Input Lag ... It's more than just the monitor...

daveswantek said:
Add to this the fact that Human reaction time is somewhere between 250ms and 400ms.

You can test it here: http://www.getyourwebsitehere.com/jswb/rttest01.html
Click to expand...

The "human reaction time" is irrelevant.
Feeling lag has nothing to do with the reaction time.

Devices that mimic different surfaces (like a 3D mouse with force-feedback) need to update the feedback thousands of times per second (less than one ms) for it to feel realistic - otherwise the lag is too great.
 
I didn't know the default Vista mouse polling rate was a piddly slow 125HZ. Gonna fix that right now, thanks for the link!
 
Thanks for the link.
Also this is teh reason the input-lag of an LCD is measured against a CRT connected to the same computer :)
 
My lowest scores were .195 and .022... and on a laptop.. hrm.. need to find a mouse :)
 
I guess this is not to bad for someone my age. 0.24020000000000002

I have a Logitech G5 lazer mouse and have the polling set to 1000. I also have a Samsung Syncmaster 275T plus paired with an ATI 4870. I feel no lag at all when playing FPS. I play Left4Dead a lot and it is just amazing on this monitor.
 
Last edited:
I averaged 247ms. I am on a 3007wfp-hc and a wireless mouse (mx revolution). Not sure if that makes a difference.

The more I try it the lower I can get. Just averaged 209ms.
 
Yeah, my problem with this one is that there's no penalty for accidentally going early; which caters to results that aren't accurate.

Still, for research purposes: I've got a Logitech MX310, corded, optical, USB mouse, in Windows XP, latest mouse drivers, checked the polling and it's set to the default 125Hz, no over-clocking here.

I'm on a Dell 2408WFP revision A02, running an Intel Core 2 Duo E6600, dual core 2.4GHz, 2GB RAM DDR2-800 CL4, and the video card is an 8800GTS 640MB Superclocked (G90).

When I first tried I was getting ~240ms to ~260ms, when I started really hankering down I had a min of about .201 w/ an average of .227.

I only played two sets in total.

So in theory, if I was on a CRT and overclocked my mouse, if my current average is .227, I might expect to go down a total of ~.041 bringing my avg score to ~0.183.

Maybe if I get really bored I'll do that to see if I get results that match my predictions.

Would like to know if someone out there is using a high Hz CRT to see if the results make sense compared to anandtech's testing.
 
MrGuvernment said:
that test is useless if you just keep clicking, i hit .016 on one....


avg 0.199
Click to expand...

Try this one it doesnt allow clicking before turning green: http://www.humanbenchmark.com/tests/reactiontime/index.php

Although monitor input lag & human reaction time are two different things this test will have better results on say a CRT's 0ms vs. LCD with 50ms input lag simply because the Green will technically appear sooner on the CRT (or an LCD with low input lag etc.)
 
Last edited:
I have the Samsung 305T and it is a damn nice monitor for $999. There is input lag (so I've read) but I am not sensitive to it. I am a gamer and do not find that lag is a hinderance. I don't normally say things like this, but I finish in the top 3 almost every round in TF2. I wouldn't blame lower performace on input lag :).
 
mathesar said:
Try this one it doesnt allow clicking before turning green: http://www.humanbenchmark.com/tests/reactiontime/index.php

Although monitor input lag & human reaction time are two different things this test will have better results on say a CRT's 0ms vs. LCD with 50ms input lag simply because the Green will technically appear sooner on the CRT (or an LCD with low input lag etc.)
Click to expand...


I tried that test. The monitor DEFINITELY makes a difference.
I've got two monitors connected to my video card, a Samsung 215TW and a Viewsonic VP912b.

Average test time for the Samsung - 339.4 MS
Average test time for the Viewsonic - 270.4 ms

That's a difference of 70ms!
 
vjcsmoke said:
I tried that test. The monitor DEFINITELY makes a difference.
I've got two monitors connected to my video card, a Samsung 215TW and a Viewsonic VP912b.

Average test time for the Samsung - 339.4 MS
Average test time for the Viewsonic - 270.4 ms

That's a difference of 70ms!
Click to expand...

your high either way
 
mnewxcv said:
your high either way
Click to expand...

The point is that there is a clear difference based on monitor used alone. The monitor input lag issue is clearly not imaginary.

Besides...

Add to this the fact that Human reaction time is somewhere between 250ms and 400ms.
Click to expand...

If we take the baseline human lag as 250 ms and add device lag, those results are within human norms. Certainly some humans will have faster than normal reflexes, but that's not even the point of the post.
 
vjcsmoke said:
The point is that there is a clear difference based on monitor used alone. The monitor input lag issue is clearly not imaginary.

Besides...



If we take the baseline human lag as 250 ms and add device lag, those results are within human norms. Certainly some humans will have faster than normal reflexes, but that's not even the point of the post.
Click to expand...

but us [H]ard folk arent normal humans :D
 
Ok, mr. cyborg. :p Take the same test in the link quoted, specify the monitor(s) and or mice you used and show us your results!
I used a wireless Microsoft Natural Mouse 6000. Default windows vista refresh rates. I'm not even sure how you 'overclock' mouse polling.
 
yup, this little "human reaction" test isn't perfect. If the monitor lags quite a bit, the program will already register some good time going by before the green light appears for you to see. Not to mention if you have a wireless/slow mouse.
 
Great article.

As a FPS (and fighting game) player I always seek the lowest input possible without compensation.
I always thought LCDs had the biggest impact on input lag but I didn't know about the CPU and GPU times.

Of corse some of those contributing factors are out of out control unfortunately.
The best settings I can do right now is
-Use CRT or fast LCD. Currently using a VX922. Planning to upgrade to a BenQV2400W.
-use a mouse with high responce times or change USB pollrate. I use a MX518 with pollrate of 500hz = 2ms input lag.
-NEVER use Vsync. Using Vsync adds a noticable amount of input delay and I personaly cannot stand it even on slow paced single player games.
-I also reduced max render frames from 3 to 1 after reading this article.
 
.196 avg on a ten year old CRT monitor (NEC multisync at work) and a whatever bare bones vid card they choose to put in these rigs. No clickclickyclicky. :)

.185 avg. on my LG TN
 
Last edited:
You must log in or register to reply here.
Back
Top