The mouse movement is calculated as follows (very much simplified for your understanding):

mouse movement counts * sensitivity * (mouseaccel ^ mouseaccelpower)

You're welcome.

lmao

ha i figured i would end up with an answer like that. perfect!

mouseaccelpower affects the distribution of accel between high and low speeds.

Power >1 and <2 amplify the effect of accel on low speeds, and reduce accel at higher speeds
Power 2 is linear. Low and high speeds receive the same amount of accel
Power >2 reduces accel at low speeds and increases accel for high speeds.

when you mention high and low speeds.. is that how fast i'm physically moving the mouse?

What else?

maybe just mouse sensitivity in general?

In this case it is to do with how fast you move the mouse, or more specifically, the count value that the mouse is producing. Higher DPI mice will appear to be moving faster to the game for the purpose of calculating accel.

thanks!

Trying to figure out the difference between mouseaccel and mouseacclepower ... in LAYMEN terms. I've done several searches on this shit and can't figure it out. I see lots of charts and graphs and mathematics being explained. I just want to know what it does.

To oversimplify quite a bit:

Mouseaccel = If greater than 0, then the faster you move the mouse, the greater your sensitivity.

AccelPower different than default and greater than 1 = the faster you move the mouse, the greater or lesser your mouse accel depending on the value.

ok that makes sense. thanks

Related question: Is MouseAccelPower the same thing as mouseAccelStyle in Quake3e (or Q3 or CPMA, wherever it came from)? Cause I love that

cl_mouseAccelPower is a parameter that only affects cl_mouseAccelStyle 0.

cl_mouseAccelStyle 1 offers a different method of using mouse acceleration. It was added to Quake Live by TTimo (http://pastebin.com/TK4FCyu6). It was subsequently removed a year or 2 later.

The ioq3 implementation of it can be seen here
https://github.com/ioquake/ioq3/blob/master/c...put.c#L471

The latest quake3e source I can find looks very similar, except it enforces a lower limit of 0.001 to cl_mouseAccelOffset.

edit: style 1 may have a problem preserving directionality of mouse movement, as X and Y speeds are used for seperate X and Y accel calculations

Thx lots. So I guess (after looking at the graph above) the closest I can get to replicating that in QL would be to decrease sens, increase accelPower massively and maybe cap it with mouseSensCap so I'm not making 360s constantly at high speed mouse movements. Actually, I think I'll try to plot the different formulas and variables to understand this better.

Senscap is a must with accelpower higher than default imo.

You can't exactly replicate accelStyle 1 in QL but I think you can come close depending on the settings. The thing is that style 1 is fps dependent, so if your fps drops by half so does your accel : /

I agree with using senscap at increased power values. Once you get your low speed tuned with power, the high speed is usually way out of control.

Also regarding style 1 in ioq3 and quake3e, it appears as though their implementation is not fps-dependent (unlike when it was in QL), as the mouse speed is calculated using the games frametime.