How sensitivity works

[Drimzi]

10 minutes ago, gbtg said:

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!

edit: also, the calculator does not let me change the WPS for pubg. could the reason be that pubg uses something similar to rawinput from csgo? so i am stuck with the high 2d sensitivity i would be ending up with if i lowered my sens and increased my dpi accordingly?

Benefits don't cap out, but I wouldn't go too extreme as mice are inconsistent at incredibly high DPI, and it can cause conversion issues for some games since many have ungenerous sensitivity sliders or suffer from negative acceleration if the sensitivity value is too low.

WPS is bypassed, that is the reason why you lower the game sensitivity instead.

If 1600 DPI is your preferred 2D sensitivity, then increase to 3200 DPI and set 4/11 WPS. In the calculator, raise the DPI to 3200, halve CSGO's sensitivity, and then convert to PUBG. I would usually recommend 3/11, but not in your case since 1600 is a pretty high baseline, and 6400 DPI is pretty extreme for the majority of cheap mice.

[gbtg]

34 minutes ago, Drimzi said:

Benefits don't cap out, but I wouldn't go too extreme as mice are inconsistent at incredibly high DPI, and it can cause conversion issues for some games since many have ungenerous sensitivity sliders or suffer from negative acceleration if the sensitivity value is too low.

WPS is bypassed, that is the reason why you lower the game sensitivity instead.

If 1600 DPI is your preferred 2D sensitivity, then increase to 3200 DPI and set 4/11 WPS. In the calculator, raise the DPI to 3200, halve CSGO's sensitivity, and then convert to PUBG. I would usually recommend 3/11, but not in your case since 1600 is a pretty high baseline, and 6400 DPI is pretty extreme for the majority of cheap mice.

i just did exactly that and i have to say i think i can feel the difference, not only when slowly moving my mouse as little as possible but it does feel better all around. some of that might be placebo, or not, but i am definitely keeping this. thanks again.

[potato psoas]

On 25/04/2018 at 20:34, gbtg said:

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!

1600 DPI should be fine. 800 is usually even good enough. You probably don't want to go too high either because some games have a sensitivity range that doesn't let you use too extreme DPIs.

[The Archlich]

This explanation is extremely simplified.

I have a few questions:

 

What is a 'dot'? How 'big' it is?

 

I understand that count is a minimal mouse movement that can be registered by machine. Right?

If so, why do we even set up a DPI (or rather Dots per Count would be a proper term), if it lowers the accuracy, while we could simply lower physical mouse travel distance per count what would increase mouse speed without lowering the precision?

 

How to calculate in game degrees per count ratio?

Edited May 21, 2018 by borysn

[The Archlich]

19 minutes ago, borysn said:

This explanation is extremely simplified.

I have a few questions:

 

What is a 'dot'? How 'big' it is?

 

I understand that count is a minimal mouse movement that can be registered by machine. Right?

If so, why do we even set up a DPI (or rather Dots per Count would be a proper term), if it lowers the accuracy, while we could simply lower physical mouse travel distance per count what would increase mouse speed without lowering the precision?

 

How to calculate in game degrees per count ratio?

Actually I just did the testing by myself cause reading this post got me thinking.

I did an experiment which proved that the 'accuracy loss' is at the least negligible (if not not existent).

So as I said I don't know what a 'dot' is, but you also use pixels as measurement (and suggest that high DPI moves cursor multiple pixels per count) so I used pixels as measurement.

So the experiment was simple: I set my mouse to 12000DPI (6 WPS), and guess what... I can still move a single my cursor by a single pixel.

That leads to two possible explanations: Either the DPI scaling works as I suggested and has no impact on 'dots per count', or the 'dots' are so small that incrementing the DPI even by extreme amounts, like hundreds of times has no noticeable effect on pointer precision.

 

Anyway no matter what the truth is I think everyone here looks for the practical explanation. And for that I would say that the limitation with the high DPI is on human side; It's just you who can't make that precise movement; not the machine. So even if there is no technical limitation you still don't want your mouse to move if someone farts in the other room, right?

[Drimzi]

When a mouse detects movement, it sends a count with information on what direction the movement was. These counts get delivered to the PC every time the USB polls. When you configure the 'DPI, or 'CPI', you specify how sensitive the mouse is in detecting movement. The value you set is the expected 'increments' of movement when the mouse is moved one inch.

The 'dot' would just refer to a pixel, since your cursor will increment by 1 pixel for every count.

I'm not sure what you are referring to with a loss of accuracy or precision. That would depend mostly on the person operating the mouse. From a hardware perspective, older mice allowed CPI outside of their recommended range, probably due to marketing, competition, etc., and these values may have undesirable properties like input lag, accel, smoothing, etc, or may just be a completely different sensitivity than what you configured it to. If you got a new mouse with a pmw3366 sensor or one of its variants, then you don't need to worry about this.

