What high DPIs are recommended? i'm sooo confused

Everything about displays and monitors. 120Hz, 144Hz, 240Hz, 4K, 1440p, input lag, display shopping, monitor purchase decisions, compare, versus, debate, and more. Questions? Just ask!
RocketJumpBlur
Posts: 1
Joined: 21 May 2022, 19:52

What high DPIs are recommended? i'm sooo confused

Post by RocketJumpBlur » 21 May 2022, 21:43

Before throwing my question. I want to greet all the people here, who share ideas and curiosities and resolve issues that are difficult to understand in this wonderful forum. I guess all of us have come here because we have discovered this site and greatly admire the amount of information provided by BlurBusters and anonymous or little known people who also share a bit of their knowledge. I'm a big fan of Chief Blur Buster's work hahahaha it's amazing that this man knows so much about what he talks about. Thanks to BlurBusters, I learned about backlight strobing and understood its value. Although I am 20 years old, I also grew up with an old CRT TV. I used to watch TV and play games a lot as a kid, I didn't know what was really going on when my parents changed the old CRT TV. There was blurring to images, barely moving text, and not at all legible. My little introduction said... I'm "new" here but I've actually been reading posts on the BlurBusters.com forums since 2018-2019.
The reason I have registered here is the doubt with the DPI of the mouse. It's hard to read all the posts and messages from the Chief Blur Buster. According to him: High DPI and low sensitivity are better in the game because this would avoid micro-stutters when using backlight strobing and also reduce mouse latency. The problem is: how much DPI is recommended? :oops: I'm going a little crazy because I don't know whether to use 1600 DPI or 3200 DPI. I haven't gotten perfect sensitivity for myself yet... I've seen videos like Battle(non)sense and Optimum Tech that prove and confirm the benefit of having a high DPI mouse because it reduces latency a bit. All my life I've been a consoler, I've been playing on "PC" (it's a disgusting laptop with a GTX 950M. You can imagine that I play Destiny 2 much worse and below 30 fps) because I like that platform better and I It allows both work and play (the prices of games on PC are considerably cheaper than on console) and more or less I play decently well but I'm wasting bullets because of it... Because I haven't found my sweetpoint with sensitivity. A year ago I used an Xtrfy brand mouse, the M42 RGB model. It is very good and I love the interchangeable cover, but I needed more buttons. Within a month my Azeron brand keypad will arrive and in October my Spectrum QHD 280 Hz. Currently I have another mouse, the Razer Naga Pro. I am very happy with it but its shape and dimensions need to be improved a bit, as for height example. I would reduce it but it's my preference. I have a medium hand and the Razer mouse is more or less small for my hand (I mean it in the comfortable sense for my hand). Now I am saving to buy the components some day this year and build my PC myself with a future RTX 4080 or RX 7800 XT. I would really love for all of Chief Blur Buster's advice to be brought together in one article or post, where you always find him at the top of all the posts and it was all the advice that he has been dropping in each post. For example in this: viewtopic.php?f=10&t=7618
Tips on how to heat your monitor and reduce screen crosstalk, configure your operating system or devices to reduce latency. Maybe brief or specific tips that anyone can understand and learn things they didn't know before. I don't know if it was 48 hours straight heating the monitor. I can't remember and there are so many posts and messages to search for something specific that Chief BlurBusters recommends. I want to inform myself about absolutely everything, to improve my gaming experience to the maximum :roll:

xPrzybyLx
Posts: 13
Joined: 03 Dec 2021, 12:53

Re: What high DPIs are recommended? i'm sooo confused

Post by xPrzybyLx » 25 May 2022, 20:39

Definitely check this video of Battle(non)sense about the topic of mouse dpi: https://www.youtube.com/watch?v=6AoRfv9W110

Check this graph (from the video):
Image

Personally I switched from 800dpi to using 3200dpi and I won't go back. Mouse movement feels so smooth now!

