. calling facts bullshit is a hype these days but that doesnt make it so.

[@MaybeBillClinton]

Man this is some serious research. Super impressed with the extent they you've gone for this. Awesome!

[@unrealfeedzz]

May I just say, you're the most meticulous tester that i've found and that explains everything in such an easy manner. I also love the fact that you mention to be a gamer and play a lot of shooters and what not, so it's like good that I can relate more and I can understand things more. You're awesome :)

[@nasser_almashjary]

Dude! You're being very modest. It's not a pasta salad that you whipped up over there but an awesome measurement device. RESPECT!

[@harambre7867]

i know right????? he simply said "okay so i made this" as if it was nothing

[@svanesk]

😂 solid reference, the pasta salad that is lol

[@Phiz787]

Imo his delivery is perfect. Hard work speaks for itself. He doesn't need to say a thing more!

[@therick8417]

not many creators i'll stop what i'm doing to watch, but yours is definitely one of them

[@ilovetinatanilovehersomuch]

@@floki1664 u act as if a 9-5 job is the only way to survive plus you dont take into account his personal life,

[@PoKeKidMPK1]

@@floki1664 his own comment entails he has something like a 9-5 therefore he chooses whats more significant for himself... lol how ironic. your own comment also entails projecting you dont have one.

[@olivercox2565]

I REALLY want to the see the next set of tests. Mouse acceleration etc. I don’t see anyone else using something well designed like this. Very cool. And thank god for putting some rest to “wireless will never compete with wired”

[@jeroenvdw]

Good job man, I'll be watching your future videos with these tests for sure. And I've learned it's better to use high DPI with lower sensitivity, will test this out now :)

[@franzjosefstakes]

The effort OT put into this video is insane. And bonus points for including standard deviation! :D

[@abyss_mk]

This is top tier content. So many unscientific videos exist on mouse DPI where someone is dragging a mouse by hand. I would love to see the raw data in some kind of spreadsheet format but otherwise nice work!

[@termitreter6545]

Just because someone uses their hand to drag the mouse doesnt mean the test is unscientific.

[@miweneia]

@@termitreter6545 any type of testing that includes human error is unscientific

[@Vysair]

@@termitreter6545 It's because it's not a controlled environment and has too many variables

[@termitreter6545]

@@Vysair It does lower the accuracy to do it by hand. But even the test in the video has flaws. Someone else in the comments mentioned it: Does the motors performance really stay the same over the stretch of the test? Like, does it get warm and response times change, does the lubrication cause artifacts, etc. You can endlessly strive for accuracy; but in the end, the question is if this is accurate enough for the information we want out of the video. That said, I am still happy that the video-makers took the time to actually set up a test rig for more accurate testing and higher reliability. Movement-delay tests are generally trickier than click-delay test.

[@termitreter6545]

@@miweneia I dont think there is a type of testing that doesnt have the potential for human error. Like, what if the rig was set up incorrectly, a mouse not correctly fastened, or the motors performance changed over the course of the test for whatever reason? Im not saying this isnt clearly better than the usual test, more that theres a scale of reliability, there is no exact point where "scientific" starts or ends.

[@Bogescianu]

Great video and testing procedure. You should do the same thing for keyboard, like checking input lag of custom keyboards and off the shelf ones.

[@EposVox]

This is absolutely wild data. Thanks!

[@flowerface9011]

The thing about turning down in game sensitivity and increasing it on the mouse, as well as DPI, was something I totally found out by accident. Thanks for explaining all of this in a way that's easy to understand and for an overall fascinating video

[@zephyr_cg_Legacy]

Just so you know, the latency "statistics" in this video for different DPI settings are flawed: this guy doesn't understand how DPI is properly used, so his "statistics" are worthless. If you properly pair your ingame sens with a DPI that is between 400 - 1600, you will be golden. If you don't understand, or have questions, ask me. Low DPI will only cause "low latency" or pixel skipping, as it should be called, if you have a high ingame sens. Low sens players do not have this issue. TLDR: DPI is your physical sensitivity button. If you have it low (i.e. 400), you better be using your arm, and if you have it high (1600+), you better turn down your ingame sens.

[@ellibre5831]

@@zephyr_cg_Legacy you need to get your comment pinned, because what you just said is what I have been taught by pros and friends for years.

