Dell 3008 input lag

Holy crap, as high as 58ms and average 46.6. Good find Rinaldo. Well, there goes the wait for me with that LCD. The only other 30" not tested is the new 30" NEC MultiSync LCD3090WQXi. Although knowing NEC and their MultiSync line, I'd put down that it won't be any faster then 33ms input lag like the rest of their LCD's.
 
Isn't the 3008WFP supposed to be IPS? I've never seen or heard of lag that bad on any other monitor with an IPS panel. I guess it's possible, but those lag results are typical of S-PVA.
 
The type of panel has nothing to do with input lag. Input lag is caused by processing and scaling. When a monitor needs to dither for 24-bit, scale an image to non native resolution, or overdrive or any ghosting technologies, processing and sometimes buffering is involved. The reason why the 3007WFP had no input lag was because it had no scaler. They added a scaler in the 3008WFP so there could be multiple inputs and so it could display non native resolutions, so it doesn't surprise me there is lag.

Isn't the 3008WFP supposed to be IPS? I've never seen or heard of lag that bad on any other monitor with an IPS panel. I guess it's possible, but those lag results are typical of S-PVA.
 
The type of panel has nothing to do with input lag. Input lag is caused by processing and scaling. When a monitor needs to dither for 24-bit, scale an image to non native resolution, or overdrive or any ghosting technologies, processing and sometimes buffering is involved. The reason why the 3007WFP had no input lag was because it had no scaler. They added a scaler in the 3008WFP so there could be multiple inputs and so it could display non native resolutions, so it doesn't surprise me there is lag.

Not to mention the larger number of pixels that have to be processed.

All 30" LCDs with scaler chips will have more input lag than smaller modles, which means it will be real bad in some cases. I agree it is the pannel type; it is the processing behind it. For PVA this means anti-ghosting, overdrive etctra, so in many cases they are slower than S-IPS, but it is not the pannel itself.

These pannels are not entended for haevy gaming. If you want to do that, get one of last years modles with no scaler while you still can.

Dave
 
Good god...didn't think it would be that high.

Glad I got my 3007 when I did, image scaling can be done in software much quicker, and I can't tell the difference between that and hardware scaling.
 
The type of panel has nothing to do with input lag. Input lag is caused by processing and scaling. When a monitor needs to dither for 24-bit, scale an image to non native resolution, or overdrive or any ghosting technologies, processing and sometimes buffering is involved.
Yep, even many of current TN monitors have some input lag, for example in this test of 22" TN monitors most of them have input lag comparable to VAs:
http://www.behardware.com/articles/689-2/22-inch-lcd-monitors-the-3rd-wave.html
No doubt that's because of all spatial and temporal dithering processing required for generating missing ~16.5 million colours.


Also Dell 2707WFP and Nec 2690WUxi have lags similar to VAs:
http://www.digitalversus.com/duels.php?ty=6&ma1=52&mo1=149&p1=1606&ma2=88&mo2=174&p2=1747&ph=12
 
That lag is a real shame, I was hanging on in there for this monitor.
I had guessed there would have been some lag at scaled resolutions but I had hoped at its native 2560x1600 no scaling would be taking place....sigh.

It seems my quest for a 30" with low lag is faltering.
The 3007 cant be found anywhere in the UK. :(
 
The type of panel has nothing to do with input lag.

you are right up to a point however VA based panels have most of thier input lag due to the panel type, S-PVA is the worst in this respect and also uses the highest amount of overdrive than any other technology,

the 3008 results are horendous and pathetic.
 
In theory, the panel type shouldn't matter, but in the real world, certain panel types tend to lag more than others. Monitors with S-PVA panels always have lag. I have yet to see an exception to this. Monitors with TN or IPS panels tend to have less lag, but there are exceptions. I didn't say panel type was the only factor. I just haven't seen or heard of any monitor with an IPS panel with that much lag until now. It's disappointing.