Inco^
Posts: 101
Joined: 20 Apr 2020, 07:47
Location: France

Re: What high DPIs are recommended? i'm sooo confused

Post by Inco^ » 26 May 2022, 02:57

It's not just about latency, it's also about the length of the movement increment. 20 years ago, I was trying to aim at an enemy far/mid distance. The enemy was standing still. I could NOT adjust the middle of my crosshair on him. Why? Because the smallest increment was too big. Too much to the left... Let me adjust... Too much to the right, now. In other words, the accuracy was too low.
The increment length, is exactly what DPI is about.
PG248QP - XL2566K - XV252Q F - PG259QN - XL2546K - Y27gq-25 - AG251FZ - LaCie Electron 22 Blue IV

User avatar
Chief Blur Buster
Site Admin
Posts: 11647
Joined: 05 Dec 2013, 15:44
Location: Toronto / Hamilton, Ontario, Canada
Contact:

Re: What high DPIs are recommended? i'm sooo confused

Post by Chief Blur Buster » 27 May 2022, 18:24

It's about finally crying time -- 400dpi and 800dpi needs to be obsolete. Been saying that for years!

Many old game engines that are not 1600-dpi friendly needs to be improved to be more 1600dpi and 3200dpi friendly, with the latest mouse sensors that does a superlative job.

400dpi mouse movements over 1/8 inch means your sniper scope slew frame rate is only 50 frames per second (since it's only 50 dots over 1/8 inch).

Low DPI can sabotage your sniper scope frame rate of sideways panning (steppy step motion) if you insist on 400dpi as a sniper in an esports game. If the game properly supports high-DPI without distortion, jack up your dpi and lower your mouse sensitivity to become a better sniper.

You need a good mouse sensor, clean mouse feet, and a good mouse pad, though.

One major side effect is rocket mouse pointer during menus and inventory screens, but that can be fixed by game developers by having separate mouse sensitivity adjustments for:
- Sensitivity for mouse pointer
- Sensitivity for sniper scope
- Sensitivity for mouse turns

Three different mouse sensitivity adjustments.
Then 3200dpi is heaven (slow mouse cursor, sub-pixel sniper scope movements, sub-pixel mouse slowturns, etc)

Even better is if mice, mouse drivers, and games would consider implementing a form of High Definition Mouse API, but it's only a proposal at this stage.
Head of Blur Busters - BlurBusters.com | TestUFO.com | Follow @BlurBusters on Twitter

Image
Forum Rules wrote:  1. Rule #1: Be Nice. This is published forum rule #1. Even To Newbies & People You Disagree With!
  2. Please report rule violations If you see a post that violates forum rules, then report the post.
  3. ALWAYS respect indie testers here. See how indies are bootstrapping Blur Busters research!

User avatar
Chief Blur Buster
Site Admin
Posts: 11647
Joined: 05 Dec 2013, 15:44
Location: Toronto / Hamilton, Ontario, Canada
Contact:

Re: What high DPIs are recommended? i'm sooo confused

Post by Chief Blur Buster » 27 May 2022, 18:27

RocketJumpBlur wrote:
21 May 2022, 21:43
The problem is: how much DPI is recommended? :oops:
Up to the lowest common denominator of (game weak link, mouse-sensor weak link, mousepad weak link, display weak link)

Go too high on an old mouse sensor, and 3200dpi is laggy or interpolated.
Go too high in an old game, and 3200dpi has math rounding errors that is worse feel than 800dpi
Go too high on a crappy mouse pad with dirty mouse feet, and 3200dpi feels worse than 800dpi
Go too high with low-resolution XGA 60Hz, and you might not be able to tell the difference.

If you have the best recent mouse sensors on market, very clean stiction-free mouse feet, a great mousepad with a high-dpi-friendly texture, and best game that has no mouse-math bugs at 3200dpi .... Then currently, that's 1600-3200 for more recent high-DPI-friendly engines such as Valorant. Then 3200dpi can feel dreamy.

I still recommended roughly ~800 for CS:GO because of very old math rounding-off bugs in the CS:GO engine. I don't know if those have been fixed by Valve yet, or if it's an unfixable limitation of the old game engine.

This may have changed, I'd love to hear more about CS:GO and Valorant players who tested 3200dpi in both games, to see if CS:GO finally does 3200dpi as well as Valorant does. In both slow-and-fast mouseturn tests, slow-and-fast mouse flick tests. Especially test slow-turns too (including sniper scope hunting). Precise slow sniper slewing on high Hz monitors is where high-DPI makes a huge difference.

