Sensitivity and Keyboard scaling on fov

[Baker]

FOV ...

Scale pitch sensitivity for mouse (mouse y delta) and the keyboard pitchspeed by multiplying it by fov_y/90.
Scale yaw sensitivity for mouse (mouse x delta) and the keyboard yawspeed by multiplying it by fov_x/90.

Or is this wrong to do? Because it ties input speed to the video mode (since fov_y is calculated in a way that depends on the aspect ratio).

Or maybe just do it for FOV < 90 (so that unreasonably small fov get scaled, but fov that someone might actually use like 90-120 are unaffected). And only use fov_x/90 for the multipler for both to keep the video mode out of the equation (I mean wouldn't it suck to have a nice setup just to find that if you changed to a video mode with a slightly different aspect ratio it messed with your aim).

[taniwha]

How about this: modify the command line parser to support math and cvar substitution (even a second command line language), and provide a cvar with the sole purpose of reporting the fov scaling as calculated by the engine. Then the user can use that value in fov changing scripts.

[Spike]

dp has a sensitivity scaler field, sent as a stat so the client can see it.
a big problem with automagic scaling is mods that do it automatically anyway. now you can't change your angle at all when zoomed in.

if its for users, just let the user set it when they change the fov.

[Baker]

[metlslime]

doesn't quake ship with a zoom alias (bound to f11) that does this? Perhaps i'm mistaken....

I know one thing I did in my autoexec.cfg was change the alias to modify m_pitch and m_yaw instead, so that I could experiment with different sensitivities without it being reset every time i zoom in/out.

[Baker]

[mh]

How about if the fov change came from svc_stufftext then don't auto-scale, otherwise assume that it's either from the user or a cfg file and do.

I wouldn't auto-scale by changing the value of sensitivity (that's evil), just add a scaling factor to the calculations. But beware of ints in the input subsystem.
I definitely wouldn't be inclined to add a second text parser either. It seems as though every new subsystem that's added gets it's own text parser and it's beyond the point of silliness.

Agree with not auto-scaling for fov > 90 too.