following stuff should be set as follows anyway:

in_mouse 2
cl_mouseacceloffset 0
m_yaw / m_pitch 0.022
cl_mouseaccelpower 2

m_cpi 0
unless you plan to switch between mice and different dpi
https://public.sheet.zoho.com/publish/sharpe1337/calc-1
http://www.quakelive.com/forum/showthread.php...l-Features


visual representation of mouse accel settings:
http://ambro.insane.si/accel.html
source: http://www.esreality.com/index.php?a=edit&...7af7a79e9a

Thanks very much, that's very helpful :)

N1. But it looks like the http://ambro.insane.si/accel.html is only showing how cl_mouseaccelstyle 1 is visualised.

there is no style 1
only cl_mouseaccelstyle 0

Just saying that the text below the graph said it was emulating 1.....

He is right, this is old stuff from the time then we had accelStyle command. Although it might still show correct curves, not sure.

sensitivity <factor>
Sets mouse sensitivity.

in_mouse [-1|0|1|2]
-1 = Windows mouse input
1 = Direct input
2 = Raw input
Controls the in-game mouse input.

m_filter [0|1]
Smoothes mouse movement.

m_pitch [value]
Vertical view movement sensitivity.

m_yaw [value]
Horizontal view movement sensitivity

cl_mouseAccel <value>
Mouse acceleration factor.

cl_mouseAccelOffset <offset>
Sets mouse acceleration sensitivity offset.

cl_mouseAccelStyle [integer]
Sets mouse acceleration style.

cl_mouseSensCap <max sensitivity>
Sets mouse sensitivity limit when using mouse acceleration.

+
cl_mouseAccelpower [integer] (Default is 2)
Sets mouse acceleration strength.

how does it differ to cl_mouseAccel ?

The cl_mouseaccelpower is the mutiplier, and the cl_mouseaccel is the amount of accel you want to have.

cl_mouseaccelpower 2 is default though, which most people use, unless you are a real tinkerer :)

There is no accelStyle anymore

http://www.funender.com/quake/files/injx/accel.png

some of the names are different in ql, (limit is sensCap, and m_accel is cl_mouseAccel).

this picture shows a default power (2) curve, which is the same as q3, where sens increases linearly with mouse speed.

accel=0 always disables accel, regardless of the other settings.

m_cpi is a separate feautre, when enabled it gives all the mouse settings 'real' units (centimetres, seconds, degrees, etc.), rather than the arbitrary units of 'counts' which depends on your mouse.

i was thinking the higher the offset is the later the accel starts to kicks in. Is it true?

Look at the graph.. ?

i don't understand it, also where is accelpower in the graph?

accelpower = gradient

no.

accelpower = curvature.

with power = 2 the gradient is constant.
with power > 2 the gradient increases with mouse speed.
with power < 2 the gradient decreases with mouse speed.

hmm, so the acceloffset do the same as cl_mouseaccel?

Look at the graph. When mouse speed passes the offset, accel kicks in (line is diagonal instead of horizontal).

i see now i understant, i think this line should be drawn under it's actual position, it's hard to understand as it is atm.

wrong

So what happens if you set the power to 1.1?

As injx said, you'd take accel to the power of 1/10 (accel would get smaller the faster you move your mouse). You raise the gradient to the power of accelpower - 1.

sens 2; accel .5; power 1.1

sens 2; accel .5; power 1.5

sens 2; accel .5; power 1.9


Complete formula that shows there offset comes into play

Thx, have a bit of a hangover today and it "was" hard to visualise.

Huh? You have the ability to visualize logarithms?

Thought everyone could?

i'm trying hard to understand all this. If i'm not mistaken when you increase cl_mouseaccel the line curve go up isn't it?

Is it possible to make the accel curve looks like this? : http://imageshack.us/photo/my-images/13/esrquestion.jpg/

also in this pic i drew a hatched line, which would represent the same line but with bigger cl_mouseaccel.

cl_MouseAccelPower 3 will produce a parabola like this.

ty. The result feels ingame to not be that, anyway what would looks like the graph of a negative accelpower?

I feel accelpower is a must but no one know how to use it atm.

anyway what would looks like the graph of a negative accelpower?

Negative accelpower is not supported in QL. I think that power has a lowerbound of 1.

A negative accel power would make an inverse parabola, such that at rest your sensitivity would be undefined, then past a certain initial velocity it would be extremely high followed by extremely fast speeds which would lower the sensitivity at an exponential rate.

Graph here: http://www.wolframalpha.com/input/?i=Plot%5B4.%2Fx^2%2C+{x%2C+-1.1%2C+1.1}%5D+

Maybe you meant accelpower < 2, then you will have less accel the faster you move your mouse.

Yes but with accelpower < 2 small slow/medium moves are faster than with accelpower 2.

They aren't (it might look like that on the plots above because the scale is different). The change in accel is more prominent, though.

If you want to start later in the curve, you can play around with the offset.

The higher the offset value the greater the velocity needed until acceleration takes effect.