Muscle memory is important too, so you may have to be forced to stay at 800dpi if one of your older games is 800dpi-friendly. But with improved mouse sensors, it is now possible to just simply scale sensitivity proportionally, e.g. 800dpi 1.000 sensitivity versus 3200dpi 0.250 sensitivity (or whatever) across your different games. To make sure FPS turns is consistent speed for a given wrist flick. Fortunately, new sensors has further improved consistency between 800dpi vs 1600dpi vs 3200dpi than in the past, and you can prioritize your higher-DPI. Also, if you disable "Enhance Mouse Pointer Precision" in control panel, pointer muscle memory can be better preserved across all poll rates and all DPI.

You can use profile-switching to change DPI on a per-game basis. But one major problem is many games sensitivity settings don't have 3-decimal-digita sensitivity adjustments, so it's hard to do exactly 1/8th sensitivity, e.g. 400dpi 1.0 sensitivity versus 3200dpi 0.125 sensitivity, for muscle-memory match (assuming sensor is good enough to scale properly). This is a big beef of mine as it wrecks muscle memory on a PC full of different games -- some 3200dpi friendly and others not quite so 3200dpi friendly.

In other words, you have many factors that interfere with ideal DPI selection. It's one of the big reasons I posted the proposal of High Definition Mouse API.
Head of Blur Busters - BlurBusters.com | TestUFO.com | Follow @BlurBusters on Twitter

Image
Forum Rules wrote:  1. Rule #1: Be Nice. This is published forum rule #1. Even To Newbies & People You Disagree With!
  2. Please report rule violations If you see a post that violates forum rules, then report the post.
  3. ALWAYS respect indie testers here. See how indies are bootstrapping Blur Busters research!

RestTarRr
Posts: 36
Joined: 08 Mar 2020, 07:08

Re: What high DPIs are recommended? i'm sooo confused

Post by RestTarRr » 02 Jun 2022, 05:20

RocketJumpBlur wrote:
21 May 2022, 21:43
I've seen videos like Battle(non)sense and Optimum Tech that prove and confirm the benefit of having a high DPI mouse because it reduces latency a bit.


No one has objectively proved that high DPI is better in terms of latency as both battlenonsense and optimum tech didn't do DPI equalized tests and the results could very well be because of significantly faster mice.

People aren't going from 1.0 sens 400 dpi to 1.0 sens 3200 dpi. They'd equalize it at 0.125 to retain the same eDPI.

I do recommend switching to higher DPI though as 1600/3200 are much nicer for browsing once you get used to the speed. I don't get how people are browsing at 400 dpi in 2022, As for games that can support the higher DPI I always recommend them there, but I recommend it only in the off chance that these youtubers are correct. You don't lose anything from having higher DPI and equalizing it with slow in game sens. If they aren't correct you don't lose anything and if they are correct, you get lower input lag.

The logical conclusion however is that they aren't correct as dpi has nothing to do with update cycles which optimum tech talks about. The update cycle is the hz which in most cases is 1000, you don't change that by changing the DPI. What the DPI changes is how much the mouse moves based on movement. By reducing the in game sensitivity to match the overall dpi you previously had, should give you the same exact result because it would require the same amount of movement to get from point A to point B. In theory it shouldn't have any difference in input latency. That's why it's a shame that battlenonsense never did a second video with eDPI equalized despite being bombarded with messages about it.

Until another youtuber decides to do another test where eDPI is equal to prove it once and for all, we are in the dark. Just use the higher DPI settings in the off chance that they were correct. Even the placebo of it possibly being better might make you hit more shots.

User avatar
Chief Blur Buster
Site Admin
Posts: 11647
Joined: 05 Dec 2013, 15:44
Location: Toronto / Hamilton, Ontario, Canada
Contact:

Re: What high DPIs are recommended? i'm sooo confused

Post by Chief Blur Buster » 02 Jun 2022, 21:19

Crosspost from other thread:
Designing Your Game to be High-DPI Mouse Friendly:
New Esports Players Using 1600-3200dpi+ Settings (non-CS:GO)

Chief Blur Buster wrote:
02 Jun 2022, 17:43
History - 400dpi in older engines like CS:GO

This approach used to be more useful back with lower Hz, lower frame rates, and older engines -- especially if you're playing with only one competitive game -- but when frame rates and refresh rates went up, the pollrate and DPI needed to max-out your refresh rate benefit started to exceed the benefits of trying to optimize a low poll rate.