[@zephyr_cg_Legacy]

@@ellibre5831 I've posted this information around a little bit, and even posted my own comment, but I doubt I'll get the video maker's attention. It's unfortunate that people are tricked by these "facts", but hey, that just makes it easier for people like us to make it out of the average ranks eh? :)

[@ellibre5831]

@@zephyr_cg_Legacy oh without a doubt my friend \m/. I almost want to call them plebs, but Im not that mean :(

[@zephyr_cg_Legacy]

@@ellibre5831 yeah me neither, unless they are too arrogant to listen. I've had a few people reply rudely and refer to the "latency" tests as proof that I'm "wrong" lol... Oh well, at least most people listen.

[@DukeSloth]

Absolutely love the tests in this series, looking forward to the next one! Now to find a wireless mouse with a shape I actually like...

[@starertree7736]

@J M if you like the zowie shape then the steel series prime wireless is a good mouse

[@TeamClutch21]

@J M oh really you didn’t like the model o? I was thinking about getting it.

[@MasalMann]

@@TeamClutch21 I like mine tho its pretty gud in my opinion ig it depends from person to person. Logitech G Pro X wireless or whatever it is, is a little bit lighter than Model O and "Better" but its like double the price so pick ur fighter. I would suggest Model O Wireless for it being cheap and pretty good. Oh and also G Pro got like bigger Skates LUL

[@unkown215]

@@TeamClutch21 I fucking love it. Best mouse NA. it’s preference honestly.

[@vitorf2303]

Orochi V2 is really good

[@Oni1975]

Very informative and interesting video. I hadn't thought of using higher DPI settings with lower sensitivity settings in the game. Thanks for this!

[@markpew8111]

This is test set up is amazing, I have never seen any other big reviewers go to this extent to build and program their own equipment for testing mice!

[@sahilmeena8018]

Awesome video.The effort and thinking behind the testing needs to be appreciated more and got you a sub. I've always wondered if attaining the same eDPI with different mouse DPIs made a difference.

[@yun8714]

The level of analysis, depth and just sheer engineering here is god tier

[@Sam-dk2dp]

Damn you're rhyming buddy haha

[@RetroPlus]

Yeah, it's amazing

[@GRID78_8]

Yup amazin

[@randomname2914]

lmfao

[@ryanrodriguez1564]

Arduino. Wow.

[@ReignSovereign]

the b roll shot at 1:21 of the mouse and the test machine you built is s-tier. just had to show some appreciation for that. GG

[@GiSWiG]

This and the PBO2 video is why I became a sub. I never thought that high DPI and low sensitivity would be a better way to go. And yeah, the Superlight is such a great mouse and the PowerPlay charging mat is so worth it. Being an old timer when we were lucky to have side buttons, the two side buttons of this mouse are great. The G502 I had before was just too heavy and to many miss-placed buttons.

[@KennethRusli]

Really impressed with how you built together the test rig. It reminds me of the many creative ideas the 3D printing community comes up with.

[@kevinverholtz2663]

You have to be my favorite KZbinr. The information you present is always clear, concise, more creative and in depth than most other youtubers or even just trying to read up on information via forums. The amount of knowledge I feel I have gained just from watching your videos is enormous considering I am not even watching to gain knowledge but to just be entertained. So being able to consistently do both you definitely deserve praise.

[@punX86]

You could also test acceleration of the sensor :) fast move right, slow move back left and the cursor should be at the same startingpoint.

[@affect3124]

its best to turn off mouse acceleration in settings

[@masonv9333]

software side im pretty sure

[@h1Lu]

​@@masonv9333they meant the actual SENSOR acceleration (G rating)

[@Heretix1989]

really awesome in depth testing!! i've always been curious about these topics, damn nice to finally see some statistics! advanced tool-testing that reminds me of gamers-nexus style, keep it up guys! :)

[@ebonilha]

Man, you always come up with content that everybody wanted but nobody made before. The effort you put into your videos, the topic choice relevancy, and production quality in general, make you my favorite overall content creator. Keep up!

[@riba2233]

Really cool test, wish more of yt was like this. Can't wait to see new one!

[@robertdeleofitness7197]

loved the test, cant wait to see more!

[@AO-yo1ke]

good stuff !!! Nice methodical testing, and supberb vid and sound quality, a pleasure to watch.

[@greybuckleton]

Epic that you made your own kit to test this, really adds validity to what you say.

