Dell UltraSharp U2711: Quality has a Price
by Jarred Walton on January 22, 2010 2:00 AM EST- Posted in
- Displays
Dell U2711 Lag and Response Time
Some users are very concerned with display lag and pixel response time. For others, they really don't notice anything unless a particular display is very sluggish. I fall into the latter camp, though I do notice the processing lag when it's above ~40ms (e.g. on the Dell 2408WFP and LaCie 324, my mouse input just felt off.) To test for display lag, we run the Wings of Fury benchmark in 3DMark03, with the output set to the native LCD resolution - in this case 2560x1440. We clone the output to our reference LCD, an HP LP3065, snap a bunch of pictures, toss out any where the time readout isn't clear, and then average the remaining results (at least 10, and usually 20 or more).
As a reminder, the reference HP LP3065 is one of the best LCDs we currently possess in terms of not having display lag. (The lack of a built-in scaler probably has something to do with this.) While we know some of you would like us to compare performance to a CRT, few users have CRTs these days and all we're really interested in measuring is the relative lag. It's possible we will find an LCD that ends up with a negative result, meaning it's faster than the LP3065, but the best we have managed so far is a tie.
So far, all of the S-PVA panels we have tested show a significant amount of input lag, ranging from 18ms up to 40ms. In contrast, the TN and S-IPS panels show little to no processing lag (relative to the HP LP3065). The BenQ FP241VW with an S-MVA panel performs similarly to the TN and IPS panels, with an average display lag of 2ms - not something you would actually notice compared to other LCDs. What about the new U2711?
We tested with the "Graphics" setting and Adobe RGB as well as "Video" and "Game" - we figured the latter might disable some post-processing and result in less lag. That turned out to be incorrect, as our measured lag actually went up 2ms. However, in practice the settings are pretty much tied. That means the U2711 has around one frame of lag relative to our best LCDs, but it still has a lot less lag than our worst offenders. As stated already, I didn't notice lag in using the U2711 - just like I didn't notice lag on the 2407WFP or the FPD2485W. I did notice slight lag on the Samsung 245T and some clear lag on the 2408WFP and LaCie 324, but the threshold for lag varies and you'll need to determine if 15-17ms is too much or not for you. If you've tried any TN panels and still noticed lag, we would expect every current LCD to be above your "lag threshold".
 |
Despite what the manufacturers might advertise as their average pixel response time, we have found most of the LCDs are basically equal in this area - they all show roughly a one frame "lag", which equates to a response time of around 16ms. Some transitions are faster than others, but the above is representative of what we found in a study of numerous photos. If you look at the tail of the center plane, you can see a slight ghost image before and after the dominant frame. Some of that will come from the camera (we use a 1/120s shutter speed), but most of it comes from the LCD panel. In this case, the U2711 panel does outperform most LCDs that we've tested, where it was often possible to see at least three frames more clearly than the two slight ghosts in the above image.
Facebook
Twitter
RSS
153 Comments
View All Comments
Mumrik - Saturday, January 23, 2010 - link
This seems very odd to me:"While we know some of you would like us to compare performance to a CRT, few users have CRTs these days and all we're really interested in measuring is the relative lag."
That is an incredibly weak argument for not getting proper numbers. I don't get it - why don't you care?
Anandtech would never only show the FPS scores of the 5870 as a percentage of a benchmark GPU, so why do this?
JarredWalton - Sunday, January 24, 2010 - link
1) I don't have a CRT.2) I have limited space.
3) I don't want to have a CRT - I ditched all of the ones I had about four years ago.
4) I would need a CRT that can support resolutions up to 2560x1600 -- none I'm aware of handle more than 2048x1536.
5) CRTs are terrible at getting correct geometry. Pincushion and trapezoidal distortion are all too common, even after lots of time spent fiddling to try to get it "just right"... and if you change resolution or refresh rate, you have to do it all over.
6) If CRTs are faster, add 20ms or 40ms or whatever to my numbers.
Lag between user and display comes from mouse, GPU, CPU, and LCD, really - up to around 200ms in some tests. You'll never eliminate all of it, and even CRTs have some "lag".
Besides, no one is making new CRTs that are worth buying, so why should we continue to compare to them? FWIW, my five year old CRT was getting dim when I got rid of it, so I couldn't even use it for comparison anyway. I'd need a new, high-quality CRT to make the comparison even remotely meaningful.
Mumrik - Sunday, January 24, 2010 - link
Its not really an argument about the merits of owning a CRT. It's about having a proper zero to benchmark against. I don't really see what most of those arguments have to do with measuring a lag time and "I don't have a CRT" is a very surprising reason to see on what I consider one of the nets premier hardware sites. Anandtech always seems to have all the esoteric equipment in the world to work with, both hardware AND testing equipment...This is not some sort of personal attack on you by the way.
Also, it looks like Sony's FW900 will do 2560x1600.