A scaler should not add lag. If it does, it's badly designed. There shouldn't even be any scaling at the native resolution, and scaling lower resolutions doesn't take as much processing power as people seem to think. The Planar PX2611W has a scaler, and it doesn't have any significant lag, even when scaling lower resolutions. The 30" monitors have less than twice the number of pixels. There shouldn't be that much lag, even with scaling. Image processing such as deinterlacing and 3:2 pulldown can add lag, but that stuff shouldn't even be enabled for PC sources.

The only reason there's that much lag is because most people don't notice it. If a lot of people complained about it, you can bet they'd find a solution.
 
I wonder if they ran the test with the option in the Nvidia control panel that says "let the video card do the scaling"

I would hope that cuts out the monitor's scaler.
 
I wonder if they ran the test with the option in the Nvidia control panel that says "let the video card do the scaling"

I would hope that cuts out the monitor's scaler.

It would not make any difference at native resolution. that is the number that tells the tail.

Dave
 
So just to clarify what that number seems to show, for gaming it's probably better to get a 3007HC and maybe plan on a video card/CPU later capable of just shoving 2560x1600 at full details in Crysis rather than get a 3008 now and plan on playing Crysis at some scaled rez full screen on a 3008?

I mean long term, hopefully, there will be a CPU/video card combo capable of a Crysis type game with full detail @ 2560x1600 right? And if so, get a 3007HC while they are still around with almost no lag - and wait on the video card/CPU availability?

Decisions decisions!

I never worried about input lag at all - then I played CoD4 on my 2405FPW and got queasy from it. First time in 25 years of computer use that happened and while I cannot say for sure it was lag related, now knowing about input lag, I'd like to have as little of it as possible. If someone else makes me loopy playing games, so be it, but it won't be something I know about and can purposely buy less of.
 
So just to clarify what that number seems to show, for gaming it's probably better to get a 3007HC and maybe plan on a video card/CPU later capable of just shoving 2560x1600 at full details in Crysis rather than get a 3008 now and plan on playing Crysis at some scaled rez full screen on a 3008?

You can play at other resolutions with the 3007FPW-HC. Your video card can scale a lower resolution to 2560x1600 and sent it to your monitor at 2560x1600. I don't believe video card scaling adds any lag.

The only reason I see to get the 3008 over the 3007 is the ability to connect it with devices other than a PC. Console games would still have lag, but it may not be as noticeable when using a gamepad. My Gamecube connected with component to my 2405 wasn't too bad.
 
Having Crysis and playing it at highest settings with a top of the line computer is sad. I would wager it will be a good three to four years before we have hardware that will even be able to handle Crysys at full settings. You can always design software that can bring any hardware to it's knees. Another example would be Microsoft Flight Sim X. It has been out since 2006 and I don't think there will be hardware that can handle that at max resolution/settings for a long, long time.
 
Having Crysis and playing it at highest settings with a top of the line computer is sad. I would wager it will be a good three to four years before we have hardware that will even be able to handle Crysys at full settings. You can always design software that can bring any hardware to it's knees. Another example would be Microsoft Flight Sim X. It has been out since 2006 and I don't think there will be hardware that can handle that at max resolution/settings for a long, long time.

Supposedly with the QX9650 and triple sli you can get 30fps in Crysis at max settings. Flight Sim X is nothing special, just crappy MS code.
 
Ya, input lag isn't a big deal if your using Word 2008. But for those of us that play competative FPS's, it is huge. When I went from the laggy Gateway XHD3000 back to a simple interim TN panel with no lag, my gameplay skyrocketed. I have been playing FPS's and fast paced games for over fifeteen years on CRT's to every type of LCD. Having input lag numbers like the Dell 3008 makes it unpurchasable for me.
 
Ya, input lag isn't a big deal if your using Word 2008. But for those of us that play competative FPS's, it is huge. When I went from the laggy Gateway XHD3000 back to a simple interim TN panel with no lag, my gameplay skyrocketed. I have been playing FPS's and fast paced games for over fifeteen years on CRT's to every type of LCD. Having input lag numbers like the Dell 3008 makes it unpurchasable for me.