To calculate how much you rotate per count, you multiply the sensitivity by the pitch/yaw value, but if the game scales sensitivity by fov, you may also need to multiply this by the angle ratio or tan ratio between the fov and the default fov. An example for Overwatch would be sensitivity * 0.0066 = degree per count. Example for Quake would be sensitivity * 0.022 = degree per count.

[Drimzi]

3 minutes ago, borysn said:

So the experiment was simple: I set my mouse to 12000DPI (6 WPS), and guess what... I can still move a single my cursor by a single pixel.

It is your Windows sensitivity setting that controls how many pixels are incremented per count. The sensitivity of the mouse sensor has no affect on this.

[The Archlich]

4 minutes ago, Drimzi said:

It is your Windows sensitivity setting that controls how many pixels are incremented per count. The sensitivity of the mouse sensor has no affect on this.

I don't understand then. It moves faster, so how wold that not affect it?

If I'm wrong, I want to know why...

Edited May 21, 2018 by borysn

[Drimzi]

16 minutes ago, borysn said:

I don't understand then. It moves faster, so how wold that not affect it?

If I'm wrong, I want to know why...

It becomes faster by sending more increments, not changing how large an increment is.

 

CPI changes the number of increments, game/windows sensitivity changes the magnitude of an increment.

Edited May 21, 2018 by Drimzi

[The Archlich]

34 minutes ago, Drimzi said:

It becomes faster by sending more increments, not changing how large an increment is.

But that means that I'm right. XD

That means that increasing the DPI lowers the mouse travel distance to count.

Edited May 21, 2018 by borysn

[Drimzi]

Yes, your conclusion is correct. I'm stating why, as your question included whether "high DPI moves cursor multiple pixels per count".

[Synd1cate5]

Looking at these and these GIFs, you might think that the lower the sense, the higher the delay in moving a certain distance. That is, if a mouse with a sense of 1 pixel per count will pass a pixel in 1ms at 1000 hz, then with a sense of 0.5 it will pass a pixel in 2ms

[fortunate reee]

1 hour ago, Synd1cate5 said:

Looking at these and these GIFs, you might think that the lower the sense, the higher the delay in moving a certain distance. That is, if a mouse with a sense of 1 pixel per count will pass a pixel in 1ms at 1000 hz, then with a sense of 0.5 it will pass a pixel in 2ms

Spoiler

We have to dig Deeper

 

Edited July 8, 2023 by fortunate reee

[DPI Wizard]

15 hours ago, Synd1cate5 said:

Looking at these and these GIFs, you might think that the lower the sense, the higher the delay in moving a certain distance. That is, if a mouse with a sense of 1 pixel per count will pass a pixel in 1ms at 1000 hz, then with a sense of 0.5 it will pass a pixel in 2ms

Moving a certain distance is of course slower with a lower sensitivity, but there's no delay.

[RedX]

On 7/8/2023 at 9:44 AM, Synd1cate5 said:

Looking at these and these GIFs, you might think that the lower the sense, the higher the delay in moving a certain distance. That is, if a mouse with a sense of 1 pixel per count will pass a pixel in 1ms at 1000 hz, then with a sense of 0.5 it will pass a pixel in 2ms

You can increase your CPI as you lower your sensitivity to maintain the same sum of speed and distance per inch. 

As an example with enhance pointer position off

400 CPI * 14 WPS (2) = 800 DPI
800 CPI * 10 WPS (1) = 800 DPI
1,600 CPI * 5 WPS (.5) = 800 DPI

With all these combinations, moving the mouse 1 inch results in 800 DPI of movement in windows. 

Older mice can have significantly more drift at higher speed, and with higher CPI. But that doesn't seem to be an issue with newer sensors like a focus+ . Which seems to have 90%+ accuracy at low, or high speeds of movement, even with a high CPI. 

On 5/20/2018 at 11:54 PM, borysn said:

So the experiment was simple: I set my mouse to 12000DPI (6 WPS), and guess what... I can still move a single my cursor by a single pixel.

I know this is old, but I use to wonder about this.

WPS 6 is a multiple of .6. So what happens is you move your mouse 2 CPI, and you see the cursor move 1 pixel. And you brain doesn't realize you moved 2 CPI. Your brain just thinks you've moved 1 CPI.  The results is the same at any CPI.  

If you want to see skipping. What you'd want to do is set a multiplier that's 2 or greater. For instance 16 WPS is a multiplier of 3. Which makes it very easy to see pixel skipping in windows. And again, this is going to happen regardless of CPI.

All of this is CPI to DPI. It's different in FPS games, because your going from CPI to Rotation. You can't have half a pixel, but you can have half a degree. And at greater distances, even a half of a degree can be a significant difference. As you lower your rotation, can increase your CPI to maintain the same speed, and sum of rotations. 

 

Edited July 30, 2023 by RedX