How sensitivity works

[DPI Wizard]

30 minutes ago, gbtg said:

hey, i got a question regarding this, but i am not sure if there is a definitive answer to it:

the vast majority of csgo pros use 400dpi. why do they not increase their dpi and reduce their ingame sens accordingly, in order to gain accuracy and smoother movement? is there such a thing as "too much accuracy", at which point having your mouse pick up on the tiniest of movement becomes a disadvantage?

They usually also have quite a long 360 distance, so they are sub-pixel accurate. Accuracy below 1 pixel should be the minimum imo, but there really is no "too much accuracy", but at a certain point the benefit is only theoretical. I plan to add a pixel accuracy calculation in the future.

 

[Skwuruhl]

The reason pros don't use super high DPI is because it simply doesn't matter. The amount of skipping you get is negligible enough to the point that you'd literally never notice it while playing.

Take this video for example: 

 

Look how tiny his mouse movements have to be to even be able to barely see the pixel skipping. This same pro player even switched back from using high DPI because it just doesn't make a difference.

 https://www.reddit.com/r/Competitiveoverwatch/comments/66ce07/why_is_taimou_using_400_dpi_x_55_sens_if_it_skips/

 

 

Edited October 13, 2017 by Skwuruhl

[Drimzi]

Edited January 28, 2018 by Drimzi

[xtendlife]

Where in the calculator do i find out pixel count?

[DPI Wizard]

2 minutes ago, xtendlife said:

Where in the calculator do i find out pixel count?

Pixel count is not in the calculator actually. I was looking to add it long ago but ran into some issues. The calculator is more suited for this now, so I'll see if I can add it in an upcoming version.

[xtendlife]

Just now, DPI Wizard said:

Pixel count is not in the calculator actually. I was looking to add it long ago but ran into some issues. The calculator is more suited for this now, so I'll see if I can add it in an upcoming version.

Ok thank you, but do you know how to get 1:1 using 1920x1080, 103 fov, 800 dpi and 24.5 monitor size in overwatch?

[DPI Wizard]

Not quite done with the math yet, I did BF3 in the example manually. So I don't have an answer yet, but it's likely in the 12-17 range.

[xtendlife]

Just now, DPI Wizard said:

Not quite done with the math yet, I did BF3 in the example manually. So I don't have an answer yet, but it's likely in the 12-17 range.

Ok thank you, do you have a recommendation which sensitivity to use in the meantime? Windowmaker scope sensitivity as well?

[DPI Wizard]

2 minutes ago, xtendlife said:

Ok thank you, do you have a recommendation which sensitivity to use in the meantime? Windowmaker scope sensitivity as well?

Do you have any game to base it off of, like CSGO, Battlefield or PUBG? If not, it's quite hard to recommend it, as people prefer anything from 2 inch 360 to 35 inch 360.

You could alternatively try to base it off of your desktop sensitivity by using Windows as the input game, and Overwatch as the output.

[xtendlife]

2 minutes ago, DPI Wizard said:

Do you have any game to base it off of, like CSGO, Battlefield or PUBG? If not, it's quite hard to recommend it, as people prefer anything from 2 inch 360 to 35 inch 360.

You could alternatively try to base it off of your desktop sensitivity by using Windows as the input game, and Overwatch as the output.

No sorry, this is my first game dealing with mouse sensitivity. Which setting should I use for windows desktop to overwatch? Monitor distance, 360, or viewspeed?

[DPI Wizard]

Try Monitor Distance 100% first. If it feels too fast, try a lower match %.

[xtendlife]

1 minute ago, DPI Wizard said:

Try Monitor Distance 100% first. If it feels too fast, try a lower match %.

Ok, thank you for being patient with me. Very new to all this stuff.

[DPI Wizard]

1 minute ago, xtendlife said:

Ok, thank you for being patient with me. Very new to all this stuff.

No problem!

[WhoCares?]

12 hours ago, DPI Wizard said:

Try Monitor Distance 100% first. If it feels too fast, try a lower match %.

Isn't it the opposite? 100% is the slowest and 0% the fastest?

Edited December 1, 2017 by WhoCares?

[DPI Wizard]

28 minutes ago, WhoCares? said:

Isn't it the opposite? 100% is the slowest and 0% the fastest?

Indeed, mixed it up sorry 

[potato psoas]

On 10/14/2017 at 4:56 AM, Skwuruhl said:

The reason pros don't use super high DPI is because it simply doesn't matter. The amount of skipping you get is negligible enough to the point that you'd literally never notice it while playing.