QFT. I can't use my XHD3000 for FPS gaming. I have been using my 19" CRT and my kill:death ratio is ridiculously higher than if I played on my XHD3000. Sad.
 
I've seen some measurements before this review site that had the 3008WFP sitting around 16ms of input lag. Can't remember where, it was just a forum post from someone who owned one, so I don't have a link. Still, it makes me wonder if digitalversus had it set up wrong (e.g. using the scalar by feeding it 1024x768 or something), and that's why it tested so slow. Basically, I'm not cancelling my order yet. I want to test it myself, and hopefully you will all benefit from independent verification.
 
You can play at other resolutions with the 3007FPW-HC. Your video card can scale a lower resolution to 2560x1600 and sent it to your monitor at 2560x1600. I don't believe video card scaling adds any lag.

Yup, you get a perfect 4 to 1 pixel mapping scaled resolution at 1280x800 and the Nvidia Panel scaling allows the video card to manipulate the output to fit the entire screen, just as a scaling chip would do.

Some complain that it's lower quality but I don't personally see that, I play a lot of my games at 1920x1200 and 1680x1050 if performance becomes a real issue and the scaling looks good, obviously not as good as the native resolution but that's impossible anyhow.
 
input lag isn't as big of a deal as it's made out to be. read this.

http://www.xbitlabs.com/articles/other/display/lcd-parameters_7.html

Thats interesting and i suspect largely accurate, however I think they're downplaying this effect in games, especially first person shooters.

They look at the sitution a little black and white, measuring 47ms from testing does not mean this is incured in real world gameplay, in fact these sorts of things tend to be amplified.

When aiming about we constantly adjust our aim based on what our senses are telling us about the game, largely our sight. If what is going on in the game is delayed by ~50ms then the information we're getting is less accurate and our response to it is not only going to be slower by ~50ms but it's also going to be less accurate.

Take for example 2 people shooting at each others heads in counter strike, headshots are quite often achieved by setting the crosshair down on the same plane as the head of the enemy and either waiting for them to strafe into your crosshairs, or strafing yourself to get their head into your crosshairs. This kind of timing for your shots makes it easier to achieve highly accurate long range shots compared to trying to move your mouse over a constantly moving target.

This has 2 serious side effects, firstly you might cleanly miss the head because the timing requires some level of constant prediction on your part, like with older online games when people played on 56k modems and you had to actually aim ahead of your target to score a clean hit (this was later removed with time stamped packets to adjust for latency)

Second of all, if you're aiming for a specific target accurately, rather than waiting for someone to strafe into your crosshairs (this is how more pro players aim) then with actions on screen happening later than your input, you might overshoot your aim but you don't realise until a small fraction of time later. This require re-adjusting your aim again which may very well overshoot again and require further re-adjustment until you finally tend towards an accurate shot.

The idea that a tiny fractional delay can't result in anything more than a tiny fractional reduction in gameplay is a logical fallacy, a small change when trying to aim for the head in CS can easily be the difference between scoring a deadly headshot and completely missing the head altogether.

If you want to experience just how bad small incriments of latency can effect your aim in first person shooters I highly suggest forcing Vsync on to sync your frames with the monitor, then enable tripple buffering which renders 2 frames ahead rather than 1 frame ahead, the same fraction of a second delay occurs but for most decent FPS gamers it makes it practically impossible to aim correctly, for me personally I always overshoot my aim, to aim with any kind of accuracy you have to learn to predict the delay time needed to alter mouse movements, and this is NOT static 47ms, it varies depending on how fast your enemy is moving and how fast your aproaching your target with your crosshair.
 
Do we have any new results for lag time with the rev a02, all these numbers are old and new comments from people state that it has improved. Any takers with a new rev?
 
