Mouse sensitivity issue in Source based games with wide screen setting

Some years ago, I liked playing Counterstrike (at last 1.6) professionally. The people who know what I’m talking about, also know that mouse/sensitivity/resolution configuration is crucial for success. These days, friends and me from time to time play Day of Defeat: Source and Team Fortress 2. In these games, I’ve lost a huge amount of skill im comparison with Counterstrike in earlier days. That’s okay, because I did not play for years and I do not know how to play DoD/TF2 as well as Counterstrike formerly. But unnecessarily one annoying thing results in an even bigger loss of skill: setting the same mouse sensitivity in CS 1.6 (“Halflife 1”) and DoD:S/TF2 (“Halflife 2”) does not result in the same distance-of-mouse-movement to distance-of-crosshair-movement-on-the-screen ratio, when using wide screen aspect ratio. Now, I’ve spent some time to analyze this issue more deeply.

Method

I’ve modified my mouse pad with some steering elements, so that I can make the same mouse movement with only right/left component with always the same distance-of-mouse-movement. It’s quite accurate.

Ingame I stand right in front of a wall and shoot to mark a point on this wall. I then go “some steps” backwards, stand still and aim at this marked point on the wall. This guarantees that I really look straightforward (and not a bit up/down). Starting at this marked point, I perform the special right/left movement of a certain distance. Then I simply measure the distance D between the crosshair position and the marked point directly on the screen using a ruler. The result of this measurement is the quantity D [cm] I base my conclusions on: I take this value under certain different conditions and compare them.

Of course this method brings along a measurement error of about some millimeters, but it’s negligible in comparison with the distances I’m measuring and with the differences of the distances I’m comparing.

Experiment, results and discussion

The following table shows the relevant configuration that is valid throughout the whole test:

General configuration
mouse MX 518
mouse dpi setting lowest
Windows mouse speed 6/11

Okay, now let’s see the results of the tests with Counterstrike 1.6

Counterstrike 1.6
resolution 1600 x 1200 (4/3)
sensitivity 1.0
others launch options: -noforcemspd -noforcemaccel -noforcemparms
D 19 cm

D=19 cm. This was measured at least 20 times on different maps and on different positions. Now I’ve changed the resolution (and the aspect ratio):

resolution 1920 x 1200 (16/10)
sensitivity 1.0
others launch options: -noforcemspd -noforcemaccel -noforcemparms
D 19 cm

19 cm, again. I somehow remembered that changing the aspect ratio in Halflife 1 results in another real sensitivity while keeping the same sensitivity setting. I’ve just measured that I remembered wrong: In Counterstrike 1.6, the real sensitivity is independent of the aspect ratio (while keeping the same sensitivity setting)!

Now let’s see what’s up with Day of Defeat: Source.

Day of Defeat: Source

Regarding Halflife 2 / the Source engine, I only consider DoD:S — I think the results should be the same as for TF2; at least I hope :-)

resolution 1600 x 1200 (4/3)
sensitivity 1.0
others m_pitch “0.022”, m_side “0.8”, m_yaw “0.022”, m_filter “0”, m_forward “1”, m_customaccel “0”, m_customaccel_scale “0”, m_customaccel_max “0”, m_customaccel_exponent “1”, m_mouseaccel1 “0”, m_mouseaccel2 “0”, m_mousespeed “1”,
D 19 cm

Again D = 19 cm. It seems that you can keep the same sensitivity setting in Hl1 and Hl2 without any worry; at least for a 4/3 aspect ratio in Hl2. Let’s change the aspect ratio here, too.

resolution 1920 x 1200 (16/10)
sensitivity 1.0
others m_pitch “0.022”, m_side “0.8”, m_yaw “0.022”, m_filter “0”, m_forward “1”, m_customaccel “0”, m_customaccel_scale “0”, m_customaccel_max “0”, m_customaccel_exponent “1”, m_mouseaccel1 “0”, m_mouseaccel2 “0”, m_mousespeed “1”,
D 16 cm

Err.. okay now we’ve seen an essential difference between Halflife 1 and Halflife 2 based games: In DoD:S the real sensitivity depends on the aspect ratio! And what now? You *want* to have the same real sensitivity as in Hl1/Counterstrike 1.6? I tried to calculate it, assuming linearity:

s1/s2 = a1/a2
s1=1.0, a1=1600/1200, a2=1920/1200
s2=1.2

Let’s see:

resolution 1920 x 1200 (16/10)
sensitivity 1.2
others m_pitch “0.022”, m_side “0.8”, m_yaw “0.022”, m_filter “0”, m_forward “1”, m_customaccel “0”, m_customaccel_scale “0”, m_customaccel_max “0”, m_customaccel_exponent “1”, m_mouseaccel1 “0”, m_mouseaccel2 “0”, m_mousespeed “1”,
D 21 cm

D=19 cm would have been desired, but I got 21 cm. When we assume linearity and calculate 16 cm*1.2=19.2 cm and compare that to the result D=21 cm, we see that the assumption of linearity is wrong.