Take this video for example: 

 

Look how tiny his mouse movements have to be to even be able to barely see the pixel skipping. This same pro player even switched back from using high DPI because it just doesn't make a difference.

 https://www.reddit.com/r/Competitiveoverwatch/comments/66ce07/why_is_taimou_using_400_dpi_x_55_sens_if_it_skips/

I notice the pixel skipping using 400 DPI and it has thrown off my aim plenty of times. You have to use at least 800 DPI. Unless you use a really low sensitivity already.

I think the pros use low DPI because it is more predictable, if you know what I mean.

Edited December 16, 2017 by potato psoas

[WhoCares?]

I might be wrong, but you should use at least:

((horizontal resolution) x (360 / hfov))/ Inches for 360°= min. DPI 

 

But I find that a bit jittery too, that might be because of shannon's law (?)

Someone here in the forum came of with shannon's law...if thats applies to mice too you should take your min DPI times 4 



 

Edited January 15, 2019 by WhoCares?

[Skwuruhl]

15 hours ago, WhoCares? said:

I might be wrong, but you should use at least:

((horizontal resolution) x (360 / hfov))/ Inches for 360°= min. DPI 

 

But I find that a bit jittery too, that might be because of shannon's law (?)

Someone here in the forum came of with shannon's law...if thats applies to mice too you should take your min DPI times 4 

 

Even the person who made the shannon's law tool for overwatch is skeptical of if shannon's law applies to mice or not. Afaik general consensus is that while stupid high in-game sensitivity causes pixel skipping it has nothing to do with shannon's law.

[IceKK]

Hey i'm a bit confused.

The whole markc 1-1 mouse fix thing is still relevant for windows and games that don't support raw input right? But wouldn't it technically be better to run low windows sensitivity and high dpi with the benifits shown in OP? As shown there high dpi with low sensitivity ingame feels much smoother.

 Why is 1-1 recommended in windows, but clearly not better in games?

[Drimzi]

9 hours ago, IceKK said:

Hey i'm a bit confused.

The whole markc 1-1 mouse fix thing is still relevant for windows and games that don't support raw input right? But wouldn't it technically be better to run low windows sensitivity and high dpi with the benifits shown in OP? As shown there high dpi with low sensitivity ingame feels much smoother.

 Why is 1-1 recommended in windows, but clearly not better in games?

MarkC fix essentially makes the 'Enhance Pointer Precision' (EPP) option useless. It doesn't do anything if you already have that option unchecked. Some games like to enabled EPP automatically, like Half-Life 1, so that's when the MarkC fix comes in handy. You are not forced to use 6/11 or anything like that.

For 2D movement, the lowest movement increment possible is 1 pixel, anything higher is going to require inconsistent movement and pixel skipping. A lower sensitivity in Windows will ignore mouse counts, but will do so in a consistent manner that makes it unnoticeable, with the end result being similar to reducing the CPI in your mouse. Any setting below 5/11 is perfectly fine to use. Also, games will usually still use the Windows sensitivity settings to handle their 2D movement, even if they have Raw Input support.

For 3D, you want the angle increment to be infinitely low, as the benefit is endless. The only way to achieve a lower increment is to reduce the game sensitivity, but in doing so, it also requires more CPI to preserve the physical distance for rotation. This also affects the 2D sensitivity, so this is where the Windows sensitivity comes in handy. For convenience sake, I highly recommend using 3/11 with 4x higher CPI if you have a decent mouse with a good sensor. This will allow you to use a game sensitivity low enough for smooth angular increments.

 

[IceKK]

27 minutes ago, Drimzi said:

MarkC fix essentially makes the 'Enhance Pointer Precision' (EPP) option useless. It doesn't do anything if you already have that option unchecked. Some games like to enabled EPP automatically, like Half-Life 1, so that's when the MarkC fix comes in handy. You are not forced to use 6/11 or anything like that.

For 2D movement, the lowest movement increment possible is 1 pixel, anything higher is going to require inconsistent movement and pixel skipping. A lower sensitivity in Windows will ignore mouse counts, but will do so in a consistent manner that makes it unnoticeable, with the end result being similar to reducing the CPI in your mouse. Any setting below 5/11 is perfectly fine to use. Also, games will usually still use the Windows sensitivity settings to handle their 2D movement, even if they have Raw Input support.