[@t1995roger10]

Seriously everytime I watch his channel, I get more and more impressed by his cinematography, his PC building and his engineering skills Ali, you rock!!

[@REMY.C.]

Handsome and big biceps 😂 This guy is the nerd we all dream of being. I have posters of him all over my room 😂

[@MrBaracu]

@@REMY.C. Really?...

[@REMY.C.]

@@MrBaracu you'll never know the truth 😂

[@Almighty_Flat_Earth]

4:18 what's the name of that game?

[@MrBaracu]

@@Almighty_Flat_Earth apex legends

[@tmmy2ootuff843]

You did a good job building that tester. even looks well made like it was engeneered and manufactured. after lots of messin', I settled on 1600 and turn down in-game sens.

[@Tr1gg3e]

super super super cool test setup, looking forward to more. subscribed

[@Hufsa]

Really great job done here letting the machines do the tests, I had in mind that higher dpi should work better but I didn't know how to measure, thank you a lot!

[@molavi65]

Great job testing and rating the mice. Love your style of presentation of data and results without shenanigan and unnecessary jargon. A true no nonsense style of testing.

[@zephyr_cg_Legacy]

The latency "statistics" in this video for different DPI settings are flawed: this guy doesn't understand how DPI is properly used, so his "statistics" are worthless. If you properly pair your ingame sens with a DPI that is between 400 - 1600, you will be golden. If you don't understand, or have questions, ask me. Low DPI will only cause "low latency" or pixel skipping, as it should be called, if you have a high ingame sens. Low sens players do not have this issue. TLDR: DPI is your physical sensitivity button. If you have it low (i.e. 400), you better be using your arm, and if you have it high (1600+), you better turn down your ingame sens.

[@dhrida5518]

I've been looking for this information for so long never found anything, today it just popped up on my feed, youtube is very weird! Thanks for the information! Very useful!

[@dannyKurt]

Awesome vid, love these kinda in-depth analysis.

[@Imabanana847]

I see that Bones Reds bearing 👀 Really cool machine you built.

[@OrcCorp]

Impressive! 5/5 cookies for your amazing work 🍪🍪🍪🍪🍪

[@theunbanned4553]

Great tool! And also great test results! I'm gonna implement this in my setup! Thank you!

[@Drew_HBK]

Taimou gave me this info years ago I'm surprised more people dont know this. Good work on the effort behind this though which is nuts

[@KG_BM]

This 1 man channel, just out researched LTT and GN

[@darkmanure]

Man actually plays games too.

[@hasanity5220]

@@darkmanure he’s actually good at them too. Gm in ow and immortal in valorant

[@harambre7867]

@@darkmanure unlike linus who reviews gaming products but can't even properly play games

[@FuckAssName]

I mean, Linus does have a family and a whole business to run, so that I can understand

[@xzraiderzx308]

Battlenonsense had done a video just like this a couple of months ago. He was the one that did all the research into it.

[@antoneverbright]

Keep these tests coming, the average customer (even though technically he wouldn't care too much) has a right to know about these things. Sensor latency and click latency tests have been somewhat neglected and I'm happy you're one of the few people that is putting it out there.

[@zephyr_cg_Legacy]

The latency "statistics" in this video for different DPI settings are flawed: this guy doesn't understand how DPI is properly used, so his "statistics" are worthless. If you properly pair your ingame sens with a DPI that is between 400 - 1600, you will be golden. If you don't understand, or have questions, ask me. Low DPI will only cause "low latency" or pixel skipping, as it should be called, if you have a high ingame sens. Low sens players do not have this issue. TLDR: DPI is your physical sensitivity button. If you have it low (i.e. 400), you better be using your arm, and if you have it high (1600+), you better turn down your ingame sens.

[@isse6790]

@@zephyr_cg_Legacy He isn't even measuring pixel-skipping. He's measuring the physical amount of time it takes for the computer to recognize a mouse movement, pixel-skipping is the game engine using the sensitivity multiplicator to make your smallest viewangle change larger than what the World2Screen function determines a function which has literally nothing to do with latency. You're the one that doesn't understand jack shit.

[@Doflaminguard]

@@zephyr_cg_Legacy thats what he said in the video. He even said if you want High DPI then lower you sense in 1:1 ratio increase of DPI.

[@zephyr_cg_Legacy]

@@Doflaminguard yes but I am focused on explaining that the latency is not an issue

[@dmnk33]

Nice insane quality videos, as expected from you king!

