Mousesettings²
•
8 May 2007, 19:43
•
Journals
I have always been interested in the technical aspects of ET. This article is ment as an expansion to 'Mousesettings for newbies". Much like Mousesettings for newbies I do not claim the stuff I talk about will work for everybody. In theory the stuff I write about is correct, in practice most ppl prefer to play with what they are used to play.
First off, I believe in consistency. ET as a game that has a certain 'lotto' factor built into it thanks to spread, antilag and the teambase factor =] You can however reduce this 'lotto' factor by playing
a) intelligently, taking calculated risks which give you a good chance of succeeding automaticly giving you enough of advantage to overcome but the biggest lotto. Playing intelligently requires both the attitide to analyze oneself and to practice practice practice!
b) aim consistent xD
The first thing is nothing I can do about. Some people were just born as plug and play, some ppl have a tendancy to try to understand. The second thing you can improve on. First you want your aim to be as consistent as possible. This is why so much people play relative low sensitivity in ET. You might be able to aim perfectly with sensitivity 5 cm = 180° but you need a way higher concentration to reach the same kind of tracking. Relative low sensitivity makes your aim more consistent, i'm saying relative as some players take this lowering of sensitivity to a new level which seriously hurts their playstyle of forces them to use fov/senstoggles.
Next you need a stable FPS, stable mouserate and a displayrefresh at least as high as your framerate. Ofcourse this is all theory. Some pc's just cant get stable fps, some monitors only give 75hz etc. But lets have a look at the theory. How can we make our aiming more consistent / predictable / stable.
As explained in "Mousesettings by newbies" removing acceleration is a good way to go. Either use accelfix or use your mousedrivers to remove the acceleration. Ofcourse if you have been using windows inbuilt accel for ages, I dont recommend turning it off. CPC showed that there are only two characteristics shared by all top players: Big ass mousemats and relative low sensitivity compared to other games like CoD or Quake. A lot of the players were using the windows inbuilt accel or just didnt give a shit and played with what they played for ages.
So how can you make your aim more consistent? Synching! So what is synching. Lets say i'm playing the piano and playing 2 notes a second, while Loekino is playing the drums, also at 2 beats a second. Both of us will be playing in synch, as we will either play a beat at the same time, or the interval between beats will always be constants. Say if loekino plays his first beat 0,2 seconds earlier, the next beat he plays will also be 0,2 earlier. Also Loekino could be playing 4 beats a second, while i'm only playing 2 beats a second. Common sense tells you he is still playing in synch with me. All sounds nice.
So lets apply synching to ET. What would we want to synch? FPS and Mouserate. Every time a frame gets rendered Q3 looks at the mouseinputs previously received. Say we got 1 frame per second and 1 mouseinput a second, the interval between frame and input will be [1-1000]ms. If we got 125 frames per second and 125 mouseinputs the interval will be in [0-8]ms. Notice that 125 fps = 125/1000ms = 8ms
I say [1-8]ms as there is no way to predict if frames or mouseinput will be 'perfectly' synched: the ms before the frame got rendered a mouseinput was received so the delay between mouseinput and frame is 1 ms. Or the 'worstest' synched: the ms after a frame got rendered and mouseinput was processed, a new mouseinput was received. The maximum delay between mouseinput and frame will be 8ms.
So how could we do better 'in theory'. We cannot increase the FPS due to config restrictions preventing spreadlaming. So the only thing we can do is increase the mouserate. As shown with loekino's drumming, we want to double the mouserate to keep the beats in tune. So if we have 125fps / 250hz the delay interval will be [1-4]ms.
The 'bestest' synch: 1ms, a frame gets rendered the ms after an input was received.
The 'worstest' synch: 4ms, just before the frame got rendered a mouseinput was received. This time the interval is 4 because mouseinputs are beeing sent at double the speed.
Continuing this line of thinking we should decide 1000hz is the 'bestestest' synched as it in theory should give you a 1ms interval as mouseinputs are beeing sent 1000 times a second and there will always be a mouseinput the ms before the frame. In rL mouse dont give stable 1000hz polling rates. You can check it out for yourself: even while doing the most fast circles on your mousemat its impossble to keep a stable 1000hz. This results in loekino skipping some beats and in the end the delay that was constant turns into one big chaos.
fps= 125 = 8ms usb= average 666
frames 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0
inputs. 1 1 0 1 0 1 1 0 0 1 1 1 0 0 1 0 1 1 1 0 0 1 1 0 1 0 1 1 0 1 0 1 1 1 1 0 1 1 1 0 1 0 1 0 1 1 1 1
So lets take the highest possible mouserate thats stable at all times?