Playing at >240Hz with many newer engines now starts to show more benefits of the "high dpi, high poll, low sensitivity" approaches a lot more clearly, especially if you do at least some occasional eye tracking (part of the time) rather than stare-at-only-crosshairs approaches. It also scales better over multiple roles (e.g. switching in/out of sniper view)

Old mouse sensors were crappy at 1600dpi+ but this is no longer the case today. If you're using a new mouse with a good sensor, a clean mouse pad, clean mouse feet, and one of the newer high-DPI friendly games, you might be at a disadvantage at 400dpi-800dpi especially with newer 500+ Hz monitors

Newer Engines, Newer Sensors, Higher Refresh Rates Start To Change the DPI Calculus

Several newer sensors also make it easier to preserve muscle memory across dpi/poll settings (with proper calibration of sensitivity setting to compensate). A lot of sensors became more opaque on muscle memory of "magic DPI numbers", so it's now usually better to scale to a preferred sensitivity by tilting your DPI:sensitivity setting and train your muscle memory based on a preferred flick speed.

In several new engines, fast flicks at 800dpi/1sens feels much more identical to 3200dpi/0.25sens (scale numbers accordingly) than they did in older engines. 1600dpi has less issues than 800dpi though.

The magic of high DPI occurs during the slower flicks -- where 3200dpi/0.25sens looks much smoother than 800dpi/1sens allowing 4x higher framerate during slow tracking (e.g. like slow sniper sight hunting / tracking) without the coarse step-step feel. Slow mouse pans start to feel TestUFO-smooth rather than grainy steppy-steppy feel.

Low DPI Can Sabotage Frame Rate Of Your Mouseturns / Sniper Scope

If your grainy low-DPI sniper track was 20-50 stepped positions during slow tracking at 60Hz, or even just 100 steppy-steppys at 144Hz, it is not as noticeable as the graininess of the same tracking at 360Hz or 500Hz. You see, if your slowturn at low DPI is causing only 50 different mouse reports per second, that's only 50 frames per second in screen mouseturn panning position updates (mouseturn). 50fps-vs-60fps is not as easy to notice as 50fps-vs-360fps.

Although enemies move at full frame rate at all times, the horizontal/vertical movements (mouse turn, sniper scope) can feel steppy-steppy especially if you're trying to do tiny amounts of tracking (e.g. moving a few pixels at a time).

Many do train to blindly shoot after a flick, using pure muscle memory alone without visual-feel as part of your muscle memory. Then congratulations, you trained on a DPI-independent approach and you can safely increase DPI in Valorant/Fortnite with newer sensors that feels much more same at 400-800-1600-3200 at scaled sensitivities. In this case, Pass GO, collect $200, and adjust your mouse settings a little, retraining for any minor differences. The best sensors in the best hi-DPI aimtrainers (not aimtrainers based on old engines like CS:GO), should feel much more identical (than they used to) at all DPI settings (sensitivity scaled) if your muscle memory was trained differently this way.

But if you're visually tracking, then that's where problems tieing you to 400dpi begin. You've just slightly sabotaged your skill portability to 360Hz-500Hz monitors and newer engines. 240Hz-vs-360Hz is also sometimes sabotaged by low DPI, and can cause people to downgrade to a lower Hz that feels better with low DPI.

The mouse begins to sabotage the Hz of the newer ultra high refresh rate monitor, and those 3200dpi new-engine-only players with new ultra high refresh rate monitors start to take a few points this year, more points next year, and so on, away from a optimized-magic-DPI rig that worked so well in the lower Hz days.