[@CoolJosh3k]

Would have been really neat to see the most popular mice tested. Also with the DPI tested up to much higher amounts. Perhaps a part 2?

[@koop3468]

it would probably be miniscule, or if they make another video just add another instance with much higher dpi compared to the rest of the sequence (e.g. 800, 1600, 3200, 12000)

[@blitZeero]

@@koop3468 Very high DPIs are usually artificial, meaning that the DPI increase usually stops at a certain point and instead just mutiply the value to get to that DPI, which introduces a level of error. Also, a lot of the info in the video is bogus either way, theres comments that explain why on this video

[@fivenightsatfreddursts2325]

I think the dpi increase was limited to 3200 because, as he said, you get very diminished returns in terms of latency decreases at that point. So he’s not testing it because the benefit is negligible at that point.

[@lachlanB323]

@@fivenightsatfreddursts2325 Cool. Don't care. Show us! That is like saying overclocking a rtx 4090 barely improves performance therefore nobody cares... I do! i want to see the most powerful hardware pushed to its limits

[@fivenightsatfreddursts2325]

@@lachlanB323 are you ok? It’s not really like saying that because it’s a mouse and it’s DPI. Do you raise your sensitivity in a game and then say “I’m pushing this shit to its LIMIT”?

[@U220E]

Thank you for such a good contribution to the mouse community! Although I saw no one wondering, I wonder if the mousepad’s micro-texture plays a role when it comes to accuracy at higher CPI. If I understood it correctly, higher CPI means that there’s less distance between each counts on the surface(mousepad). While Mouse sensor’s development is clearly going on, I wonder if mousepad’s surface are truly able to support high CPIs. If they're not, I think that explains why there are some people who believe that they feel better and more accurate/consistent at 800 CPI than 1600 CPI. Well it’s fair to point out that it might be because that’s what they got used to, but I think looking into relationship between mousepad’s micro-texture and higher CPI is an interesting topic. ...or I’m just a stupid monke. Thank you for reading.

[@notblazedfordays9536]

Uhh DPI….?

[@turbocomfy]

@@notblazedfordays9536 dpi and cpi are same in that context

[@notblazedfordays9536]

@@turbocomfy good to know. Never heard of it as cpi

[@ImportedPatriot91]

STOP SAYING CPI!!! JUST SAY DPI!!!!

[@Parkourcrazed]

@@ImportedPatriot91 CPI is a more accurate measurement as it translates to "Counts Per Inch" Whereas DPI translates to "Dots Per Inch" which is not nearly as self explanatory. Dots per inch refers to pixels as "dots" and Counts per inch refers to each pixel as a count. Which at least in my head makes way more sense when doing calculations.

[@furdstahmania]

Absolute dope! The effort you put on your videos is commendable.

[@GewelReal]

I did that for close to a decade. Running at native DPI but lowering the in-game multiplier makes for a small but very noticeable difference in smoothness

[@LeBassDiaries]

I rewatch every mouse video. Thanks for all the info / testing.

[@ambhaiji]

RocketJumpNinja: Well I'm Retiring to make my own mouse Optimum Tech: It's Free Real Estate

[@OnlySinusodial]

Honestly since rocket jump retired I feel this is the only other channel that's on his level or surpassed him

[@Serkong]

@@OnlySinusodial how did he retire? He still making vids

[@ryurenyama]

Really great tests, would love to see how different sensors perform on different mousepads/surfaces now

[@ikunaruLIVE]

THIS.

[@hemlock2343]

THIS. ☝

[@DeanVarney]

Yes, please!

[@mikem2022]

YES

[@CoolJosh3k]

I’d be interested in where the line is between physical DPI of the sensor and the interpolation used to reach higher DPI specs. For example, the actual sensor resolution might be 800DPI while interpolation allows it to figure out up to 3200DPI.

[@arconreef]

That's a misconception about mice. Interpolation is perfectly normal and happens at all dpi values. In fact, dpi has no effect on how the mouse calculates it's current position. Modern gaming mice always calculate their current position with the highest degree of accuracy that they are capable of, irrespective of the dpi value that you set. The dpi setting only changes how the mouse reports positional data to your computer.

[@eniff2925]

@@arconreef for example on my deathadder elite with pixart 3389 smoothing is first applied at 1900 dpi, then it doubles at 6000 and doubles again at 11300. Now in my pixart 3370 mouse it is only applied at and above 6500

[@captain-cool2275]

this is actually really interesting and useful information. thanks for making this video!

