+ Reply to Thread
Page 1 of 3 1 2 3 LastLast
Results 1 to 10 of 21

Thread: yet another Q3=QL sensitivity thread..

  1. #1
    Junior Member davay_harosh is on a distinguished road
    Join Date
    Nov 2012
    Posts
    18

    yet another Q3=QL sensitivity thread..

    As title says, I would like to know how to convert exactly my Q3 settings to QL.Ive went through Yakumos guide, which brought me close, but not quite there yet and as Im not using m_cpi thing I assume SyncErrors post about this is irrelevant.

    These are my q3 sensitivity settings:

    sensitivity 1.6
    cl_mouseaccel 0.3
    m_pitch 0.017
    m_yaw 0.014

    And this is from yakumos guide
    [q]q3accel = cl_mouseaccel setting when using cl_mouseAccelStyle 0 (or accel in quake3 arena)

    When cl_mouseAccelStyle 1
    Set cl_mouseAccel 2
    cl_mouseAccelOffset = (1 / q3accel ) * .8[/q]

    I assume that offset should be multiplied with 80 and not 0.8??

    would my accel then look like this:

    cl_mouseaccel 2 or 0.3???
    cl_mouseacceloffset 266.6666666666667

    and rest of the settings stay the same..

    The thing that is so confusing to me that it does feel close sensitivity while playing, but base sensitivity from ql is slower than in q3.I need to make longer distance for 360 than in q3.And if I increase the sensitivity approx. to the distance level from q3 the sensitivity feels too fast!?Can someone elaborate on this.. Lorfa and such??Thank you!

  2. #2
    Senior Member Lorfa is a jewel in the rough Lorfa is a jewel in the rough Lorfa is a jewel in the rough Lorfa's Avatar
    Join Date
    Aug 2010
    Location
    Kepler-22b
    Posts
    8,902
    Quote Originally Posted by davay_harosh View Post

    These are my q3 sensitivity settings:

    sensitivity 1.6
    cl_mouseaccel 0.3
    m_pitch 0.017
    m_yaw 0.014

    And this is from yakumos guide
    [q]q3accel = cl_mouseaccel setting when using cl_mouseAccelStyle 0 (or accel in quake3 arena)

    When cl_mouseAccelStyle 1
    Set cl_mouseAccel 2
    cl_mouseAccelOffset = (1 / q3accel ) * .8[/q]

    I assume that offset should be multiplied with 80 and not 0.8??

    would my accel then look like this:

    cl_mouseaccel 2 or 0.3???
    cl_mouseacceloffset 266.6666666666667

    and rest of the settings stay the same..

    The thing that is so confusing to me that it does feel close sensitivity while playing, but base sensitivity from ql is slower than in q3.I need to make longer distance for 360 than in q3.And if I increase the sensitivity approx. to the distance level from q3 the sensitivity feels too fast!?
    cl_MouseAccelStyle doesn't exist in QL anymore (and hasn't for some time).

    It doesn't exist in vanilla Q3 either (point releases). It does exist in ioquake3 however. Can you confirm what client you were using to play Q3?

    The accel code that QL uses is most similar to style 0.

    One thing that is different about q3 is the mouse input code. I believe that by default Q3 uses direct input.

    QL uses raw input (WM_Input) by default in the windows version.

    You can make use of direct input with:

    in_mouse "1"
    in_restart

    Then these numbers should be identical:

    sensitivity 1.6
    cl_mouseaccel 0.3
    m_pitch 0.017
    m_yaw 0.014

    To keep things just like Q3, also make sure the following are set (also these are their defaults):

    cl_mouseAccelOffset "0"
    cl_mouseAccelPower "2"

  3. #3
    *fonsxe
    Guest
    Article about WM_MOUSEMOVE, WM_INPUT & DirectInput ;-)
    http://msdn.microsoft.com/en-us/libr...=vs.85%29.aspx

  4. #4
    Junior Member davay_harosh is on a distinguished road
    Join Date
    Nov 2012
    Posts
    18
    Quote Originally Posted by Lorfa View Post
    cl_MouseAccelStyle doesn't exist in QL anymore (and hasn't for some time).