Such optimized 250-500dpi rigs start to be overwhelmed by the other players' refresh rate advantage in a "high poll + high dpi + low sensitivity" approach now recommended with newer game engines (i.e. you're not only playing CS:GO). The players that train with the "high poll + high dpi + low sens" approach starts to produce better results now. The esports champions that are now using 240Hz monitors + 1600dpi are a great example of this.

Obviously higher DPI is more sensitive to weak links (e.g. game engine based, mouse pad based, mouse feet based, mouse sensor based, dirt, etc) so keep things as good as you can if you want the silkiness of high-DPI tracking to "keep up with the refresh rate race".

More Professionals in Esports Recognizing Higher DPI Benefits In Newer Engines

The game engine and its mechanics with high-DPI can vary a lot in the professional leagues -- so sometimes you want to train with settings everyone else is using. If nobody is using 3200dpi in Valorant, then you can still have an advantage with an optimized magic-DPI system. But as more and more players start to migrate DPI usage in games that are friendly to high-DPI, this can change.

It's all over the map, though. On sites like prosettings.net, you see many CS:GO players use still 400dpi but a lot of LoL players use 1600dpi-3200dpi. Almost nobody is using less than 1600dpi in LoL though -- but it's still relatively rare in Valorant (which feels more high-DPI friendly than CS:GO).

Be noted prosettings.net sometimes is more than a year behind for some teams that gave stale data (e.g. pre-Razer-8KHz) so some players have already increased DPI in some FPS games like Valorant and Fortnite, once they finally figured out that those engines are more high-DPI friendly than CS:GO ever was. Far more Fortnite esports teams use ~1000 to ~4000dpi Fortnite than in CS:GO, even though many still stick to 400 & 800.

Low DPI Coarseness Was Visual Cue In Classical Muscle Memory Training
(Especially when refresh rates were lower)

CS:GO has less precise mouse coordinate mathematics than Valorant and Fortnite does. Preserving visual muscle memory from CS:GO can also be a factor (even the steppy steppy feel of low frame rate may be important, like mentally feeling the number of frames clicking during a tiny slew of a sniper scope).

The frame-rounding effect of a low framerate, may be able to snap-to-pixel more precisely on tiny distant enemies, if you've trained your muscle memory visually on the steppy-steppy feel of low DPI. Low DPI is sometimes kinda like ruler if you use the visual frameratefeel of a slow short-distance flick or sniper scope slew, as part of your muscle memory training. In keeping DPI numbers really low in other FPS games, as a matter of habit -- transitioning to Valorant and Fortnite from CS:GO is muscle-memory-easier at 400dpi-vs-400dpi.

CS:GO mousefeel definitely feels more different 400-vs-800-vs-1600 (sensitivity scaled appropriately) than it does in Fortnite/Valorant. The same distance flick doesn't scale as consistently as it does in newer engines with better floating-point mouse mathematics.

Low DPI During Slow Slew (e.g. Sniper) adds More Latency Disadvantage During Competing With 360-to-500Hz Hi-DPI Players

The steppy steppy behavior adds input latency. When you are mouseturning at only 20 kliks (mouse coordinates) a second, you might have 1/20sec latency during your slow sniper slew. Ouch. This is fine for 60Hz with 1/60sec latency when the advantages of the muscle memory benefit of the grainyfeel outweighed the latency of low-DPI. Especially since CS:GO was so ultra-heavily optimized at 400dpi and 800dpi

However, the latency reductions of "high dpi, high poll, low sensitivity" on new ultra high refresh rate monitors, now massively outweigh the old-fashioned muscle memory training technique. If you spent 10 years at 400dpi, it will be tough to transition, especially since 10 years is often longer than most esports careers. But, not everyone fits into this, and can easily re-train.

If your aimtrainer app is based on Source engine (same as CS:GO) then the aimtrainer is possibly only 400-800dpi friendly because it uses the less accurate mouse mathematics of the older Source engine. So you have a training problem too; if you use an aimtrainer to retrain to high DPI, stick to newer aimtrainers that are high-DPI friendly, or use one based off a high-DPI-friendly game engine.