I ordered a 3008 refurb yesterday (15% discount coupon + tax + free shipping = $1150 total). I'm planning to use it as my home office PC monitor and HDTV. Connecting consoles isn't an issue for me. There's no lag problem when running the monitor at native resolution, so I'm interested to see how it does with games. I've never noticed any lag on my 2407, or the 2405 I had previously, even while playing FPS games like COD4, Dead Space, L4D.

IMO, input lag is highly subjective. I've been using LCDs exclusively since 2003 so maybe my eyes are "LCD-ized" for lack of a better term. If I were a competitive FPS gamer who still used CRTs, maybe I'd be more sensitive to lag.

My real concern is how well the 3008 handles HDTV, videos and DVDs. According to PRAD's review, the monitor is excellent in this regard. We'll see.
 
There's no lag problem when running the monitor at native resolution...

Is that true?
Never made sense to me that a scaler would add lag in the native resolution, but noone ever writes at which resolution the comparison is made so I always had assumed native.

If it only has a bad input lag when scaling then why not use the graphic card scaler for all PC use - and consoles shouldn't be that sensitive for input lag anyway? I don't think I'd matter to me at least.

With the HP LP3065 now costs about as much as Dell 3008WFP (here at least) I'm thinking of buying the Dell - but the input lag really feels like a deal-breaker.
 
Is that true?
Never made sense to me that a scaler would add lag in the native resolution, but noone ever writes at which resolution the comparison is made so I always had assumed native.

We usually only refer to the scaler cause monitors without them are less likely to lag.
The lag can be caused by various kinds of signal processing. NECs for example buffer a frame for the LUT. Overdrive also needs a frame or two for calculation (reason for the lag tendency of PVA panels and probably why the Samsung 30'' without scaler lags a bit.), etc.

There are smaller models with scaler, without lag. But it seems the electronics on the 30'' models with scaler tend to be rather elaborate, hence the lag.
 
True, although when it comes to 30" IPS displays I haven't seen any exception of it. It's quite odd that just the monitors with scalers do have the signal processing that causes lag.
Though thats not the only thing thats different, I assume that all monitors with a scaler have an OSD which none of the monitors without a scaler do (right?).
In case of the scaler I don't see why it would introduce lag, blaming the lag on the high resolution of the screen barely makes sense to me - one could just use two 'ordinary' scalers that processed half the image and you wouldn't have lag due to the high resolution.
A poor mans scaler could also just incorporate a regular scaler but instead of scaling to 2560x1600 you could just do it to 1280x800 - with the loss of quality yes but it's very easy to do and at the time where graphic cards didn't have scalers and when 2560x1600 scalers wasn't feasible (for any reason) that would have made sense.

In any case it would be nice to know.

Gateways 30" who also suffer from great inplut lag apparently doesn't do this in it's native resolution (according to some). So the question of how the input lag affects by using different resolutions is still a valid one I think - as well as the question of whether the new revision has the same characteristics as the old one when it comes to lag.
Just the fact that the lag often is blamed on the scaler should put a focus on how this affects different resolutions and different inputs.

Almost all input-lag tests I've seen people have been mirroring the screen so that the same image is displayed on both monitors (the second monitor usually being a CRT). I don't know if modern graphic cards can mirror an image with different resolutions but assuming that they can't or that people haven't been doing that (quite probable, I think) then the 30" displays haven't been tested at their native resolution. This could be a real issue.

Edit:
Nec buffers a whole frame just for the LUT? :eek:
 
I can assure all of you that I'l give my 3008 a vigorous testing for lag, ghosting, etc. I had the Gateway XHD3000 for about a month before returning it when its PSU fried itself. I never noticed any lag on it during FPS gaming while many others claimed to see excruciatingly intolerable input lag. Can't think of a better example that lag is in the eye of the beholder. I'll post pics/comments when I receive my 3008 next week.
 
thanks to the link to the other thread. I do notice some lag when playing online against people vs just playing the single player mode. I always feel I am a step or 2 behind. I would only use the monitor at native resolution unless hook up a console to it but as of now I am strictly PC. I am looking forward to your tests next week.
 