For 3D, you want the angle increment to be infinitely low, as the benefit is endless. The only way to achieve a lower increment is to reduce the game sensitivity, but in doing so, it also requires more CPI to preserve the physical distance for rotation. This also affects the 2D sensitivity, so this is where the Windows sensitivity comes in handy. For convenience sake, I highly recommend using 3/11 with 4x higher CPI if you have a decent mouse with a good sensor. This will allow you to use a game sensitivity low enough for smooth angular increments.

 

Thanks for replying!

I see, so i actually got it right the first time around i think. I calculated that in order for me to get my favourite sensitivity of (WPS:6 - DPI:1600)  at the highest possible dpi i would have to settle for (WPS:3 - DPI:6400) as 6400 x 0.25 is 1600. But then i tested it with the mouse recorder software and discovered all the skipped pixels where the cursor doesn’t move at all, unsurprisingly, because of it being lower than 1-1. So i bailed and wen’t back to standard 1600.

Am i understanding it correctly that those values would be better in games and equal in windows? (Sorry about my english)

Edited March 9, 2018 by IceKK

[Drimzi]

1 hour ago, IceKK said:

Thanks for replying!

I see, so i actually got it right the first time around i think. I calculated that in order for me to get my favourite sensitivity of (WPS:6 - DPI:1600)  at the highest possible dpi i would have to settle for (WPS:3 - DPI:6400) as 6400 x 0.25 is 1600. But then i tested it with the mouse recorder software and discovered all the skipped pixels where the cursor doesn’t move at all, unsurprisingly, because of it being lower than 1-1. So i bailed and wen’t back to standard 1600.

Am i understanding it correctly that those values would be better in games and equal in windows? (Sorry about my english)

There wouldn't be any skipped pixels. What you saw were skipped counts, which is intentional, otherwise you will be moving 6400 pixels per inch instead of 1600 pixels per inch as you can't increment less than a pixel. Mouserecorder should have highlighted the skipped counts as green, which is ok.

[IceKK]

7 hours ago, Drimzi said:

There wouldn't be any skipped pixels. What you saw were skipped counts, which is intentional, otherwise you will be moving 6400 pixels per inch instead of 1600 pixels per inch as you can't increment less than a pixel. Mouserecorder should have highlighted the skipped counts as green, which is ok.

Yes thats what i meant, didn’t know the terminology. Skipped counts are not an issue? I assume it’s only affecting windows and 2D, but have you noticed any drawbacks?

Edited March 9, 2018 by IceKK

[DPI Wizard]

Games that have issues with acceleration or similar can not be fixed with an external program. Four typical issues I run into with games are:

1. When the sensitivity get low enough, single counts are simply ignored. Very easy to notice with scripts, and reproducible with low DPI and slow movement which result in no or jagged movement in the game.
2. Inconsistent distance moved with single counts. This is the worst one probably, Arma 3 comes to mind that have (or at least had) this issue. When moving back and forth 1 count you obviously expect the crosshair to end up at the same place, but several games have issues with this. When a single count is not registered properly there's something fundamentally wrong with the game. This is not how acceleration typically works either, that's the next issue.
3. Acceleration, meaning the distance moved varies depending on either packet size, report rate or a combination.
   Packet size is when sending 100x1 counts does not result in the same movement as 1x100 counts.
   Report rate is when 100 counts in 100 ms does not move the same as 100 counts in 1000 ms.
4. "Negative" acceleration, which is basically packets dropped.

None of these issues can be fixed with any universal external program, as the issue is what the game engine does with the information from the mouse, not what information it gets.

[gbtg]

On 3/9/2018 at 3:57 AM, Drimzi said:

For 3D, you want the angle increment to be infinitely low, as the benefit is endless. The only way to achieve a lower increment is to reduce the game sensitivity, but in doing so, it also requires more CPI to preserve the physical distance for rotation. This also affects the 2D sensitivity, so this is where the Windows sensitivity comes in handy. For convenience sake, I highly recommend using 3/11 with 4x higher CPI if you have a decent mouse with a good sensor. This will allow you to use a game sensitivity low enough for smooth angular increments. 

i am currently playing with 1600dpi and ~22 ingame hipfire sens in pubg. would you say that lowering the ingame sens even more (and adjusting by increasing dpi) would make a noticeable difference? do those benefits cap out at a certain point or should i just go as low as possible? thanks!

Edited April 25, 2018 by gbtg