New Players Who Go Straight To Newer Engines (Skipping Old Engines like CS:GO)

But as more new players who's never played CS:GO enter esports directly into newer higher-DPI-friendly engines (Valorant, Fortnite, and others), is more likely to be more comfortable going straight to higher DPI approaches, where the pros can outweigh the cons if your refresh rate is high enough. With refresh rate stratospheres involved, there are additional gaming tactics that benefit from the "high DPI, high pollrate, low sensitivity" approaches.

Also, the longer mouse travel distance required to keep low-DPI approaches low latency, starts to fail against high-DPI players in good-high-DPI-compatible engines, who can physically move their hands only a medium distance (e.g. two inches instead of ten inches) to do their own differently-muscle-memory-trained flick, allowing a frag sooner than the longer flick distances required for low-latency low-DPI.

Other players don't need to retrain for different flick distance. They can easily scale sensitivity settings for exactly the same flick distance at 400-800-1600-3200 (or intermediate numbers like 2000 etc) -- new sensors, clean mouse feet and clean mouse pad lets you do that in newer high-DPI-friendly engines. So less muscle memory concern. So fewer or zero aim-retraining compromises for these players.

Also, real life frame rates and real life snipers don't have the existence of a "coarsefeel" muscle memory training tricks, so are thus more real-life realistic, for real-world shooting versus virtual shooting. Even though Fortnite is fantasy, Valorant is less so, and some games are more shooting-simulator. This can be relevant, since coarsefeel is an artificial muscle memory training trick only enabled by a finite frame rate and a finite refresh rate -- but real life has no frame rate or a low-DPI coarsefeel to train from.

The ever-higher refresh rate however tips the domino of a fence-sitter who's unsure of whether or not to go above 800 dpi, especially if they decide to discontinue playing in any of the legacy engines. If you're already a low-DPI high-sensitivity player, it's easier to transition to high-DPI low-sensitivity, because there is not always enough decimal digits in the sensitivity sliders.

Game Developers: Use 3 Decimal Digits in Sensitivity Sliders For Easier Transition Training

Blur Busters recommends game developers to use at least 3 decimal digits in sensitivity settings, to allow easier division -- e.g. 400dpi @ 1 sensitivity translates to 3200dpi @ 0.125 sensitivity. If you can only choose 0.1 or 0.2 then it's much harder to transition to 3200dpi from 400dpi.

Simple workarounds:
- You can keep your UI simple (0.1 steps), but at least allow your gamer to edit the configuration file to add decimal digits as needed
- You can let users click the sensitivity number to manually edit it, instead of slider the slider.
- Etc.

Please make sure to also additionally QA test your game with 1000dpi-4000dpi+ esports players, who have increased massively in numbers in the last few years. Also recruit newer-blood esports players who's never played CS:GO, for a more accurate vet of your high-DPI friendliness of your engine. You need to test all types of esports players too.

If you are designing or improving your game engine, use double instead of floats for scaled mouse coordinates -- to preserve maximal precision when calculating mouse coordinate mathematics. There are situations where math rounding errors can build up -- with an 8000 Hz poll rate mouse, 8000 math roundoff errors per second of a "float" accumulator instead of "double" accumulator, can spike up to human visible pixel-offset misalignments. Also, performance of doubles is identical to float on most modern computer systems, especially since you're not doing any more than 8000 sets of mouse mathematics per second anyway. It's not like a GPU doing billions of math calculations per second -- you can afford to preserve a double accumulator (not float) for 8000 accurately-microsecond-timestamped mouse deltas per second, that's virtually zero impact on performance -- just cast to a float at the final step if your rendering of new mouseturn position requires floats.

At 60Hz using a 125Hz mouse, fewer mouse math roundoff errors occured. But at 500Hz refresh rate and 8000Hz mouse, especially when multiplied by a sensitivity factor, can generate enough math rounding errors occurs in one second (in floats) to create full-integer pixel misalignments! This can make it worth transitioning your game engine to use doubles for all mouse mathematics. Also keep it as high precision for as long as possible when accumulating mouse coordinate math over a longer time period.

In designing your game engine, keep it sub-pixel accurate between adjacent polls as you can, because it adds up to many pixels of error over a long flick. Many GPU APIs use floats, but you need to keep mouse coordinates accumulated in doubles across a long time period, and only cast-off to floats or integers as needed at the last minute (e.g. passing along to an OpenGL or DirectX API etc). This is part of the reason why 3200dpi works so much better in Valorant/Fornite than it did in CS:GO. We already cover this problem in High Definition Mouse API.

