Page 8 of 44
Re: flood's input lag measurements
Posted: 21 Nov 2014, 19:16
by Dragonheart
Well, you can always stop

Re: flood's input lag measurements
Posted: 22 Nov 2014, 19:19
by flood
new idea: use arduino as a virtual mouse that also activates an led at the same time it sends the usb message
Re: flood's input lag measurements
Posted: 24 Nov 2014, 16:04
by silikone
This method of measuring lag could be used to compare mouse sensors. In this case, the G100s is known to be responsive, but other mice can perhaps have a bunch of well hidden milliseconds added. Even the click tests that don't necessarily measure the sensor can reveal a surprising amount of input lag. There was that test on Esreality that compared the G100s with a Zowie mouse, and the difference was significant.
flood wrote:new idea: use arduino as a virtual mouse that also activates an led at the same time it sends the usb message
Is USB even possible on a vanilla Arduino? If not, I know there is a PS/2 library.
Re: flood's input lag measurements
Posted: 24 Nov 2014, 17:07
by flood
silikone wrote:This method of measuring lag could be used to compare mouse sensors. In this case, the G100s is known to be responsive, but other mice can perhaps have a bunch of well hidden milliseconds added. Even the click tests that don't necessarily measure the sensor can reveal a surprising amount of input lag. There was that test on Esreality that compared the G100s with a Zowie mouse, and the difference was significant.
yea but the issue is that the sensors must not malfunction under the high accelerations of this test (0 to 1m/s in less than 1ms means >100g)
the only mice i have are a g100s, a old logitech g3, a pre ordered ninox aurora which i'll be receiving in approximately 7 years.
flood wrote:new idea: use arduino as a virtual mouse that also activates an led at the same time it sends the usb message
Is USB even possible on a vanilla Arduino? If not, I know there is a PS/2 library.
YUP
http://arduino.cc/en/Tutorial/ButtonMouseControl
already ordered one... it looks ridiculous simple.
Re: flood's input lag measurements
Posted: 25 Nov 2014, 04:05
by flood
cs1.6: with fps_override 1 and fps_max 10000, the game runs at 2000fps.
i measured consistently between 2 and 4ms of lag. this is pretty much the limit...
anyway ill try drivers, gtx460, and my old computer once i get the arduino stuff set up
i'm also planning to use the arduino to check if csgo has any mosue smoothing as some people claim there is with rawinput
Re: flood's input lag measurements
Posted: 25 Nov 2014, 23:23
by flood
Re: flood's input lag measurements
Posted: 26 Nov 2014, 00:04
by flood
measured 1-2ms in cs1.6 with arduino setup
will post details later
Re: flood's input lag measurements
Posted: 26 Nov 2014, 08:23
by flood
so i hooked it up so that when i press a button on my breadboard, a row of leds light up (instantly), and the arduino which is acting as a mouse, moves 100 dots to the right.
i set fps_override 1 in cs1.6 to get unlocked framerate, which led to a fps of ~2000
and... here's the result:
https://www.youtube.com/watch?v=jhK5_Rr1Nyw
Re: flood's input lag measurements
Posted: 27 Nov 2014, 04:13
by LagBuster
Awesome!

Re: flood's input lag measurements
Posted: 03 Dec 2014, 10:54
by treach
Oh man thats really next level stuff, GREAT work!!!
1-2 ms is sick, looks like you have the most accurate setup on the planet at the moment.
Now you can really tell wich variables de- or increase lag