JarredWalton - Monday, January 25, 2010 - link
Here's an email I just sent to someone else on this subject (with a few edits):As far as input lag and the HP LP3065, if you trust another source you can find an LCD where we overlap, look at their result and look at my result, and then add the difference. But then, I'm not sure what other source I would trust, because I have seen sites report some of the LCDs I have as a "0" as anything from 0 to 20ms. None of them go into detail as to how they're testing (i.e. what stopwatch program they use, and how many pictures they take and average).
Anyway, here's one comparison to a CRT with the LP3065:
http://www.widescreengamingforum.com/forum/viewtop...">http://www.widescreengamingforum.com/forum/viewtop...
"I noticed no input lag at all with this thing. I took around 20 pics comparing input lag with a CRT -monitor and the worst lag I got was 24 ms, and the best was 0 ms."
So my choice of reference LCD has at most 24ms of lag compared to a CRT, and as little as 0ms, or perhaps an average of around ~12ms. But then, all CRTs aren't created equal, so what CRT do you use as a reference point?
I've seen other LCD vs. CRT tests where they post images of an 2405FPW and it gets an average of 35ms versus a CRT (http://www.hardforum.com/showthread.php?t=1047842&...">http://www.hardforum.com/showthread.php?t=1047842&..., which would mean the LP3065 is somewhere in the vicinity of 10ms average most likely.
Ultimately you're stuck trusting some source for your information about input lag. All I can tell you is that I've played a lot of games on the LP3065 and have never, ever noticed any form of lag. I've played games on a 2405FPW and barely noticed lag, and I've played games on the 2408WFP and definitely noticed lag. I could list a dozen more LCDs, and the fact is that not a single one has had less lag than the LP3065 in testing; they have only managed to tie it -- and many of the TN panels I've tested were recorded as "no lag versus CRT" at other sites.
If I ever find an LCD that has less lag than the HP LP3065, I'll make sure to mention it, but when I can confidently say that there's no current LCD that does better in that area, why beat a dead horse any more than we already have?
Every chart on the internet can be misleading if you don't know what you're looking at, which is why I explain in detail exactly what I'm doing. Plus, not all stopwatch programs are created equal; at best they are accurate to 17ms on an LCD, since it only updates the display 60 times per second. I tried one program and had results that were off by as much as 100ms, where I showed a difference of only 10 to 30ms with 3DMark03 as the time source.
So why not accept that the LP3065 is my zero point, and unless and until another LCD can beat it there's not much point in worrying about it. TN panels with "0ms lag" tie the LP3065, so it must therefore also have 0ms lag. If you trust my testing procedures, of course.
(And will I need to go over this AGAIN when I do another LCD review? A reference point is just that, and I can't find anyone that can give me a clearly better reference point. 10ms at most doesn't count....)
redbone75 - Saturday, January 23, 2010 - link
I don't even see how this can even be listed as a complaint, no matter how minor. To me, this ranks right up there with the classic complaining about the speakers on monitors. Should makers of ultra premium displays cater to people with less than stellar eyesight? Isn't the point of it all to be able to resolve finer and finer detail? It just sounds funny: "Man! This monitor is too sharp!"san1s - Saturday, January 23, 2010 - link
you are crazy, this is a valid complaint that should be noted in the reviewI'm glad that you have perfect vision- but many people, like me, don't. I would be angry spending my money on something only to find out later (since it wasn't noted) that I would have to change settings so becomes actually usable!
strikeback03 - Monday, January 25, 2010 - link
Well, resolution/pixel pitch is something each consumer should educate them on first. As it is a major point of this monitor, it probably isn't so much a complaint ("This is bad") as a caution ("This is something you have to know how to deal with")CSMR - Monday, January 25, 2010 - link
Yup, just "This is something you deal with by increasing the dpi setting" would be sufficient.Amazing that you get posters on AnandTech not understanding this. What will they say next? "Don't get that 1080p screen, your 720p movies will look small and your 480p dvd you can hardly see."
Voo - Saturday, January 23, 2010 - link
I disagree. Just because you have a stellar eyesight, doesn't mean the notice isn't justified, because this IS interesting for a lot of people.If it's so small that I have problems reading text that IS a problem for me and should be at least mentioned in the text. If it's not a problem for you, that's fine, you don't have to agree with the review in every point.
It just sounds funny: "I don't have a problem with it, so it's perfect for anybody!" Talk about empathy..
Voo - Saturday, January 23, 2010 - link
I disagree. Just because you have a stellar eyesight, doesn't mean the notice isn't justified, because this IS interesting for a lot of people.If it's so small that I have problems reading text that IS a problem for me and should be at least mentioned in the text. If it's not a problem for you, that's fine, you don't have to agree with the review in every point.
It just sounds funny: "I don't have a problem with it, so it's perfect for anybody!" Talk about empathy..