fps= 125 = 8ms usb= 500 stable
frames 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0
inputs. 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0
OR
frames 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0
inputs. 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1
As you can see no matter what the the mouseinputs will be stable, the delay will be a constant 1 or 2 ms. But what happens if you got a shit pc that doesnt do 125fps?
fps= 100 =10ms usb= 500 stable
frames 1 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0
inputs. 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0
As you can see there is no problem when using lower fps as long as the 500/ fps = integer 1,2,3, etc. The result is you will have a fixed amount of beats every frame, a fixed amount of mouseinputs per frame. This results in a more consistent tracking.
500fps
250fps
166fps
125
100
83
71
63
Long story short:
use 500hz
cap fps: 125, 100, 83, 71, 63
Stuff missing:
prerendered frame limits ;)
direct input, buffering
To be continued:
Antilag and me, I have known :'(
First off, I believe in consistency. ET as a game that has a certain 'lotto' factor built into it thanks to spread, antilag and the teambase factor =] You can however reduce this 'lotto' factor by playing
a) intelligently, taking calculated risks which give you a good chance of succeeding automaticly giving you enough of advantage to overcome but the biggest lotto. Playing intelligently requires both the attitide to analyze oneself and to practice practice practice!
b) aim consistent xD
The first thing is nothing I can do about. Some people were just born as plug and play, some ppl have a tendancy to try to understand. The second thing you can improve on. First you want your aim to be as consistent as possible. This is why so much people play relative low sensitivity in ET. You might be able to aim perfectly with sensitivity 5 cm = 180° but you need a way higher concentration to reach the same kind of tracking. Relative low sensitivity makes your aim more consistent, i'm saying relative as some players take this lowering of sensitivity to a new level which seriously hurts their playstyle of forces them to use fov/senstoggles.
Next you need a stable FPS, stable mouserate and a displayrefresh at least as high as your framerate. Ofcourse this is all theory. Some pc's just cant get stable fps, some monitors only give 75hz etc. But lets have a look at the theory. How can we make our aiming more consistent / predictable / stable.
As explained in "Mousesettings by newbies" removing acceleration is a good way to go. Either use accelfix or use your mousedrivers to remove the acceleration. Ofcourse if you have been using windows inbuilt accel for ages, I dont recommend turning it off. CPC showed that there are only two characteristics shared by all top players: Big ass mousemats and relative low sensitivity compared to other games like CoD or Quake. A lot of the players were using the windows inbuilt accel or just didnt give a shit and played with what they played for ages.
So how can you make your aim more consistent? Synching! So what is synching. Lets say i'm playing the piano and playing 2 notes a second, while Loekino is playing the drums, also at 2 beats a second. Both of us will be playing in synch, as we will either play a beat at the same time, or the interval between beats will always be constants. Say if loekino plays his first beat 0,2 seconds earlier, the next beat he plays will also be 0,2 earlier. Also Loekino could be playing 4 beats a second, while i'm only playing 2 beats a second. Common sense tells you he is still playing in synch with me. All sounds nice.
So lets apply synching to ET. What would we want to synch? FPS and Mouserate. Every time a frame gets rendered Q3 looks at the mouseinputs previously received. Say we got 1 frame per second and 1 mouseinput a second, the interval between frame and input will be [1-1000]ms. If we got 125 frames per second and 125 mouseinputs the interval will be in [0-8]ms. Notice that 125 fps = 125/1000ms = 8ms
I say [1-8]ms as there is no way to predict if frames or mouseinput will be 'perfectly' synched: the ms before the frame got rendered a mouseinput was received so the delay between mouseinput and frame is 1 ms. Or the 'worstest' synched: the ms after a frame got rendered and mouseinput was processed, a new mouseinput was received. The maximum delay between mouseinput and frame will be 8ms.
So how could we do better 'in theory'. We cannot increase the FPS due to config restrictions preventing spreadlaming. So the only thing we can do is increase the mouserate. As shown with loekino's drumming, we want to double the mouserate to keep the beats in tune. So if we have 125fps / 250hz the delay interval will be [1-4]ms.
The 'bestest' synch: 1ms, a frame gets rendered the ms after an input was received.
The 'worstest' synch: 4ms, just before the frame got rendered a mouseinput was received. This time the interval is 4 because mouseinputs are beeing sent at double the speed.