10e
Thanks! :)
Now when I'm actually considering the 3008 I guess I'll have to go through that thread :eek:

Mr. Wolf
Sounds great, looking forward to your findings :)
I think that unlike the CRT the image doesn't go black after each frame has a big role in how people experience ghosting. Which I guess is why 120 Hz and black-frame-inserting is coming to the LCD (well, if theres anything to it except for PR).
For me, ghosting is never an issue on a modern LCD though :)
 
IF i got a refurb 3008 and the lag was too much for me would I be able to trade it back in and get the 3007 HC? I am really leaning toward the 3008 so I can use it for future things but my main use 90% plus is PC.
 
The 30-40 ms input lag readings from the other thread don't seem too bad to me. At the moment I have the 22" 1920x1200 Lenovo L220X that's supposed to have an average input lag of around 30ms. I just played Unreal Tournament 3 in single player against bots and noticed no problems. Then again I'm not a hardcore, reflexes like a freakin' cat super speed online player type anyway.

Would the input lag be higher or lower with a closer to native resolution like 1920x1200? What about if the graphics card is used for scaling instead of the monitor? Not that it is of utmost importance to me since I already have a 40" HDTV with around 10ms input lag for movies and games.

Apparently the only bad things about the 3008 are the input lag and some kind of anti-glare coating. Do other monitors like the HP 3065, Apple 30" or Samsung 305T have the anti-glare thingy? Because when looking at any of those in a store I didn't notice anything bothering about them.

What also bothers me is whether I'd feel that the 30" is too big. I'm used to sitting fairly close (about an arm's length away) to my monitors (21.3" 1600x1200 and the afore-mentioned 22" Lenovo).
 
I can assure all of you that I'l give my 3008 a vigorous testing for lag, ghosting, etc. I had the Gateway XHD3000 for about a month before returning it when its PSU fried itself. I never noticed any lag on it during FPS gaming while many others claimed to see excruciatingly intolerable input lag. Can't think of a better example that lag is in the eye of the beholder. I'll post pics/comments when I receive my 3008 next week.

Sorry for being impatient, just wondering if you've gotten it yet :)


Been thinking of throwing together a simple device consisting of a simple light sensor connected to the serial-bus or something, then accompanied by a simple application one could measure the lag of the screen, that'd be testing the response time and the input lag together though. Problem is that I won't be able to test any monitor that I don't already have bought :p

I'm wondering how people have been testing 30" displays that doesn't have a scaler. Either they'd test the 1280x800 resolution or they were relying on the graphic cards scaler, is it even possible to mirror the same resolution to two monitors but scale just one of them to another resolution? Sad thing is that half of the reviews of non-scaler 30" haven't even realized that they don't have a scaler...
Both ATi and nVidia does good job with my HDTV which only supports pure HD resolutions and they automatically sense this and never sends anything else (was a pain before accessing BIOS or installing the OS on the HTPC). But I don't know if that'd remain if I were to connect two screens and mirror them. In any case with a 30" screen with scaler you'd most likely would have to force this manually somehow. I lack a decent graphic card so I can't even check the options.
 
Actually, I got it yesterday!!! :D I probably won't be able to set it up until the weekend, tho. I've gotta do some rearranging to make space for it.

Here's a pic to tease you:

 
Okay, I've been playing with my 3008 for a full day. It's going back to Dell. The screen is full of dead/stuck pixels. I stopped counting at 50 and there are a lot more. A few comments:

1. Brilliant colors out of the box without any calibration
2. The monitor displays HDTV extremely well
3. I tried several games to test for input lag. My observations are based simply on what I saw with my own eyes.
  • Dead Space: Noticeable lag
  • Left4Dead: Less lag than Dead Space; had to scale back the graphics settings to get smooth gameplay
  • The Witcher: No lag
  • Mass Effect: No lag
  • FarCry 2: No lag
  • Deus Ex: No lag
Is it unusual for some games to show lag while others don't? I've got friends over to watch the games, so I'll post some pics a little later tonight.
 
Back
Top