    It doesn't exist in vanilla Q3 either (point releases). It does exist in ioquake3 however. Can you confirm what client you were using to play Q3?

    The accel code that QL uses is most similar to style 0.

    One thing that is different about q3 is the mouse input code. I believe that by default Q3 uses direct input.

    QL uses raw input (WM_Input) by default in the windows version.

    You can make use of direct input with:

    in_mouse "1"
    in_restart

    Then these numbers should be identical:

    sensitivity 1.6
    cl_mouseaccel 0.3
    m_pitch 0.017
    m_yaw 0.014

    To keep things just like Q3, also make sure the following are set (also these are their defaults):

    cl_mouseAccelOffset "0"
    cl_mouseAccelPower "2"
    Sorry, but this is ridiculous.If I keep the values identical as you are suggesting with pwr2/offset0 the accel becomes uncontrollable, without applying any senscap, do you honestly believe someone can play with such fast sensitivity?I know that Yakumos guide refers to accelStyle1, Im not illiterate, but as said I have nothing else out there to refer to.. SyncErrors post is not relevant either thats why Im asking this.The thing is that with your suggestion, accel feels closer to what m_accel is in q3 and not cl_mouseaccel.Please test the client I provided below with m_accel/cl_mouseaccel 0.3, see the diference.. then go to ql and test the same.

    I was using this client,I assume its accepted as standard.. at least in esr:
    http://www.esreality.com/post/156593...input-support/

    Im not certain if accels are identical or not going yakumos road, but im satisfied with it as I dont need to use senscap which kills natural q3 accel behavior.The problem is which im trying to point-out, that BASE SENSITIVITIES are different!!As banal and inaccurate as it sounds, I actually tried that 360cm formula with wolfram and was surprised at the accuracy, but in a negative way.Heres why, the result really is accurate, but it applies only to Q3 and not QL!!!With values I provided result is around 25cm(accounting 1600dpi).. when testing in ql i run out of my mousepad, while in q3 it really is around 25/26 flux.So you probably didnt understand me in the first post, my question is.. If both clients are derived from the same RawInput, why are the base sensitivities without any acceleration applied to them so much different!?!?!?!?!?This is the most confusing thing to me and Id really like someone to explain this problem.Thank you for the answer!

  5. #5
    Senior Member Lorfa is a jewel in the rough Lorfa is a jewel in the rough Lorfa is a jewel in the rough Lorfa's Avatar
    Join Date
    Aug 2010
    Location
    Kepler-22b
    Posts
    8,902
    Quote Originally Posted by davay_harosh View Post
    Sorry, but this is ridiculous.If I keep the values identical as you are suggesting with pwr2/offset0 the accel becomes uncontrollable, without applying any senscap, do you honestly believe someone can play with such fast sensitivity?
    I don't know what dpi your mouse has, at 400 or even 450 dpi these numbers are certainly playable.

    I know that Yakumos guide refers to accelStyle1, Im not illiterate, but as said I have nothing else out there to refer to..
    You could also see the guide in my sig.

    I assure you that accelstyle does not exist in Q3A or Quake Live. It was removed from Quake Live ages ago.

    SyncErrors post is not relevant either thats why Im asking this.The thing is that with your suggestion, accel feels closer to what m_accel is in q3 and not cl_mouseaccel.
    m_accel doesn't exist in that Q3A client, nor does it exist in Quake Live.

    Any cvar will appear to exist if created with set or seta, however it will not be effectual.

    Please test the client I provided below with m_accel/cl_mouseaccel 0.3, see the diference.. then go to ql and test the same.
    I just tested that Q3A raw client vs. Quake Live. They are identical in terms of sensitivity and mouseaccel. There is no senscap in that Q3A which is basically point release 1.32 with the raw input added (in_mouse 3).

    I was using this client,I assume its accepted as standard.. at least in esr:
    http://www.esreality.com/post/156593...input-support/
    That is in_mouse "2" in QL.

    The difference between dinput and raw isn't that large imo, it's more of a change in feel.

