Error in the Calculator (Source Engine, not sure about others)

[Luckylock]

Hello DPI Wizard, I found through some testing that the calculator is wrong somehow.

 

Game -> Counter-Strike: Global Offensive

 

According to the math, these two should be the same overall sensitivity (distance travelled with the mouse to achieve same rotation)

1600 DPI + 1 in-game sensitivity

400 DPI + 4 in-game sensitivity

 

But in fact, they are not the same. The math is wrong. The same distance travelled with the mouse does not equate to the same rotation, you can try it out yourself. 1600 dpi + 1 in-game is effectively a lower sensitivity than 400 dpi + 4 in-game. It's somewhat close, but it's definetly not accurate.

 

Perhaps that's why pro players stick to 400 DPI. If they tried to switch to a higher DPI / lower in-game sens, the sensitivity feels "wrong" because it is actually different.

 

Extra notes;

- m_rawinput 1 or 0 doesn't change anything (using 6/11 windows sens and markv fix)

- the mouse sensor isn't the issue

- getting way above 300 FPS

- test was done by 2 friends and confirmed by both

- other dpi / in-game sens settings were tested

- test was done by swiping mouse from left to right from one end of the mouse pad to the other (fast, medium and slow speed). I got consistent results using the Logitech G402 and the razer abyssus.

 

I ordered a Logitech g303 (best sensor money can buy), it's coming monday. I'll do a new test with the mouse to confirm that the sensor really isn't the issue here.

Edited December 18, 2015 by Luckylock

[DPI Wizard]

1600 DPI + 1 in-game sensitivity and  400 DPI + 4 in-game sensitivity is exactly the same, just tested to verify nothing has changed.

Tested with both scripting and physically moving the mouse.

The math for Source games is well known, Counts/360 = 360/(SENSITIVITY*0.022).

1 = 16363.64 counts/360

4 = 4090.91 counts/360

 

If you are using Logitech, you can easily verify this by running these script (macro assigned to G1 on M1 on the keyboard), make sure raw input is on:

For sensitivity 1, moving 16362 counts (shortened a little over 1 count to make the macro run faster)

if (event == "G_PRESSED" and arg == 1) and GetMKeyState() == 1 then
	for i = 0, 161 do
		MoveMouseRelative(101,0)
		Sleep(5)
	end
end

For sensitivity 4, moving 4090 counts (shortened a little under 1 count to make the macro run faster):

if (event == "G_PRESSED" and arg == 1) and GetMKeyState() == 1 then
	for i = 0, 408 do
		MoveMouseRelative(10,0)
		Sleep(5)
	end
end

Are you sure you have disabled acceleration?

 

How much off are your measurements?

[Luckylock]

Acceleration is in fact completely disabled.

 

I made a video for you. My mouse pad is the Steelseries QCK Heavy

 

https://youtu.be/6_BBnW12IyI

Edited December 18, 2015 by Luckylock

[DPI Wizard]

Acceleration is in fact completely disabled.

 

I made a video for you. My mouse pad is the Steelseries QCK Heavy

 

https://youtu.be/6_BBnW12IyI

I see, thanks for the video.

 

Could you try a couple of things;

For this, find the movement you need to do a 360 with both 400DPI/sens 4 and 1600 DPI/sens 1.

Now with 400 DPI and sens 1, check that you need to lift and swipe 4 times to get a perfect 360, and with 1600 DPI and sens 4 see that you do exactly 4 360's with one swipe.

 

Also, have your tried the DPI Analyzer at both 400 and 1600 DPI and checked if the real DPI discrepancy is the same? There is something going on either between the mouse and the game, or some 3rd party application interfering. 

 

Maybe you can try with vsync on?

[Luckylock]

I don't have enough mousepad space to do 4 360's at that kind of sensitivity. Same goes for the DPI analyser, I don't have enough monitor space to properly test it...

 

I had 2 of my friends test this and they confirmed, I am not sure how you are getting accurate sensitivity. I will test with the g303 (best sensor on the market) in a couple days. There's no third party application interfering, I know what is running on my computer.

 

As far as vsync is concerned, rawinput is on, so framerate shouldn't affect the sensitivity at all. I highly doubt that is the problem.

[DPI Wizard]

Same goes for the DPI analyser, I don't have enough monitor space to properly test it...

You should be able to keep dragging the cursor beyond the edge of the screen with the analyzer, it might depend on your browser though.

 

EDIT: I might need to update something to make that work again actually...

[Luckylock]

After thinking a bit, I see two possible reasons for the difference in sensitivity.

 

Either the the value in the algorithm of the calculator is wrong (I doubt it at this point), or the game is missing reports / counts on higher DPI. In this case it would make sense, since the sensitivity is in fact lower and the problem (difference) is worse the higher you go in DPI.

 

I wonder if it's only in the source engine or other games as well.

[DPI Wizard]

or the game is missing reports / counts on higher DPI. In this case it would make sense, since the sensitivity is in fact lower and the problem (difference) is worse the higher you go in DPI.

I think this is the case, however I haven't seen any issues myself, going as high as 8200 DPI with 1000 Hz. What polling rate are you using? Maybe try going to 125 Hz if you haven't already, and see if the results are the same.

[Luckylock]

Actually I just tested 4000 dpi and 0.4 in-game. It's a higher sensitivity than 1600 dpi 1 in-game but still lower than 400 dpi 4 in-game

 

It seems to be back and forth (I've tested other settings rigorously). I don't think it's missing counts / reports anymore, I'm starting to think rounding errors.

 

Anyway, this is beyond my knowledge and testing abilities at this point. I'll just do like the pros and stick with 400 dpi master race. I'm tired and need sleep.

Edited December 18, 2015 by Luckylock

[gourami]

maybe it's your mouse? i just tried this with a g303. 4@400 and 1@1600 moves exactly the same distance

[Luckylock]

Yea it doesn't do that with the g303.

 

the g402 is shit.

[Xbye]

Can confirm. I was his friend who helped verify as well using my G402. Inconsistency between the sensitivities/DPI while using it. Currently using the G502 as of a few minutes ago... tested and the difference in distance is so small it might as well be human error.

 

G402 sensor is pretty balls it seems.

[DPI Wizard]

I didn't realize the sensor could fluctuate so much with the DPI. They are usually off compared to the configured DPI, but this has been a flat percentage regardless of DPI on the mice I've tested.

 

Thanks for the heads up

[Xbye]

I didn't realize the sensor could fluctuate so much with the DPI. They are usually off compared to the configured DPI, but this has been a flat percentage regardless of DPI on the mice I've tested.

 

Thanks for the heads up

I remember reading that percentage somewhere but lost it. For knowledge's sake, could you give me the average % off when using same distance but different counts?