[@Nomvd90]

I’ve been trying to explain this for a while. Finally, someone who can do it. Thank you

[@Rafael_Boog]

My man Ali deadass built a modern futuristic engineering type of shit to tell which mouse us faster in ms Really a legend

[@B.K.7]

"modern futuristic engineering type of shit"

[@Rafael_Boog]

@@B.K.7 indeed

[@Weisguy]

Great vid. Thanks for doing the research on this.

[@C0sm1c_Owl]

Amazing work! So well done and explained, thanks!

[@vladislavkaras491]

Really thanks for those tests! It was interesting to watch, and especially thanks for the last lifehack to turn DPI up and game sens down. Your testing equipment is shocking. Great job!

[@zephyr_cg_Legacy]

The latency "statistics" in this video for different DPI settings are flawed: this guy doesn't understand how DPI is properly used, so his "statistics" are worthless. If you properly pair your ingame sens with a DPI that is between 400 - 1600, you will be golden. If you don't understand, or have questions, ask me. Low DPI will only cause "low latency" or pixel skipping, as it should be called, if you have a high ingame sens. Low sens players do not have this issue. TLDR: DPI is your physical sensitivity button. If you have it low (i.e. 400), you better be using your arm, and if you have it high (1600+), you better turn down your ingame sens.

[@vladislavkaras491]

@@zephyr_cg_Legacy yeah, I have heard from a friend, that with high DPI you skip some pixels, what cause in-accuaracy, or as you said pixel skipping. So, my practical question, did you feel the difference yourself with low DPI, high sens and vice versa?

[@zephyr_cg_Legacy]

@@vladislavkaras491 if you stick to the 400-1600 DPI range you'll be fine. Just don't exceed 1600. My only point was that latency is not affected by DPI!

[@honguyenbao7720]

thank you so much for doing the test, seriously!! ichange my settings from 2 sens 400 dpi to 0.375 sens 1600 dpi (which is 1.5 sens with 400 dpi), and my new "1.5 sens" is so smooth and feels like im playing at 2 sens. Keep up the good work man!!!

[@Z3t487]

Why not 0.5 1600 dpi?

[@agspda]

@@Z3t487 because the multiply factor at 1600 dpi is big.

[@Z3t487]

@@agspda I don't get it: 2 400 dpi is equals to is 0.5 1600 dpi.

[@druzzi9054]

@@Z3t487 yes i think he fucked up his maths

[@jukka1996]

I'm absolutely in love with your content. All of your data would be enough to write a bachelors thesis and you do it "just" for informational purposes and your youtube account.

[@robhoward7041]

This is fantastic work, well done!

[@BeesOnAimlab]

Btw your thumbnails are completely insane

[@David-ym2vk]

Love the combination of tech, programming, and 3d printing. Very well done as usual.

[@haitsung6574]

*Gaming

[@luigiadler]

wow, dude really went full science. Congrats on the effort bro, I was a lot but it was needed

[@abrahamalviarez5870]

nice vid, been using my wireless G603 in a 2.0 port with 3800 dpi/1000 polling rate and compared to 800, theres way much smoother (playing rainbow six extraction with a RTX 3060 at 165 hz with DLSS set in Perfomance/Nvidia Reflex ON) I feel like the system latency decreases a lot, even getting more fps (but maybe that last one about fps is just placebo) great video as always. Maybe with a 3.0 the results are even better? going to try that out

[@Quake199]

Thanks, hopefully this improves my Bronce V aim skills.

[@obeliskt1024]

Ali is on a breakthrough here. This will take him on the top of the chain when it comes to reliable subjective AND objective takes on mouse. It's great to see how much this channel and Ali has improved over the years.

[@outofahat9363]

Just discovered this channel. I love how in depth he goes with these technical comparisons. They put my "comparison ocd" to rest

[@zephyr_cg_Legacy]

The latency "statistics" in this video for different DPI settings are flawed: this guy doesn't understand how DPI is properly used, so his "statistics" are worthless. If you properly pair your ingame sens with a DPI that is between 400 - 1600, you will be golden. If you don't understand, or have questions, ask me. Low DPI will only cause "low latency" or pixel skipping, as it should be called, if you have a high ingame sens. Low sens players do not have this issue. TLDR: DPI is your physical sensitivity button. If you have it low (i.e. 400), you better be using your arm, and if you have it high (1600+), you better turn down your ingame sens.

