some time ago i bought a 226bw samsung monitor (22" tft)

since then my qw-game droped a bit (might be due to inactivity ) anyways im thinking its in the hardware...

so i did a quick test (as i dont have cameras and crt monitor to do real testing)
http://www.iconinteractive.com/clients/react/

I dont really know if these results were the expected ones or not. And how accurate they are.
But would be interesting if anyone knew some tricks to reduce the input lag further (yeah, im not 100% sure the scaling actualy improved anything). (in qw, vsync off)

(ms)
nvidia scaling:
0.233
0.217
0.229
0.212
0.219

no scaling:
0.203
0.209
0.211
0.210
0.231

It's not impossible that the screen lags more than it should. If you are able to hook up a CRT and then clone the image, you can run a digital clock (with milliseconds) on screen and take photos. That way you'll know if it's the screen or if you're simply experiencing some kind of placebo effect. (Also, this post is possibly pure bullshit since I have no idea what nvidia scaling is.)

in nvidia controlpanel there is 4 settings. no scaling, monitor scaling,nvidia fixed aspectratio scaling and nvidia scaling. Think it decides whetever screen or gpu does the scaling.

just for the sake of completion..
i attached my old CRT monitor again and use them in dualscreen.

crt-150hz: avg 194.3ms
tft--60hz: avg 216.9ms (no scaling)
difftotal: 22.6ms

we can see that the avarage for tft is still about the same, but the crt monitor is 22ms faster.

According to tests (behardware) there is only slight video lag (10ms) present in the 226BW. The frametime difference between 150Hz and 60Hz modes is around 10ms. When you add these together you'll get your 20ms video lag consisting of the frametime difference and the time it takes TFT to process one image.

At this point there is nothing to do unless you can run the monitor at higher frequency. Since it's TN panel I'd guess it would go to 75Hz but you'll need to test it properly, as some TFT monitors can't do 75Hz (they skip every 5th screen update making them effectively 60Hz) even if they seem to sync at 75Hz. Even so you would be rather close to 20ms video lag.

This test is addictive...

Brought me to an idea of better measurement test. A blinking item on screen would give you a rythm, let's say 60 beats per minute, and you would have to press mouse button in the same rythm as the thing on your screen flashes. It would eliminate "human reaction lag". It could also get implemented into the qw client

johnny, rather implement the random reaction test.

If someone is offbeat, it wont tell you how far off you are

I also "upgraded" from a CRT to this very monitor and I dropped a lot in skill after that.

QW is just jerky nowadays.

Do you run it in its native resolution? I tried lowering the res but that just made it worse.

i used to run it in native res. 1680*1050

Still do, but have the old monitor for QW.

I got a similar monitor recently, a Samsung 22" LCD Syncmaster 223BW, and I found it completely impossible to play with. I'm sure I don't have my settings tweaked right, but still it was like a slideshow.

Yeah, my duel skill has dropped like crazy since I went from my CRT to this piece of crap. Sure, it looks good as long as you're not playing anything but for Quake it's like defecating in your underwear and then wearing them on your head for a day.

Anyone - TWEEKZ PlZ?!!?

Heard some crap that the Syncmaster 226BW has a random lottery with the quality? Any tricks & tips around that mess?

Well it's not totally random. They ended up having four different panels and only one of these was really good for gaming, two were pretty good otherwise and from what I read the fourth one was just crap all around. Haven't really digged into it further as I have had no intention of buying said monitor, but iirc there was information on batches and on for some panels markings on the back of the monitor indicating what panel it had but it's lottery in the sense that you can't see it from outside the box and possibly not always inside either.

And there is no trick The best panel has faired good in gaming latency tests, the three others not so great. One of the problems with all these tests is that they aren't testing qw. Qw is its own beast when it comes to shit like this

Yeah I understand QW is really lag-dependant. But I think this particular monitor had ~2ms lag which sounds totally fine by me

Actually no. Response time on TFTs means how fast the color of the pixel can be changed from black to white or from grey to grey. In addition to that, there is of course frametime (16,7ms at 60Hz and 13,3ms at 75Hz). On top of that there is also internal processing time for the image, and this one is the worst of the kind. This particular monitor had maximum of 10ms videolag due to internal processing. And not to be fooled by the manufacturers response time claims, it's not really 2ms response time except in few cases (certain color code transitions).

Add all this shit together (10ms+16,7ms+2ms) and compare it to CRTs frametime (depending on the refreshrate of course, CRTs have no video- or inputlag) and you'll see how much later the images can be.

On a related note, bear in mind that many TFT monitors will have an increased video lag when you use scaling (because it needs to process/stretch the image prior to display).
So assuming your video card is powerful enough I would ALWAYS recommend using the native resolution for QW (even compared to cases where the image should display correctly, e.g. using 840x525 on a 1680x1050 screen).