using zoomsens 0 it converts the accel using the old broken algorithm

but with zoomsens 1 it wont touch the accel and will convert the sens and any number above and below one besides 0 gets multiplied by the conversion result



cg_zoomsensitivity 0 (old broken algorithm):
zoomsens = sens * 2 * arctan[ 0.75 * tan[cg_zoomfov/2] ] / 75
cg_zoomsensitivity 1 (corrected version):
zoomsens = sens * arctan[ 0.75 * tan[cg_zoomfov/2] ] / arctan[ 0.75 * tan[cg_fov/2] ]


thanks to injx for replying on the QL forums.


nice boxxy pic

also algorithm in the link is wrong but is consistent and and would work nicely

any idea if accel should be converted with the same algorithm or just left untouched?

it shouldn't really but i use accel with my zoom script that uses zoomsens 0 and i use a small amount of accel (well small to me) 0.15 @800 dpi with sens 2.6 and it changes inconsistently but feels decent to me



but using a consistent algorithm with medium to high amounts of accel especially the one you linked in the thread would give big accel value jumps

while the QL zoomsens 1 algorithm seems to have slightly less exagerrated values to the algorithm you linked

but its all up to feel

an example

0.15*tan(100/2) / tan(90/2) = 0.178763038


0.15*tan(90/2) / tan(90/2) = 0.15 (same value)

this is a correct algorithm since it give consistent values and this also takes you fov in account which is one reason why its consistent


now the QL algorithm give consistent values but they are lower and less exagerrated

example:


0.15 * arctan[ 0.75 * tan[100/2] ] / arctan[ 0.75 * tan[90/2] ] = 0.170020116


0.15 * arctan[ 0.75 * tan[90/2] ] / arctan[ 0.75 * tan[90/2] ] = 0.15





and here are some increased values of 1 accel to show you how big a difference between them


1 * arctan[ 0.75 * tan[100/2] ] / arctan[ 0.75 * tan[90/2] ] = 1.13346744

1*tan(100/2) / tan(90/2) = 1.191753593

but the difference get bigger with higher numbers



now if were to convert accel with my zoom i would go with QL's algorithm cause its values are slightly lower but i wouldn't recommend it because it would change you feel for every weapon completely unlike converting the sensitivity .


TL;DR
even though i convert my accel with zoom ( mind you because of zoomsens 0) i still wouldn't recommend converting it cause it changes you feel for the weapon completely but you can get used to it proof of that is fox in Q3 and noctis

but if you were to convert it i would say use the same algorithm you zooming with and QL's algorithm would be the best in my opinion

good luck

great explanation thank you mister :3

welcome

don't use zoom and just change fov and pitch and yaw values.

why change pitch/yaw at all if i dont need to have different sens/accel at x/y axis?

these re only simple dividers so it doesnt matter whether i half the pitch/yaw or half the sens and accel

i was thinking actuslly about zooming manually by vstr script (press space go fov 40, sens x, accel y / release space go to the previous default fov/sens/accel)

what do u mean exactly please explain

you effectively have +vstr these days. Check out noctis's config at www.ncts.at

He is using the ancient alias command, I don't know when this previously eradicated command came back into ql, or has it been there from the start and only just been noticed? iirc it was in q1 and q2, and got removed for the quake 3 engine, and was replaced by the vstr command (not sure why).

And echo is back! Well, via the print command, when did that appear?

I know its a bit irrelevant, but its been annoying me :(

they re-introduced those with one of the latest updates, iirc the ql goes gold one.

messing around with your m_yaw and m_pitch setting mess with the the amount of real accel you get
http://www.esreality.com/?a=post&id=1370847

just so there's no confusion, the zoom factor multiplies BOTH the sens and the accel in the same way. (and this is certainly how it should be).

this is the case for both the old q3 algorithm (zoomsensitivity 0) and the corrected version (zoomsensitivity 1).

what both these algorithms do is to calculate the ratio of vertical zoomfov to vertical un-zoomed fov. however the q3 algorithm assumes your vertical un-zoomed fov is 75degrees (equivalent to a horizontal fov of 90 [ish] at 4:3). the corrected version does it properly.

yeah and thats why i stick to 4:3 ratio resolutions mainly because my zoomscript has me zoomed all the time
.

aspect ratio doesnt affect your zoomsens in QL, it only depends on the values of cg_fov and cg_zoomfov.

nah I meant in Q3

I just use a 4:3 aspect ratio resolution in both games for consistency since i use Q3's zoom algorithm

bump, after switching to widescreen I noticed that it does and seems like noone noticed this before

i.e.
set cg_zoomfov same as cg_fov
set cg_zoomsensitivity 1
on 16:9 when zoomed your sens is faster
on 4:3 it's the same

reported to syncerror

We could use a sample calculation of accel at least with the zoomsens 1 algorithm

you mean an example?

zoomSens = sens * k
zoomAccel = accel * k

where

k = arctan[ 0.75 * tan[cg_zoomfov/2] ] / arctan[ 0.75 * tan[cg_fov/2] ]

e.g.

fov = 120
zoomfov = 60

gives

k = 0.45

if you dont know how to use trig functions, google will do it for you: google

" fov = 120
zoomfov = 60
gives
k = 0.45 "
i wonder about this magic k, why its not just 0.5 in this case?does it suppose to give the feeling of a bit lower sens while zooming?if so the algoritm should be much simplier

Hey, injx!

What calculation wich I can I use for to added an amount of accel ideal for my current sensitivity settings?

-eg:
if I have a mouse with 400 dpi and Sens 5, then how much should I put in my "cl_mouseaccel"?

Or

If I have a 1000 dpi mouse than does 10cm (cm/360) and with sensitivity 4.2, without accel?

thanks in advance