[@megaboy2811]

@@zephyr_cg_Legacy You're wrong, you are the one who doesn't seem to understand how DPI works. DPI is basically the resolution of the mouse so the lower the DPI, the lower the resolution. Thus a lower DPI is going to have higher latency as it takes more movement for the mouse to be able to detect it. Well, "technically" speaking you could argue that it isn't really latency, but practically speaking the difference it makes is like increased latency which is the reason why he measured it as such. This is exactly the reason why he made the measurements at 4:34 in the first place. Besides calling his measurements "worthless" and copypasting it all over the comment section is not only highly disrespectful to the effort that went into those measurements, it's not even a valid point...

[@zephyr_cg_Legacy]

@@megaboy2811 so you at least understand somewhat. My only point is that if you use a DPI that properly compliments your preferred sensitivity or aiming method, the "latency" is not even a consideration. People with low sensitivity will make up for the longer gaps and even use it to their advantage by swinging their arm. You get insane precision and you make flicky accurate aim a habit. I should know, I use 400 DPI. I'm not wrong. Technically neither is he. But he's the one spreading misinformation; I'm just clearing it up.

[@megaboy2811]

@@zephyr_cg_Legacy ​ "I'm not wrong. Technically neither is he. But he's the one spreading misinformation" You're saying he isn't wrong in one sentence and then saying he is spreading misinformation... what? Besides thats exactly what he said as well. That the difference is so small, it should barely be considered, but that technically something like 800DPI and 0.5 ingame-sens is better than something like 400 and 1 ingame-sens. Which is correct.

[@JulianSloman]

Based on how good the results are do you think mouse manufacteurers already run similiar tests or perhaps less end-to-end (i.e. not counting the windows / screen etc layers) as hardware tests?

[@Gamermonk999]

i found this video on accident and out of curiousity i changed my settings from 400 dpi 8.2 sens to 1600 dpi 2.05 sens in ow2 and it definitely feels waaayyyy smoother. awesome work on this video!

[@xXJeReMiAhXx99]

wow dunno how you pulled this off, impressive engineering.

[@RabidChasebot]

I'm legitimately curious to see what the latency would be when cranked to like 16,000 DPI. I know diminishing returns and all, but still. Would be cool to see! Anyway, awesome video and really informative! Gonna experiment with turning my DPI up a bit and lowering in-game sens and see how it goes.

[@daniellake2377]

In Valorant I now use 20,000 DPI (max on Razer Viper Ultimate), 1 Windows Sens, 0.01 in-game sens. It's 200 eDPI. I can't say it's noticeable since I'm casual and use 120hz, but it's nice to be scientifically optimal

[@devlinsiantama1162]

@@daniellake2377 can u even move ur mouse like that?(the cursor i mean)

[@daniellake2377]

@@devlinsiantama1162 With Windows sensitivity down to 1 it feels like a normal person's 1000 DPI

[@ItsdaFittySimp]

@@daniellake2377 that’s awful

[@msmc5379]

@@ItsdaFittySimp wait why? i just put my dpi to 8000 but 0.1 sens overall using rawaccel so it feels like a smoother 800dpi whats wrong so i can fix it

[@tahirsanli]

Great work man. Big kudos!

[@101rockfreak]

Possibly the best PC gaming related vid I've seen in a while! Great scientific method. I presumed the results would be what you found. But it's great to see someone do the work!

[@pedrocaballero7590]

Man, great video. Two weeks ago a changed my DPI from 800 to 1600 and the sensitivity in valorant from 0.4 to 0.2 because I thought the input delay might be lower that way, and I was right! Your video is gold!

[@zephyr_cg_Legacy]

You wouldn't be able to notice the 1.7 millisecond difference in so-called latency between 800 and 1600 DPI. The latency "statistics" in this video for different DPI settings are flawed: this guy doesn't understand how DPI is properly used, so his "statistics" are worthless. If you properly pair your ingame sens with a DPI that is between 400 - 1600, you will be golden. If you don't understand, or have questions, ask me. Low DPI will only cause "low latency" or pixel skipping, as it should be called, if you have a high ingame sens. Low sens players do not have this issue. TLDR: DPI is your physical sensitivity button. If you have it low (i.e. 400), you better be using your arm, and if you have it high (1600+), you better turn down your ingame sens.