    Im not certain if accels are identical or not going yakumos road, but im satisfied with it as I dont need to use senscap which kills natural q3 accel behavior.
    I suppose. I enjoy its added benefit personally.

    The problem is which im trying to point-out, that BASE SENSITIVITIES are different!!As banal and inaccurate as it sounds, I actually tried that 360cm formula with wolfram and was surprised at the accuracy, but in a negative way.Heres why, the result really is accurate, but it applies only to Q3 and not QL!!!With values I provided result is around 25cm(accounting 1600dpi).. when testing in ql i run out of my mousepad, while in q3 it really is around 25/26 flux.So you probably didnt understand me in the first post, my question is.. If both clients are derived from the same RawInput, why are the base sensitivities without any acceleration applied to them so much different!?!?!?!?!?This is the most confusing thing to me and Id really like someone to explain this problem.Thank you for the answer!
    Ahh 1600 dpi, in that case sens 2 with accel 0.3 should be extremely high.

    Anyways, I strongly suspect that something odd is going on with your QL config. The most likely culprit is senscap (although that's not necessarily it).

    If you could pastebin your Q3 config and your QL config, I could compare them and make sure that there aren't any mitigating factors.

  6. #6
    Senior Member Lam has a spectacular aura about Lam has a spectacular aura about Lam has a spectacular aura about Lam's Avatar
    Join Date
    Aug 2010
    Posts
    3,763
    Quote Originally Posted by Lorfa View Post
    The difference between dinput and raw isn't that large imo, it's more of a change in feel.
    For QL's needs, on current Windows+DX versions, DirectInput is just a wrapper over the WM_INPUT messages. The difference in feel may come from the fact that DirectInput runs in separate thread, so for some people (single-core CPUs?) it can have slightly delayed "feel", for others it even works better (no idea why, though). You can switch between in_mouse 1 and 2 and get same sensitivity (after all it's using same API underneath), and in most configurations, similar if not the same feel.

  7. #7
    Senior Member JustLOL is on a distinguished road JustLOL's Avatar
    Join Date
    Feb 2012
    Location
    Croatia
    Posts
    288
    Can it have anything to do with FOV? I remember reading somewhere that if you're using 16:9 res in QL, and you'd want to keep the same feel in Q3, you have to add 15 to your FOV (103 at 1920x1080 in QL = 118 at 1920x1080 in Q3). Don't know if it only affects widescreen, or there are calculations that might also apply for 4:3.

  8. #8
    Junior Member davay_harosh is on a distinguished road
    Join Date
    Nov 2012
    Posts
    18
    I assure you that accelstyle does not exist in Q3A or Quake Live. It was removed from Quake Live ages ago.
    Please dont patronize me, Ive been playing the game since my 486DX.. I think I know enough about each from the series, I know what Im talking about.

    m_accel doesn't exist in that Q3A client, nor does it exist in Quake Live.
    This is shocking to me, why do you keep convincing me in such manner??I will leave this scrnPrint as an answer as it describes the problem better than my words:

    http://s8.postimg.org/c417qhjo5/m_accel.jpg <-----Note the "RawQuake" from the right side!!

    I didnt fall from a tree, Im not stupid, I know what Im talking about, if you know something more than me please provide a solution dont persist on convincing me.

    I never said there is m_accel in QL, what I said was.. That with your suggestion of values to be applied in QL, the accel feels much closer to what m_accel is in Q3 than to cl_mouseaccel.While going yakumos path the accel feels closer to what cl_mouseaccel is in that Q3 client.But thats not the main issue more on that later..


    I just tested that Q3A raw client vs. Quake Live. They are identical in terms of sensitivity and mouseaccel.
    At this point I just cant believe you, because you didnt even bother checking out the m_accel thing.There are hundreds of others that played with me years ago and can clearly tell there is a BIG DIFFERENCE!!

    Anyways, I strongly suspect that something odd is going on with your QL config. The most likely culprit is senscap (although that's not necessarily it).
    I dont know, as said I keep my senscap at 0.

    If you could pastebin your Q3 config and your QL config, I could compare them and make sure that there aren't any mitigating factors.
    Could you leave me your email address on priv message, I never used this pastebin thing?

