Originally Posted by nlmiller0015
So does this mean 400 dpi 2 sense is slower than 800 and 1 sense in terms of speed
"Speed" can be confusing way to call it: the actual speed
at which you move your mouse is a factor speed
wise as "speed" is measured in m/s (for mices) (or inch/s).
The 400/2 vs 800/1 example posted by Glymbol on page 3 is to equal the distances (meters or inches) one has to move their mouse to achieve the same angle of view change, e.g. 180° or 360°. But that despite the same physical movement was made, 51,95 cm/360° in his example, the precision that can be achieved is not the same (sub-pixel precision in 3D environment).
To get back on the original subject of this thread "Which DPI is the best". I'm a low sensitivity player on a Quake 3 engine based game (as is DeMS
). I've been lately looking into what best possible configuration for me. Finding out my previous sensitivity i was using was actually too high for precise tracking and aiming (over-aiming & under-aiming). I have lowered it since (with much surprise that my aiming improved because of that).
There are a few limiting factor here for me:
- My max screen resolution is 1680x1050
- I'm using a 'fall back' system which has a low performance graphics card (HD5450). So running a game at 1680x1050 is impossible to get a proper stable FPS (125)
- The resolution at which I can get relatively stable 125 FPS is 800x600
- 125 FPS is a sweet spot for the Q3 based engine game which has a limit of \cl_maxpackets 42 max
- Low sensitivity player
- Out-of-the-box game binary doesn't support raw input, only WM_MOUSEMOVE or DirectInput (which has a higher response latency, so not a good choice ... for me)
I'm using a DeathAdder 3G (1800dpi/cpi), 1000hz polling, default windows and driver sensitivity with Cheese Mousefix installed (for 75Hz since my screen CAN do 75Hz at resolutions of 1280x1024 and below), with an in-game sensitivity of 0.625.
During testing it at 1800 cpi settings, i will hit negative acceleration fairly easily (read: all the time). As confirmed by the calculations on http://www.funender.com/quake/mouse/index.html
At 900 cpi the negative acceleration effect is still very noticeable.
At 450 cpi the negative acceleration effect is less often noticeable, but it does kick in regularly while making 90° ~ 180° twitch movements. Makes the twitch movements precision very 'random' due to the negative acceleration ... or malfunction speed of the mouse -> From ESReality testing of the DeathAdder, it was shown that the DA at 450 cpi does has a lower malfunction speed than compared to at 1800 cpi (link
DeMS pointed me out the existence of the Enotus mouse test. Which according to it, I'm reaching speed well above the 4.4 m/s malfunction speed
. Albeit several people doubt (including DeMS himself) about the accuracy of Enotus mouse test.
But let's assume the numbers are right. According to the calculations, i need a 200 cpi mouse at a sens of 5.5555556 (for WM_MOUSEMOVE input based games)? But at which I'm probably going to lose the sub-pixel precision the 1800 cpi settings gives me right now for tracking? Or is sub-pixel precision only matters if i were using raw input since that is capable for high resolution input?Edited by Nexu - 1/29/12 at 12:30am