[@gabrieldasilvamoreira9452]

After i saw a video from the battlenonsense channel a few months ago, i changed my dpi from 800 to 1600 and have noticed a big difference since then, the response time is noticeable and you can get a lot better accuracy.

[@ernestfrigelj2876]

sure

[@vasiliynkudryavtsev]

Outstanding research!

[@photaiguy22]

Loved the video. Very informative thank you! This is a bit niche but I'd love to watch a video comparing input latency on the various cloud gaming services available. You'd have a great eye at discerning how good or bad they are compared to a local machine

[@paulapeno.]

"Why Gamers are Switching to High DPI" Me who started high DPI "Welcome, to paradise"

[@zipzinger9159]

i have been playing on 2400 for so long and people called me a fool i still pown em every day XD

[@hapstipo5301]

nice bro, good work, love ur content

[@MikkoRantalainen]

Great work and you explained the methodology well!

[@Shkreych]

Thank You for this test. Very useful information.

[@Pericles777]

I started using 1600dpi as opposed to 800dpi back in 2016, havent looked back, when you play with high framerates you can feel and see the difference quite a bit more

[@abrahamalviarez5870]

Im using 2100 the last 2 weeks with 1000 polling rate and so far is amazing, also high dpi benefits from high hz monitors, with 165 hz monitor 2100 dpi feels better that 800 dpi

[@hxd9321]

I’ve been using 8000 dpi and I’m dropping people like fliesssss

[@hmdak]

Yk isa good day when Ali uploads a video

[@gdbroque]

Best day

[@nuclearsydrome9926]

fr

[@fuzz11111111]

I remember the thing that initially made me go for high DPI back in the day, was that as GPU's got better and video resolutions got higher, it eventually got to the point I noticed that I was pitching/yawing by more than 1 pixel at a time. This was in Quake 2 and 3, where I ran higher sensitivity than I do these days (granularity did still matter though, it's not like there was a scope on the railgun so precision was up to you).

[@themainevnt3419]

Dude is on point with his stuff. So dig this channel.

[@TcSomers]

I'll try and do some controlled testing of my own, but from personal experience of playing at 3200 DPI for 8ish years, switching to 400 DPI 2 years ago might as well have been magic with how much it improved my gameplay. Yes in sterile testing higher DPI looks like a straight up improvement. But the reality is that humans are....human. Higher DPI means you're more prone to micromovements and jitters affecting your aim more. If you're a professional level gamer, higher DPI might be more beneficial, but for even above average gamers it really isn't helping. When my friends stream their games, we can easily tell if their DPI is too high by how twitchy their mouse movements are. We've convinced nearly all of our friends to drop their DPI to the 400-500 range and every single person has noticeably improved their ability to click targets.

[@ukultur9237]

I really love this testing! This is some GN level testing :D Pure curiosity here, no engineer at all; With the stepper motors, and belt drive, could it not POTENTIALLY have slight variations in acceleration, or break-in/wear degradation, that could skew the results over time? I am sure that in this testing there are multiple tests take for each unit that is averaged in the end, so I don't suspect any true fault to be at hand. It was just that in my mind, when testing pure input latency over a path measuring up to 10 cm, I would think a dual solenoid piston action, one on each side, would remove any potential acceleration latency involved. Yes I have realized that the human hand does not snap 6 cm in any direction near instantaneously, but in measuring only the actual mouse sensors, I believe it would be more accurate. Thoughts?

[@termitreter6545]

There definitely are some potential inaccuracies, but if the tests are consistent enough, then it probably doesnt matter. Just gotta do a control group and its probably fine. Frankly, the actually message of the result seems to be mostly that it really doesnt matter. 400 DPI is a tiny bit slower, but with 800 DPI and beyond youre reaching differences that probably arent relevant for humans. Which makes sense and is consistent with testing that eg Battlenonsense did.

[@199t8]

I'm also wondering if he accounted for the time sound takes to travel, since sound takes nearly a millisecond to travel one foot :/

[@EssillusionsUHD]

What a great video, dude great work taught me things I didn't even give a second thought😂

[@playstationRLZ]

Excellent work. I'd be curious to see what kind of difference 8k polling vs 1k polling would make at 1600 and 3200 dpi since the gap is already so close.

[@ivannikitin2476]

Great approach! Thank you. It would be interseting to see the results @ 20k++ DPI. Will it give further latency decrease?

[@zephyr_cg_Legacy]

The latency "statistics" in this video for different DPI settings are flawed: this guy doesn't understand how DPI is properly used, so his "statistics" are worthless. If you properly pair your ingame sens with a DPI that is between 400 - 1600, you will be golden. If you don't understand, or have questions, ask me. Low DPI will only cause "low latency" or pixel skipping, as it should be called, if you have a high ingame sens. Low sens players do not have this issue. TLDR: DPI is your physical sensitivity button. If you have it low (i.e. 400), you better be using your arm, and if you have it high (1600+), you better turn down your ingame sens.

[@isse6790]

@@zephyr_cg_Legacy He isn't even measuring pixel-skipping. He's measuring the physical amount of time it takes for the computer to recognize a mouse movement, pixel-skipping is the game engine using the sensitivity multiplicator to make your smallest viewangle change larger than what the World2Screen function determines a function which has literally nothing to do with latency. You're the one that doesn't understand jack shit.

[@hxd9321]

@@zephyr_cg_Legacy All of what you just said didn’t make sense, and just mostly contradict yourself...

[@zephyr_cg_Legacy]

@@hxd9321 ask away.

[@ponaskompetencija]

This was already known back in 2008, that is why i went from 400 dpi to 800 in quake. Plus tracking is far superior on higher dpi, less pixel skipping.

[@trustyTankadin]

Right? Like great to have some data to back this up and cool testing rig, but this has been known for well over a decade. Title of the video seems like click bait.

[@Dizzeke]

current mice will 1:1 pixel no matter what so long as your windows pointer speed setting is 6/11 (the middle) so you can use any dpi, i use 500 sometimes 550. no pixel skipping, thats just how many exact pixels will be covered per inch of mouse movement.

[@willtull]

Outstanding work!

[@realchaopeng]

Love the test kit. I think the test kit has issue: the kit depends on the monitor flashing which is the more likely an End-to-End latency, it includes the time in message queue (Windows), application polling rate. I know some test kit use the signal in USB cable which maybe more accurate for hardware only latency.

[@manuelmoscardi6606]

The speaker to microphone also can be easily bypassed, by connecting directly the wires of the speaker with the jack on the LDAT

[@DanilErofeev]

Not sure if you could call it latency actually in dpi testing - you need to cover more distance to register event in lower dpi vs higher. It's like measuring the time it takes to cover keystroke length in the keyboard before actual switch signal goes from it. If one keyboard has long keystroke and another is based on touch interface but you start to measure time when you touch it then it's not latency I believe?

[@ppeez]

It is latency in fact

[@edddie7563]

you are completely correct people are just npcs in these comments

[@martinzhang5533]

Crazy how I myself as a gamer and a software/hardware engineer have never thought of this. Now that I think of that, Higher DPI +lower multiplier works like supersampling and lower DPI + higher multiplier is kind of like up sampling. I will definitely try doubling my DPI, although I doubt it would be able save my poor aim.

[@Bldzera]

lower dpi is more accurate dont double ur dpi 🤣

[@ItsdaFittySimp]

Stick to low sens

[@RequiemOfSolo]

@@Bldzera stop spewing ignorance.

[@ProFow]

but with higher sampling rate results in a sensor that's more susceptible to noise aka. (the slightest hand movement that you didn't intend to be picked up). This means that even the slightest movements will throw off your aim at the same eDPI.

[@RequiemOfSolo]

@@ProFow false because the lower in game sensitivity that you are using to compensate for the higher dpi means that the in game camera has much smaller angle increments it can look at. So whereas on say 800 dpi 3 sensitivity when you "hand jittered" you would skip say 1.5 angle increments, on 1600 dpi 1.5 sensitivity your "jitter" would smoothly travel across a finer degree of angles like 0.75 per count. Not to mention this makes micro adjustments way easier and smoother too so recovering from a jitter is easier, also not to mention having a smaller angle increment means everything visually looks smoother and higher fps since you aren't chopping over big angle jumps for every mouse movement which also makes aiming easier as visual fluidity is improved

[@CraaaaaabPeople]

Super interesting.... I might have a crack at dropping my sensitivity and raising DPI. Although I would argue that if using a mouse bungee the cable will never bother you again on a wired mouse.

[@sohighsosavage6070]

Nice work !