There is a nonlinearity between the sensitivity setting and the distance-of-crosshair-movement-on-the-screen. I did not expect this nonlinearity and I don’t think that it makes any sense. It means that when you change your sensitivity setting from 1 to 2, you won’t get double speed! In this case, you will get more!

The dumb thing now is that we don’t know and I don’t want to do many measurements to find out the almost-exact form of this nonlinearity. Hence, because of this nonlinearity, we cannot calculate the sensitivity setting for DoD:S in widescreen mode to get exactly the same real sensitivty as in 4/3 mode of DoD:S or as in widescreen/normal mode of Counterstrike 1.6. Since it is a matter of fact that we cannot set it more exactly than to one digit (DoD rounds it then), an exact calculation would even not help much in many cases. With this widescreen aspect ratio, you’re simply not able to get the same real sensitivity as in CS 1.6 or as in DoD:S 4/3 screen mode :-(

Conclusion

Counterstrike 1.6, which is based on Halflife 1, produces the same real sensitivity for normal or widescreen aspect ratios, while keeping the same sensitivity setting. Day of Defeat: Source, which is based on Halflife 2, produces the same real sensitivity as Counterstrike 1.6, while keeping the same sensitivity setting and playing in 4/3 aspect ratio mode.

If you play Day of Defeat: Source in widescreen mode, you have to adjust your sensitivity setting to get the same real sensitivity as in 4/3 mode of DoD:S or as on widescreen/normal mode of Counterstrike 1.6. But:

  • it cannot be calculated easily, because of an unknown nonlinear relation
  • if you find it out by “trying”, you often cannot set it exactly, because only one digit is allowed

I want to have the same real sensitivity on both, Counterstrike 1.6 and DoD:S. Hence, I have to play DoD:S in a 4/3 mode on a wide screen. Urks.

If you have more information about this problem, then please let me know. Any feedback is appreciated.

  • me

    If movement is not linear it seams that mouse acceleration is on and btw the resolution is 1920×1200. I don’t see any real problem with this learn to master widescreen or play in 4:3 aspect with black bars.

  • Thanks for telling me about the resolution; this was a careless mistake. I’ve corrected the post.

    Your mouse acceleration comment tells me that you’ve not read carefully. I compare two independent mouse movements. In both cases the mouse is moved by exactly the same real right->left trajectory on the mousepad and hence by the same real distance. Both cases are performed multiple times. I am sure that I did not move the mouse with the same way-over-time-curve every time, but I reproduced the measured value every time. This indicates that acceleration had no impact here (in fact, it was switched off). Again, the results that indicate the nonlinearity:

    • sensitivity 1 -> D=16 cm
    • sensitivity 1.2 -> D=21 cm

    Hence, the nonlinear behaviour I mention does not characterize one mouse movement. Only in this case talking of acceleration would have made sense. Again: mouse acceleration is turned off in all cases (look at the settings I used, they’re listed.) and it’s switched off in Windows, too.

    I think a more intelligent comment would have been that I must not compare distances, but instead I should compare angles (but these are more difficult to measure). But hey, this nonlinearity is the less important topic.

    The question still is why Source-based games don’t do the adjustment internally when switching the aspect ratio. While CS 1.6 (or Hl1) does this internally, with DoD:S (or Hl2) this adjustment is left to the user and — as shown — this is not easy to accomplish, or even not possible to accomplish exactly.

  • Christian Gehrcke

    Hi, Jan-Philip!

    Just nice to see that you do have other – quite normal – problems beside your Master’s thesis :-)
    BTW: Good luck at ICMRM in West Yellowstone! Have a nice time there!

    Greetings from Bückeburg,

    Christian

    • Ha! :-) Ich machs mal auf Deutsch.. ja, ich hoffe, dass ich irgendwann mal wieder bisschen zum Spielen komme, DoD und TF2 machen echt Spaß! :D

      Danke für die Grüße! Zurück aus der Pampa.. Amerika ist genau so wie man sich es vorstellt. Man man man…, “die spinnen die ***”

  • Fabian

    ich habe gerade einen ganzen colakracher auf einmal verschluckt und das tut richtig weh. die sind so hart, das denkt man gar nicht.

    auch ich habe mir viele gedanken über oben genanntes problem gemacht und bin zu folgendem schluss gekommen:

  • GARRRRNOVE

    Very nice work you ve done here, this issue in source engine games is really annoying if you are forced to switch between screen ratios.