Continuing this line of thinking we should decide 1000hz is the 'bestestest' synched as it in theory should give you a 1ms interval as mouseinputs are beeing sent 1000 times a second and there will always be a mouseinput the ms before the frame. In rL mouse dont give stable 1000hz polling rates. You can check it out for yourself: even while doing the most fast circles on your mousemat its impossble to keep a stable 1000hz. This results in loekino skipping some beats and in the end the delay that was constant turns into one big chaos.
fps= 125 = 8ms usb= average 666
frames 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0
inputs. 1 1 0 1 0 1 1 0 0 1 1 1 0 0 1 0 1 1 1 0 0 1 1 0 1 0 1 1 0 1 0 1 1 1 1 0 1 1 1 0 1 0 1 0 1 1 1 1
So lets take the highest possible mouserate thats stable at all times?
fps= 125 = 8ms usb= 500 stable
frames 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0
inputs. 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0
OR
frames 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0
inputs. 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1
As you can see no matter what the the mouseinputs will be stable, the delay will be a constant 1 or 2 ms. But what happens if you got a shit pc that doesnt do 125fps?
fps= 100 =10ms usb= 500 stable
frames 1 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0
inputs. 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0
As you can see there is no problem when using lower fps as long as the 500/ fps = integer 1,2,3, etc. The result is you will have a fixed amount of beats every frame, a fixed amount of mouseinputs per frame. This results in a more consistent tracking.
500fps
250fps
166fps
125
100
83
71
63
Long story short:
use 500hz
cap fps: 125, 100, 83, 71, 63
Stuff missing:
prerendered frame limits ;)
direct input, buffering
To be continued:
Antilag and me, I have known :'(
This is easier
Edit : 2nd :<
I DONT BELIEVE IT
edit: lowsenser :[
prerenderlimit.... petition against opengl to have a prerenderlimit less than 2 frames!
direct input.... would be nice if prerenderlimit = 0.
My monitor can 'only' do 100 hertz at a playable resolution(r_mode 6, everything below that looks too bad for me). So which mouseherts would be the best then?
first choose an fps you can keep stable, try it out on a pub running goldrush and see if its stable.
mouserate at 500
FPS = Maxpackets
If FPS is higher then 100, maxpackets will be halved, ET doesnt support maxpackets higher then 100.
So at 125 fps you better cap at 63, so if you do get fps lag and go under 100 it will stay at 63 rather then jumping to 100 during the fps lag. When using 100 or lower fps, just keep maxpackets at 100 as it wont affect anything and it will prevent ppl whining.
Display = as high as you can get it
How do these things transfer to other games such as quake 4? (or any other game that doesn't use the Q3 engine.)
I can set it up to 160
I have tried 100 fps but it's just not smooth enough :)
and don't u get tearing with that?
wouldn't 100fps with 120Hz be better?
and if fps is stable, could it be rather higher than 125?
125fps will be the max you will be able to cap it at, just like in rtcw. In a slow paced game like ET there's really no need of 1ms flickshots etc, so why allow more if ppl could abuse it?
i'll stick with
1000hz @ 72fps @ 75hz thxs vmuch
well done?
use maxfps 85 packets 85 and hz 85 if 85 is your max hz
even if my mouse cant do it really stable?
dont understand anything about refreshrate, mhz, etc...
That is what i use :p dunno if it's perfect
***Edit***
If I try to lock my FPS at 63 it will raise to a 66.7 stable FPS
So it will be 66.7FPS >500hz > 1024x768 > 100maxpackets > rate is 25000
because 63 aint possibole :<
edit: why would you want to cap your fps that low?
>:D
The magical 500hz/125fps/120hz is something to aim for rather then fucking up all your settings just to get something 'synched'.
If you buy a new pc and get 125 fps stable, this is what I recommend to any player. Got a tft, np still keep playing 125fps. Cant tweak usbport at lan, 125hz will still do fine. And if you are playing 76fps 75hz 500hz you are using what a lot of the pros are using, its not perfect but reality never is ;)
and thats the best what i've get
125/125/120/1000 in Quake, theres really not much tracking in Quake and the uberhigh polling rate is nice for fast flicks and rails.
G5 @ 1000Hz / 400dpi
ET @ 125fps (unstable on radar only) r_mode 8
LCD @ 75Hz
cl_maxpackets 100
i've big problems with shooting on long range (fov 100) maybe it sounds stupid but when opponent hit me my crosshair fly away (around 3cm ingame) and its hard to point it back before i get killed
the crosshair shaking is just an animation, you should just continue to aim as if your crosshair were still in the same position
now got new Reflex record also :D