    Now I want to clear this other thing thats bothering me in the first place.Following this formula below, knowing how superficial it is measured, I am surprised of its accuracy, but it only applies to Q3 and NOT QL!!

    The formula to compute sensitivity in inch/cm for a 360

    Assuming no acceleration is used:

    Sinch=360 / (M*R*B)
    Scm=360 / (M*(R/2.54)*B)

    with

    Sinch=sensitivity in inch
    Scm=sensitivity in cm
    M=m_yaw
    R=mouse resolution (dots/counts per inch)
    B=ingame sensitivity
    Each time I drew my mouse I was able to pinpoint the result with fluctation around +-1cm accurate!!!BUT ONLY IN Q3!!!!!
    When I attempted the same in QL the error was of such scale that I actually ran out of my mousepad and didnt even bother measuring it as it was evident something is utterly wrong with the sensitivity(no no m_cpi is 0), but lets say the error is more than 10cm(5inch).Wouldnt you agree that is not just a "slight" difference, but serious glitch??That was compared with all accels off.

    There are only 2 logical conclusions I can draw from this..its either the formula that is incorrect and only applies to q3(despite being writen for ql use), or the sensitivity in ql is just different.There is no 3rd option.However I find it surprising both of these are supposed to be derivated from RAWINPUT.Its either Formula or QL..

    Thank you for respond!
    Last edited by davay_harosh; 06-08-2013 at 03:31 AM.

  9. #9
    Senior Member Lorfa is a jewel in the rough Lorfa is a jewel in the rough Lorfa is a jewel in the rough Lorfa's Avatar
    Join Date
    Aug 2010
    Location
    Kepler-22b
    Posts
    8,902
    Quote Originally Posted by davay_harosh View Post
    Please dont patronize me, Ive been playing the game since my 486DX.. I think I know enough about each from the series, I know what Im talking about.
    What does this have to do with anything?

    cl_MouseAccelStyle does not exist in Q3A, Quake Live, or CNQ3. That's the truth.

    This is shocking to me, why do you keep convincing me in such manner??I will leave this scrnPrint as an answer as it describes the problem better than my words:

    http://s8.postimg.org/c417qhjo5/m_accel.jpg <-----Note the "RawQuake" from the right side!!

    I didnt fall from a tree, Im not stupid, I know what Im talking about, if you know something more than me please provide a solution dont persist on convincing me.
    I said Q3A, not CNQ3 (CPM or CPMA). That is different.

    This screenshot isn't really telling though because like I said you can create cvars with seta. You'd have to start with no configs to see if a cvar exists or not.

    m_accel does exist in CNQ3 though.

    I never said there is m_accel in QL, what I said was.. That with your suggestion of values to be applied in QL, the accel feels much closer to what m_accel is in Q3 than to cl_mouseaccel.While going yakumos path the accel feels closer to what cl_mouseaccel is in that Q3 client.But thats not the main issue more on that later..
    At this point I just cant believe you, because you didnt even bother checking out the m_accel thing.There are hundreds of others that played with me years ago and can clearly tell there is a BIG DIFFERENCE!!
    I DID test it with Q3A vs. QL.

    You're talking about CNQ3 instead.

    Just now I also tested CNQ3.

    This is what I found:

    cl_mouseaccel has no effect in CNQ3 (per feel and cl_showMouseRate)

    m_accel does (per feel and cl_showMouseRate).

    The base sensitivity was also identical.

    I measured in CNQ3:

    m_yaw "0.022"
    sensitivity "3.5"
    m_accel "0"
    in_mouse "3"

    vs. in QL:

    m_yaw "0.022"
    sensitivity "3.5"
    cl_mouseaccel "0"
    in_mouse "2"

    Both were ~30 cm/360 within a very small margin of error.

    This is what it should be, per funender, and simply 400 cpi at 0.077 (0.022 * 3.5) degrees/count 360/0.077 = 4675.32467532 counts, divided by 157.480314961 counts/cm which is 400 cpi. That comes to 29.6883116882 cm/360.