But What About Turbo Mouse Cursors? Any Fix For Too-Fast Cursors?

This is a known problem, so you have to use your mouse profiles or DPI button to lower your DPI when exiting to Windows. If your game involves large amounts of inventory menus, then hopefully you can adjust your pointer speed setting without interfering with the DPI of other things.

Ideally, I recommend 3 separate sets of sensitivity settings for FPS -- a pointer sensitivity setting, a sniper scope sensitivity setting, and a mouseturn sensitivity setting (and X/Y axis where applicable, as some games use separate sensitivity settings for different axes). In some cases, if the newer engine is using rawinput and the pointer is using old mouse APIs, you can use the Windows sensitivity setting (but disable Pointer Precision checkbox). This is safe for certain FPS/RTS games so you can keep your pointer speed unchanged when increasing mouse DPI.

However, older games (like Warcraft with rawinput disabled) didn't use rawinput for mouseturns, so be careful adjusting your pointer sensitivity setting if you have a lot of games installed. Just two years ago, Quake Champions wasn't even using rawinput, but a recent patch has transitioned it to rawinput. If all your games are rawinput-compatible, it becomes safe to adjust the Windows Pointer speed setting in Control Panel to slow down your mouse pointer while increasing DPI -- since it no longer interferes with your mouse turns in games, but helps your mouse pointer in inventory screens, etc.

In the longer term, a future HD Mouse API could fix this problem permanently.

About one-third of the Faze clan now use at least 1000dpi+ in Fortnite!

The old adage that esports athletes never use more than 400dpi-800dpi in FPS is becoming outdated. It was useful for CS:GO (and transitioning from CS:GO) which is super-heavily optimized at 400-800dpi, but newer engines are now high-DPI friendly.

As refresh rates go up, the benefits of using "high dpi, high poll, low sensitivity" starts to exceed other muscle memory training techniques on former coarsefeel techniques that had other issues (e.g. increased latency caused by fewer mouse coordinate reports per second during low dpi approaches).
Head of Blur Busters - BlurBusters.com | TestUFO.com | Follow @BlurBusters on Twitter

Image
Forum Rules wrote:  1. Rule #1: Be Nice. This is published forum rule #1. Even To Newbies & People You Disagree With!
  2. Please report rule violations If you see a post that violates forum rules, then report the post.
  3. ALWAYS respect indie testers here. See how indies are bootstrapping Blur Busters research!

daniloberserk
Posts: 37
Joined: 22 Aug 2017, 16:50

Re: What high DPIs are recommended? i'm sooo confused

Post by daniloberserk » 18 Jun 2022, 16:45

TL/DR: Enough DPI is enough. It depends on your overall dexterity, sensitivity, screen size and subjective taste.

--

First thing that should be clarified once for all. Higher DPI doesn't trigger less INPUT LAG, it's just more "reactive".

You're just lowering the threshold for a single count of movement, that's why CPI/DPI is mouse RESOLUTION. Every count will STILL be hardcapped by your polling rate, it doesn't matter how fast you move.

Of course if you're comparing an 800 DPI to a 400 DPI movement at THE SAME SPEED. 800 DPI will REACT faster, but if you're mantaining your overall eDPI, it will just move half the movement. So in the end... Moving from A to B will take the same time.

NO ONE FLICKS A SINGLE COUNT OF MOVEMENT, so it's mostly useless. First onscreen reaction doesn't make sense in this case because every meaningful movement uses dozens of counts.

And that ARE drawbacks using high DPI with low sensitivity that people doesn't talk about. Since you'll be dividing an angle of movement at degrees smaller then a single pixel on the screen, it does create an "shimmering" visual artifact, specially at the center of the screen. Which may be visually annoying for some people. You can even see what I'm talking about on those examples from mouse sensitivity: https://www.mouse-sensitivity.com/forum ... ity-works/

So, while it DOES make the movement more granular, it also introduces an new visual artifact.

And the extra granularity will not raise your precision most of the time. Even hitboxes aren't precise enough to start with...

