perfect sensitivity, a question to developers / coders

23 replies [Last post]
Cyrax
Offline
Joined: Apr 2009
Posts:
Re: perfect sensitivity, a question to developers / coders
animalchik wrote:

but the view angle steps depends on resolution? aren't they?

And fov ( - opengl perspective projection settings)

animalchik wrote:
so there is 1024 steps in 90degres of horizontal movement, is that correct?

No, pixels on screen have non-linear dependency from view angles

animalchik wrote:
crosshair moved 1px

Enable antialiasing and you will not see that Happy

I am proud of spreading a pirated Excessive Plus version and claim to be the original author, yay!

GG
GG's picture
Offline
Joined: Apr 2009
Posts:
Re: perfect sensitivity, a question to developers / coders

All these reasonings I think in will not result. Q3 engine is too complex. I will add only the one that is confirmed by experience: my shots are exact only when values of variables mpitch and myaw come to an end with a number in the period and have 6 signs after the comma.

✽ ✾ ✿ ❀ ❁ ❃ ❋
¡Adiós!
\'o♥o'/

.:Green:.-=Girl=-

animalchik
fala.q3's picture
Offline
Joined: Jul 2009
Posts:
Re: perfect sensitivity, a question to developers / coders
Cyrax wrote:

animalchik wrote:
but the view angle steps depends on resolution? aren't they?

And fov ( - opengl perspective projection settings)

animalchik wrote:
so there is 1024 steps in 90degres of horizontal movement, is that correct?

No, pixels on screen have non-linear dependency from view angles

animalchik wrote:
crosshair moved 1px

Enable antialiasing and you will not see that Happy

if we do playing in 1024px horizontal resolution and have cg_fov 90 how many view angles that is?

and then the same settings but cg_fov 110?

btw. even if i have AA:16 i can easily notice crosshair moves one px

Was trying to figure out how the engine count, and how many steps is in view angle while having fov 90 and fov 110, i was convinced the amount of steps (precision/resolution of move) depends from r_mode and in our situation it will be 1024 steps in fov 90, and it seems, linear or not, these things are connected

I am addicted to life.

JosNobody
Offline
Joined: Jul 2010
Posts:
Re: perfect sensitivity, a question to developers / coders

I've found this

http://www.esreality.com/?a=post&id=1638450#pid1638450

R = ( pi * W ) / ( I * tan[ F / 2 ] )

where

W = screen resolution width

I = real sensitivity (distance per 360 turn)

F = horizontal fov

R = mouse resolution required

.xls is here http://www.esreality.com/?a=post&id=1681607#pid1681607

but when we have fixed DPI on mouse I think we should play only with our m_yaw, m_pitch and sensitivity so:

I = 360 / ( M * S * DPI )

from 1st equation

I = ( pi * W ) / ( DPI * tan[ F / 2 ] )

then

360 / ( M * S * DPI ) = ( pi * W ) / ( DPI * tan[ F / 2 ] )

M = ( 360 * tan[ F / 2 ] ) / ( pi * W * S )

where

S = Sensitivity in game ( Windows sens default )

M = maximum values of m_yaw and m_pitch to get 1px precision for given S

OK, injx said it's correct

BOZO
Bozo's picture
Offline
Joined: Feb 2010
Posts:
Re: perfect sensitivity, a question to developers / coders

I've also used this formula and to me it seems quite accurate.

I = 360 / (0,03 * 157,48 * 22) = 360 / 103,9368 = 3,46 cm/360°
J = (0,03 * 157,48^2 * 0,75) / 1000 =0,56 accel

Before I did this calculation I used a ruler Big grin to measure the move when I'm doing a 360 and I came to the same result.

Btw here is the full thread where he explains his formula a bit:
http://www.esreality.com/?a=post&id=1370847

3M'darkangel
darkangelspa's picture
Offline
Joined: Feb 2010
Posts:
Re: perfect sensitivity, a question to developers / coders

i will re read again this topic and also that website Tongue and make a tests....as now i cant find a sensitivity for my new mouse :'(


Gloria Mōmentum
Talent wins games, but teamwork and intelligence wins championships.”



“There is no hunting like the hunting of man, and those who have hunted armed men long enough and liked it, never care for anything else thereafter.”
—Ernest Hemingway

XV!BENEDYKT
xvibenedykt's picture
Offline
Joined: Jul 2008
Posts:
Re: perfect sensitivity, a question to developers / coders
9d1c65341c2e4ed7 wrote:

JosNobody wrote:

= 3,46 cm/360°

OMG i was playing with that rule: 36cm/360°.
but now with new(shit) mouse i play with 18cm/360° cuz that mouse is interrupting and loseing motion trajectories

JosNobody
Offline
Joined: Jul 2010
Posts:
Re: perfect sensitivity, a question to developers / coders

my post was wrong

there is no such thing as pixel precision in FPS games

from esreality.com

"you're right, it's just a model (only we have?)
with default windows sens and no mouse driver scalling m_yaw/m_pitch are elemental angles that 1 mouse tick rotates the scene we see and ingame sensitivity is simply a multiplicator of those values

so we should calculate m_yaw/m_pitch - values should be smaller than angle between lines connecting observer with edges of object (like enemy head) placed in maximum distance we encounter in particular game that would be represented by at least 1px at a given resolution
right?

without good knowlege of game engine that's impossible
and maybe mouse DPI would matter if m_yaw/m_pitch were really small :]

in conclusion: set all settings way they fits you and look at configs of best aimers :]"

animalchik
fala.q3's picture
Offline
Joined: Jul 2009
Posts:
Re: perfect sensitivity, a question to developers / coders

there it is a pixel precision Happy Yours assumption is wrong Happy but the topic is more about in game counts than pixel precision

as u already know pointer jumps one view angle = one pixel
and because the steps on the borders of the screen are different there are (3.14 * horizontal_resolution) steps in 90degres of pov

in game multiplets affects the counts, and this topic is mainly about this

compare the mouse feeling on m_yaw m_pitch 0.022 to 0.0011 or even 0.01 if u set multiplier of sensitivity to the way mouse going to made exactly the same distance there will be feeling of the mouse change only

the main thing witch was interesting was the method of calculations used in game witch had influence on the mouse feel

I am addicted to life.

[h]iper[s]low
etkq3's picture
Offline
Joined: Jul 2007
Posts:
Re: perfect sensitivity, a question to developers / coders

Here you have code responsible for mouse movement:http://pastebin.com/v0LfetG8

So if we make assumption that accel = 0 then view angles are calculated in this way:angley = mx * some_const_based_on_current_fov * sensitivty * m_yaw

Simple equation there may be only loss in accuracy when rounding number but it should not matter what sens/yaw you actually use. Scaling that depends on fov is probably to make sensitivity feel same way between different fovs. m_yawspeed, m_pitchspeed it’s for keyboard handling (epsi).
And yes 1 cpi can be 6 pixels based on this equantions and there is 3rd multiplayer as you said but evenwithout it you would not have 1px jump it need to be scaled for proper angles.

Acer Aspire 5920G
Razer Deathadder
Bayerdynamics DT-990 Pro