    I am sure that they are identical with these settings. I tried faster motions to try and detect any built-in accel and there wasn't any.

    Are you saying that the base sens is off _and_ the accel? I tested CNQ3 with my accel values and they felt the same as Q3A and QL too. There could be an issue there though, it's more difficult to test but I believe we have the source code so we could at least look up the difference between Q3A and CNQ3, and like I said Q3A and QL were the same.

    Could you leave me your email address on priv message, I never used this pastebin thing?
    It's at the bottom of the guide in my sig (there is no private messaging on this forum).

    By pastebin I mean in QL and the other client you are using, apparently CNQ3, do \writeconfig somename

    Then open up somename.cfg and paste the contents into a pastebin.com.

    Should produce two separate links so that I can compare them. I suppose you could e-mail them as well.

    Now I want to clear this other thing thats bothering me in the first place.Following this formula below, knowing how superficial it is measured, I am surprised of its accuracy, but it only applies to Q3 and NOT QL!!
    I think when you say Q3, you don't mean Q3A but CNQ3 and/or CPM.

    Each time I drew my mouse I was able to pinpoint the result with fluctation around +-1cm accurate!!!BUT ONLY IN Q3!!!!!
    When I attempted the same in QL the error was of such scale that I actually ran out of my mousepad and didnt even bother measuring it as it was evident something is utterly wrong with the sensitivity(no no m_cpi is 0), but lets say the error is more than 10cm(5inch).Wouldnt you agree that is not just a "slight" difference, but serious glitch??That was compared with all accels off.

    There are only 2 logical conclusions I can draw from this..its either the formula that is incorrect and only applies to q3(despite being writen for ql use), or the sensitivity in ql is just different.There is no 3rd option.However I find it surprising both of these are supposed to be derivated from RAWINPUT.Its either Formula or QL..
    It's likely some config setting causing the problem. That is why I asked to see them so that I could look for the culprit.

  10. #10
    Junior Member davay_harosh is on a distinguished road
    Join Date
    Nov 2012
    Posts
    18
    cl_mouseaccel has no effect in CNQ3 (per feel and cl_showMouseRate)

    m_accel does (per feel and cl_showMouseRate).
    I agree on this one.. just discovered.

    The base sensitivity was also identical.
    I disagree on this one.. when drawing the mouse with slow speed it seems similar, but when twitching or trying to do rail flicks with fast movement, this is where it becomes evident that these are 2 different sensitivities.

    Are you saying that the base sens is off _and_ the accel? I tested CNQ3 with my accel values and they felt the same as Q3A and QL too. There could be an issue there though, it's more difficult to test but I believe we have the source code so we could at least look up the difference between Q3A and CNQ3, and like I said Q3A and QL were the same.
    Something is different in BASE SENSITIVITY and im 100% certain about that!!!!I dont know if its mouse specific going from low dpi to high dpi or if drivers related I will take time in the next following days to test this through with WMO and abyssus aswell.I assume that ure using WMO(400dpi) so will have to test this through carefully cant say anything now.But regarding the code comparison, the thing is that there is no need in comparing the Q3A to CNQ3 code.. because as I said in above post, its the CNQ3 client that is accurate to +-1cm and not QL!!I massively overdraw the length distance for 360 in QL it doesent compute in practice with the formula.

    It's likely some config setting causing the problem. That is why I asked to see them so that I could look for the culprit.
    I doubt its config related, because I tried deleting my config and "reset to defaults" in my Game settings.As far as I understood that erases everything I had in qzconfig/repconfig and everything else.Then I simply binded my setup, but sensitivity was still same flawed QL sensitivity, I couldnt replicate CNQ3 sensitivity.The problem at which I sumble is.. that when my mouse hits a certain speed it just stops accelerating sensitivity, like it hits the wall and stops tracking further like its supposed to, I dont turn to such degree as I do in CNQ3 while at small movements it does seem pretty close to Q3 feel.Its difficult to describe exactly, I will try to get a proper camera and make a clip of this phenomenon, but I dont promise anything because I dont own any, will have to try and borrow one.

+ Reply to Thread

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts