http://s221.photobucket.com/albums/dd258/mooocus/asus%20vw246/
lag test for asus vw246h
for you lazy fucks, here was the run down:
out of 21 pictures total:
15 had ZERO lag
3 had 16 milliseconds of lag
2 had 15 milliseconds of lag
1 had 31 milliseconds of lag
lcd was connected on its dvi. i really didn’t feel like disconnecting my xbox and hooking up a dvi-vga adaptor.
:woot:
Hitachi V700 (CRT) [DVI out to dvi->dsub adaptor->dsub in]
/ right rear / top connection to gfx card, primary OS monitor Versus
Asus VH236H / [DVI out->DVI in] / GAME MODE / Overdrive=100 sharpening=31
/ left front / bottom conn to gfx crd / 2ndary os monitor
1280x720 60hz (720p)- Nvidia Clone Mode Nv 9600GT (dual DVI out) exp=1/250 s
In both 720p tests, the primary monitor was connected to the top port of the graphics card, the secondary to the bottom port. However, in this test, the Hitachi CRT was the operating systems primary monitor, whereas it was the Asus in the first test.
The center of all 3 of the the hundredths place digit was centered horizontally on the screen, with the center of middle timer also centered vertically.
The numbers were determined with the same logic as my prev posted test (see legend/key) however, no descriptor is listed and the numbers are the average of the 3 timers, rounded to the nearest tenth. So, 0.3333=.3 and 0.6666=.7
So, a result of 1 clock being 1 frame behind, another half behind, and a third equal would be (1 + 0.5 + 0) / 3 = .333 or .3
avg/3|run total|result #
.8 .8 1
.3 1.1 2
.7 1.8 3
.3 2.1 4
.3 2.4 5
.5 2.9 6
.7 3.6 7
.7 4.3 8
.3 4.6 9
.5 5.1 10
.3 5.4 11
.3 5.7 12
.4 6.1 13
.7 6.8 14
.6 7.4 15
.2 7.6 16
0 7.6 17
.4 8 18
.5 8.5 19
0 8.5 20
.5 9 21
.5 9.5 22
.5 10 23
.7 10.7 24
.3 11 25
.5 11.5 26
.3 11.8 27
.2 12 28 top slightly ahead, weak .688 with strong .672; .672 on CRT
0 12 29
.5 12.5 30
.3 12.8 31
.2 13 32
.7 13.7 33
.2 13.9 34
.3 14.2 35
14.2/35 = .406 avg frames behind CRT
.406*16.67ms=6.77 or ~7ms
These are great results! Especially as it’s assumed the secondary monitor gets the short end of the stick.
I didn’t notice any different patterns in lag, meaning it didn’t seem as though the upper or lower portion of the screen seemed behind relative to the CRT. However, I think 3 timers increases the accuracy, as you get 3 numbers per picture to even out the results. Or to put it another way, I’m guessing it allows for about the same accuracy as a single timer with 3 times as many pictures.
Averaging these results with the 4 ms estimated average with the primary/secondary and top/bottom connections reversed, you get 5 to 6 ms average lag. Not bad at all. And this isn’t the monitors native resolution!
I’d guess the vice-versa primary/secondary 1080p would probably show a similar results, based on it showing a similar result to the 720p test above (4 vs 5 ms, which is probably actually equal, and just a variation that wasn’t ironed out by a larger sample set of numbers)
I’m looking forward to shinshoryuken’s results. I’m wondering how much they’ll deviate from my numbers.
Unless the HDMI numbers come up significantly worse, I’ll be happy with my purchase unless the monitor breaks or something. Oh, and I’ve gotta try lag in “scenery” mode, as it produces much darker blacks… I’ll try to get around to it sometime. I’d prefer to game in scenery mode if there is no or little penalty.
They’ll fit if you get rid of the status bar and put the bottom edge of the higher timer on top of the top edge of the lower timer.I shrank the top one a bit though, so that I could have the center timer perfectly centered. I don’t think it really made any difference though, after looking at the results, but I can’t say so conclusively.
OK, so 5 images 1 frame behind, and 1 two frames behind, so 7 total behind out of 21 means an average of .333 frames behind so 1/3rd of 16.7ms would be 5 to 6 ms avg. Just curious, was this in game mode? And what resolution?
Has anyone seen any results of the 26" version of that monitor?
Someone did a few pages back. They say its just like the “evo model” but its 16:10 not 16:9 so you will get black bars on the top and bottom, unless you do full screen and stretch the image…like make honda and rufus look skinnier.
resolution was 1920x1080 in standard mode.
as far as i know, the different modes are more or less just different preset color/contrast/brightness levels and don’t affect post-processing much, if at all. feel free to correct me, though.
RESULTS THROUGH HDMI…
Hitachi V700 (CRT) [DVI out to dvi->dsub adaptor->dsub in] (AT 1280X1024 WITH 1080p SIZE DESKTOP)
/ right rear / top connection to gfx card, primary OS monitor Versus
Asus VH236H / [DVI out->DVI to HDMI cable->HDMI in] / SCENERY MODE / Overdrive=100 sharpening=31
/ left front / bottom conn to gfx crd / 2ndary os monitor
1920x1080 60hz (1080p)- Nvidia Clone Mode Nv 9600GT (dual DVI out) exp=1/250 s
This time rather than rounding 3 frames to 1, I added them up and showed the result out of 3.
So. a frame ahead (-1) equal (0) and a half frame behind (.5) would add up to (-.5) A few frames did have examples where part of the image was a full frame ahead. Note that though it shows 75 examples, this was only out of 25 pictures.
For information on how I came up with my numbers, see the Legend/Key from my first test.
-.5 -.5 3
1.5 1 6
1.3 2.3 9
0.3 2.6 12
-.5 2.1 15
1 3.1 18
1.5 4.6 21
1.5 6.1 24
1 7.1 27
1.5 8.6 30
2 10.6 33
0 36
-1 9.6 39
1.5 11.1 42
2 13.1 45
1.5 14.6 48
-.5 14.1 51
-.5 13.6 54
1.5 15.1 57
0 15.1 60
-.5 14.6 63
1 15.6 66
1 16.6 69
-1 15.6 72
1 16.6 75
16.6 frame values behind of 75 examples
(or relative to the previous test, 5.53 frames out of 25 had I scored it the same, the results add up the same, just look different)
16.6/75= 0.2213 = 0.2213*16.67 = 3.69 = about 4ms average delay, assuming the ports average out as “synced” over the set span of the 25 images taken. Now, this was done with the Asus as a SECONDARY monitor, which ranked it a slower result in the 720p test relative to it being the primary. So, if it were the primary, we MIGHT even see LESS lag!
HOWEVER, these results might be skewed by the fact that the CRT couldn’t do 1920p and was doing 1280x1024 showing a partial desktop of a 1920x1080 desktop size.
Based on the first 1080p results I had averaging only 1ms higher at native resolution (than at 720p) (4 vs 5ms), I’m guessing the CRT in this mode is still about as accurate as a CRT in 1920x1080. Feel free to point out if I’m wrong.
ALSO… this is very nice, that these tests were run in SCENERY mode and not GAME mode! So, for darker blacks, I think I’m going to be gaming in scenery mode now. :lovin:
As far as my results show, I’m happy to assume that you are correct! :rock:
Wondering what my results would be if I did the "throw out any non-clear results (mixed/split numbers on timer) and reinterpreted the raw data from my first test. (720p)
It throws out 40% of the results:
0 0 0 1
1 -1 -1 2
0 0 3
0 0 4
1 -1 -2 5
0 0 6
1 -1 -3 7
0 0 8
+ +1 -2 9
0 0 10
0 0 11
1 -1 -3 12
0 0 13
0 0 14
0 0 15
0 0 16
0 0 17
0 0 18
0 0 19
0 0 20
0 0 21
0 0 22
0 0 23
0 0 24
1 -1 -4 25
0 0 26
0 0 27
0 0 28
0 0 29
1 -1 -5 30
5/30 = 0.1667 = 0.1667 frames lag average = 0.1667*16.67 = 2.78ms average lag, so a bit lower of an estimate. Keep this in mind when you go over my numbers, if I had crunched them like other sites do, they’d likely be even lower, as there are many results that have two numbers superimposed, or split down the center (horizontally).
***Because of this, you must factor this in when comparing my numbers to other sites (like Digital Versus) as my numbers aren’t as forgiving. ***
Your non forgiving numbers still look good, I have to say.
I’m a little confused though, 1080p first test netted you 5ms, 1080p second test HDMI secondary netted you 4ms?
What shutter speed is ideal to use on the stop clock test?
I’m camera illiterate and I have no diea how to change the shutter speed.
Also, how come the last ##X of the stop watch moves really slowly and for me when I take a picture of the LCD vs CRT, the last X of the stop watch is always the same.
Starting on my Custom Vewlix SF4 Cabinet…
Thinkin about using THIS: http://www.newegg.com/Product/Product.aspx?Item=N82E16824254026 monitor. anybody have experience with it? feedback? any reviews?
28" is an odd size…why won’t you go for 32’?
This new lg monitor looks pretty interesting
http://www.lge.com/products/model/detail/W2363V.jhtml
They actually advertise no input lag
Well, there are two big differences.
The first test the monitor was through the DVI input on the monitor (as opposed to the HDMI)
The first test was with the LCD being the Operating system’s primary monitor. The second had it as the secondary. It would seem by the 720p results that the secondary is a tad slower. Other factors lurk I’m sure that are beyond my understanding of the complexities of graphics rendering.
Now, this might be huge, or it might be nothing, I don’t know. But for latter tests, I had the nvidia driver set to from 3 frame max pre-render to zero. Anyone have insight on this in regard to relevance?
This is probably up for speculation. One could argue that the human eye’s visual perception could not be capable of really distinguishing something that happened in 1/250th of a second from the things before an after it (save for a huge change in exposure, like a flash or strobe).
A coin spinning on a counter looks somewhat like a shiny translucent sphere to the eye, yet in a 1/250th second image, it looks like a coin standing upright and still.
However, one could also argue that as you get too low in shutter speed, you’ll have to much information blurred together. Take a picture at 1/15th of a second, and you’ll get around 4 different numbers superimposed on eachother. Worthless for this test.
So, I figured the higher the shutter speed, the clearer the results. I would have gone higher, with my cameras max aperture of 2.8 and max ISO of 1600 combined with the fastly dimming pixels of the CRT…
All things said and done, I picked the highest shutter speed where I could still make out what the dimmest of the numbers where on the CRT clearly. Which just so happened to be 1/250th of a second.
The room doesn’t need to be bright. better too dim than to bright, so that reflections from the screen don’t obscure the numbers on the CRT screen. Also, a flash is worthless and counter productive. Disable it for the images.
What model camera do you have? I can probably figure out how to set it manually.
LG seem to have a good reputation with having low input lag.
I’m starting to see a pattern…it seem like most 24" and under TN monitor have the least input lag. Every thing above 30" seem to have more lag.
is that really true? or is the testing method favors the smaller monitors and the bigger monitors gets the skewed results?
I’m using a nikon coolpix p80 and a cannon sd550.
THANKS!!
I like that they have recognized a niche and are actively thinking about addressing the needs of minimal lag for gamers.
I don’t like the fact that there aren’t any specifications at all dealing with input lag. If it were actually a great minimal lag monitor, they would give us some numbers because if those numbers were good, they’d sell the thing. The fact that they don’t leaves me thinking of that monitor like I think of any other LCD monitor. Assume it has significant input lag until you see solid numbers showing otherwise.
2ms response time means nothing by itself. It could have ~60ms lag if it buffers a few frames. As does “None Delay thru mode” which is a trademark, not a specification. As is, I see it as marketing hype.
I’d love to see this be a great gamer monitor. I’d love for it to make huge sales because of minimal lag and have other manufactures take note that they’re loosing a slight chunk of market share to purpose built minimal lag gaming monitor. And I’d love for those manufacturers to create there own, competing for whose is best, and there bringing lag and prices down.
However, I just won’t believe any of it until it’s proven.