-

To keep it simple. It's mostly an subjective choice TBH. Use whatever value you feel confortable to use in the desktop while browsing, because going higher then this already shows that you don't have enough dexterity to take any advantage of the extra granularity that higher DPI achieves (and you would ONLY take advantage anyway if you're an high sens player).

DPI value is like most measure tools. If you only need a clock to measure seconds, you don't need an clock that can measure nanoseconds. Both will be equally precise to do it's job.

As every thing that is measurable in the material world. Enough is enough.

---

I'm honestly sick to see this battlenonsense video spreading misinformation. The same guy who spreaded misinformation about AMD Chill, where an ACTUAL developer of the Chill function answered in the comments and he never cared to answer him or to address the bad methodology he was using.

hkngo007
Posts: 33
Joined: 28 Nov 2017, 01:40

Re: What high DPIs are recommended? i'm sooo confused

Post by hkngo007 » 19 Jun 2022, 03:59

Thank you.

hkngo007
Posts: 33
Joined: 28 Nov 2017, 01:40

Re: What high DPIs are recommended? i'm sooo confused

Post by hkngo007 » 19 Jun 2022, 04:10

daniloberserk wrote:
18 Jun 2022, 16:45
TL/DR: Enough DPI is enough. It depends on your overall dexterity, sensitivity, screen size and subjective taste.

--

First thing that should be clarified once for all. Higher DPI doesn't trigger less INPUT LAG, it's just more "reactive".

You're just lowering the threshold for a single count of movement, that's why CPI/DPI is mouse RESOLUTION. Every count will STILL be hardcapped by your polling rate, it doesn't matter how fast you move.

Of course if you're comparing an 800 DPI to a 400 DPI movement at THE SAME SPEED. 800 DPI will REACT faster, but if you're mantaining your overall eDPI, it will just move half the movement. So in the end... Moving from A to B will take the same time.

NO ONE FLICKS A SINGLE COUNT OF MOVEMENT, so it's mostly useless. First onscreen reaction doesn't make sense in this case because every meaningful movement uses dozens of counts.

And that ARE drawbacks using high DPI with low sensitivity that people doesn't talk about. Since you'll be dividing an angle of movement at degrees smaller then a single pixel on the screen, it does create an "shimmering" visual artifact, specially at the center of the screen. Which may be visually annoying for some people. You can even see what I'm talking about on those examples from mouse sensitivity: https://www.mouse-sensitivity.com/forum ... ity-works/

So, while it DOES make the movement more granular, it also introduces an new visual artifact.

And the extra granularity will not raise your precision most of the time. Even hitboxes aren't precise enough to start with...

-

To keep it simple. It's mostly an subjective choice TBH. Use whatever value you feel confortable to use in the desktop while browsing, because going higher then this already shows that you don't have enough dexterity to take any advantage of the extra granularity that higher DPI achieves (and you would ONLY take advantage anyway if you're an high sens player).

DPI value is like most measure tools. If you only need a clock to measure seconds, you don't need an clock that can measure nanoseconds. Both will be equally precise to do it's job.

As every thing that is measurable in the material world. Enough is enough.

---

I'm honestly sick to see this battlenonsense video spreading misinformation. The same guy who spreaded misinformation about AMD Chill, where an ACTUAL developer of the Chill function answered in the comments and he never cared to answer him or to address the bad methodology he was using.
Thank you. It took me a while to understand the misinformation from battlenonsense video but I agree the initial movement "feels" faster from a count at high dpi (say 400 vs 1600 at same edpi) but from point A to B it the same.

I think battlenonsense video was great for generating him views and income etc and average Joe like me will take his video as holy grail. We need ppl like you to spread the word and fix his mess.

Also I find that at a lower dpi it feels alot easier to maintain straight lines. Ie. Maintaining head level which feels better for say CS GO where ground is perfectly flat/even.

However this also might give tracking heavy games a slight preference to slightly higher dpi like Apex but nothing to do with input lag.

Iv used 400 to 16000 dpi (viper 8khz) and i still feel most comfortable at 400-800dpi. I'm about 45cm/360turn.

Post Reply