With the default 0, acceleration has an effect at any velocity.

Negative offset values will start you higher up on the acceleration curve, or linear progression in the case of accelpower 2.

Isn't negative offset pointless in the case of accelpower 2?

Well, it would have the same effect as increasing your base sensitivity.

Are you sure about this?

Well let's see:

B + (A(v-c))^(P-1)

B is the base sens
A is cl_MouseAccel
v is velocity
c is cl_MouseAccelOffset
P is cl_MouseAccelPower

Let's set A, P, B, and v to constants: 0.2, 2, 4, and 0.5 respectively.

I will assume that m_cpi is 0, so that v is in counts/millisecond. To give you an idea, 10 counts/ms is 10000 counts/sec, at 400 cpi (157.48 counts/cm) that represents a velocity of 0.635 meters/sec.

Here I want to demonstrate though that the offset is the same (or extremely close to the same) as increasing base sens when MouseAccelPower is 2, so I want to use a small velocity value. If the mouse is polled at 500 hz, or every 2 ms, and only one count is registered during that time it would represent a velocity value of 0.5 counts/ms (which is 3.1 cm/sec, slow and steady).

I am sure it detects velocities much smaller than this, but since I do not know how it is calculated we will use this as the minimum for now.

That gives us:

4 + (0.2(0.5 - c))^(2-1) =

4 + (0.2(0.5 - c))

Now let's set a negative offset value of say -5:

4 + (0.2(0.5 - - 5)) =

4 + (0.2(5.5)) = 4 + 1.1 = 5.1

The sensitivity starts at 5.1when we reach the minimum velocity.

Next let's use an offset value of 0, but increase the base sensitivity to 5 instead:

5 + (0.2(0.5 - 0))

5 + (0.2(0.5)) = 5 + 0.1 = 5.1

The sensitivity starts at 5.1 when the minimum velocity is reached.

So, yes I'm sure.

It would be really obvious with an interactive graph like the ambro one if it was made to use the modern cvars.

Nicely demonstrated. I am not going to try to get technical cus u lost me a little after "B + (" [which is an ironic symbol] :)

I just supposed that a constant would not change, I had guessed that the feeling that the sensitivity had increased was only because the acceleration was kicking in at an earlier point, to an extent that it was somewhere up the graph...

Ok can you work this out, cus you clearly look more adept at equations than I will ever be!! I just tested -10 offset to see what it was like, with a sens of 1 so that it wasnt too fast. What would be the equivalent sensitivity at 0 offset? I would like to see if it feels similar.

using a negative offset with a power of 2 is the same as using no offset with a base sens of

B + A*(-c).

with a power other than 2, this is not the case (there are cross terms).

here's a picture of what is happening when you set a negative offset:

http://www.funender.com/quake/files/injx/accel-negativeoffset.png

Dammit injx beat me to it by like a half hour

He is substantially better than me at math :-(

Thankyou, so effectively if you want the acceleration to come into play faster then setting offset to 0, or immediately then set a negative value. So if you were using -10 offset and you wanted a sens of 1 at the start, what would you divide your sensitivity by?

As long as you have accelpower set to 2, use offset 0. The only thing you're changing by manipulating it is effectively your base sens. Accel stays the same.

the conclusion you *should* have drawn from my post is that there's no point in using a negative offset with power 2.

are you asking me about powers other than 2?

Well maybe it doesn't seem as black and white today! The thing is that the acceleration curve doesn't really start going up radically until the mouse has been moved over a certain speed. (when set at offset 0.)

Which I have taken from the link given by T1E
http://ambro.insane.si/accel.html

From this curve and your image it looked like the accel was kicking in earlier when you set the offset to a lower value and a negative value?

Might not ever get this, so maybe if the question should be:-

How do I get the acceleration steep part of the curve to come in to play faster, negating or reducing the offset (just base sens, with no accel) period?

What accelpower are you using?

do not pay any attention to that link, it's not a correct visualisation of the current mouse acceleration.

with a zero offset, the acceleration kicks in immediately.

furry muff TY.

i'm dieing for a graph with sliders that would show how all this works tbh i don't understand everything right there :l

If someone could pull that code off ambro and modify it for the modern accel cvars it could be uploaded to phoon.us/mouse probably and would be a nice addition.

I'm not familiar with java script but I will take a look if I have time.

So my last post was just one case.

Here's my idea for all cases, assuming mouseAccelPower is 2:

Previous Base + (Accel * Offset) =
New Base with offset 0

What was the default mouseacceloffset in q3?

0

Should ql sens feel like cpma sens, I don't think it does. Do you know of any way to get that feel. I can aim in cmpa but am really poor in ql.

See my wild post here for some ideas:

http://www.quakelive.com/forum/showthread.php...post161401

m_mouseaccel 0
nothing else required !

wish there was a way to only allow cl_mouseaccel to M_pitch.

have you tried m_yaw 0?

have you tried militia?

Yeah, back in 1998...