LCD input Lag problems: new tests of 2405fpw and 244t

Sluuut said:
This is exactly why i choose for an S-IPS screen and 23" instead of 24"..

Very nice review/proof !

I wish there were more like this. And i dont understand why reviewers on like for instance Toms Hardware Guide don't even notice this !!
Click to expand...

I wish there were more like this. And i dont understand why reviewers on like for instance Toms Hardware Guide don't even notice this !![/QUOTE]

The panel technology being S-IPS, TN-film or PVA/MVA and this input lag of which we speak are two different things. Depending on the type of LCD panel, there can be large differences in the pixel refresh depending on whether it’s going from black to white, or rom gray to gray. This can cause a burry image or color shift on fast moving images but the overall image (on newer faster displays) is where it needs to be, for the most part. We would all love to see sub ms responses time for all color changes and may well some day soon but that is not the issue here, I think.
 
chuey101 said:
I was very disappointed at first when there was a really noticeable input lag to the point of the game being unplayable, but after doing the following things:
- uncheck the following boxes in ATI Catalyst Control Center
Reduce DVI frequency on high-resolution displayes
Alternate DVI operational mode

Another thing you can do is to go in to the OSD of the monitor and select the name of the input as "Game", supposedly this results in the monitor operating in a more optimized fashion, reducing the overdrive usage that is apparently to blame for the horrible input lag.

After following these steps, i found no noticeable input lag in HL2 and it was quiet playable, there was maybe a smidgeon of input lag, but after I turned up my mouse sensitivity (it was really low b/c my previous monitor was tiny) I couldn't notice it anymore. I ran through COD2, GRAW demo, HL2, and will try BF2 tonight. So far, so good!
Click to expand...


Could anyone please test and measure how much those things affect input lag?
 
oRdchaos said:
The 2405 noticably lags behind the 2005...
Click to expand...

I don’t have a good way to know for sure but that is the same kinda thing I saw when using different outputs and why I split the composite signal to all three displays. When doing so, I see that they look exactly the same. Look at the pic I took, the image on the two Dells look exactly the same to me.
 
Cuthalu said:
Could anyone please test and measure how much those things affect input lag?
Click to expand...
Those settings won’t affect input lag at all. Input lag would be the time from the video card sending a new frame and you seeing that frame on the display. Any card tweaks may help reduce how noticeable it is by bring the over all time between creating a frame and seeing it but I think/hope that the video cards lag or buffering would be way under a ms or .001 seconds and we are talking about display introduce lag or buffering of .05 seconds or 50ms. Also the video cards buffering should be a constant as when the card is over loaded it drops frames, they don’t get backlogged.
 
Tanquen said:
Those settings won’t affect input lag at all. Input lag would be the time from the video card sending a new frame and you seeing that frame on the display. Any card tweaks may help reduce how noticeable it is by bring the over all time between creating a frame and seeing it but I think/hope that the video cards lag or buffering would be way under a ms or .001 seconds and we are talking about display introduce lag or buffering of .05 seconds or 50ms. Also the video cards buffering should be a constant as when the card is over loaded it drops frames, they don’t get backlogged.
Click to expand...

That may be the theory, but above stated empiric observation says something else, and therefore it needs to be tested and measured. And changing those things do change what the monitor receives and it takes less time to output that image, and there could be an explanation for possible difference?
 
Cuthalu said:
That may be the theory, but above stated empiric observation says something else, and therefore it needs to be tested and measured. And changing those things do change what the monitor receives and it takes less time to output that image, and there could be an explanation for possible difference?
Click to expand...

Your are right I can’t know for sure but my limited knowledge of the signals involved and the scalars wont let me believe that you could lesson the displays input lag by tweaking the signal. The scalar should be constant, 60 frames in 60 frames out. Maybe something like a static image or a sold blue screen might get processed a little faster but should not change anything. Maybe… but it’s encroaching on stereo salon territory with guys telling you that a gold plated $1000 fiber optic audio cable is directional, eliminates all jitter and gives you a much brighter sound stage.

Just joking around stereo guys. :)
 
I may see if I can get a VGA splitter to test this. No offense, but I think composite output is too different to be a valid test for the presentation capabilities of these monitors. Since the delay seems to increase as the number of pixels on the monitor increase, and it seems to be a function of the processing done in the monitor, I would think the fewer input 'pixels' of an analog composite input would be simple for the monitor.

It does help to rule out the scaler as a factor in the displays maybe... I may try to run tests at different resolutions and see what comes up.
 
There is no way to know but I am already seeing an input lag with the composite and I did see that when I put the 2405 on DVI and ran composite to the 2005 and TV. The 2405 was a little faster with the 2005 being the slowest and the TV was still faster. Now, that could be the frame is coming soon from the cards DVI output than the composite or the Display has less lag on the DVI input than the composite. I think the display would be able to scale the composite and the VGA basically the same.

Of course sending composite to all three is not my first choice. But I don’t have a DVI or VGA splitter and even if I did I felt that the TV would be a desirable base (least likely to add in any input lag if any) and I’d have no way to compare it to the others without relying on the card to send out the same exact frame on the DVI and the composite.

On a side note I did see a guy that did some real detailed test on the 2407 and found that it has less input lag that the 2405.
 
Tanquen said:
On a side note I did see a guy that did some real detailed test on the 2407 and found that it has less input lag that the 2405.
Click to expand...

Do you have a link to that?
 
Sorry no, it was many weeks ago but I'd not get it as there are many reports of color banding and soft text. :( All the info that I've seen says that the input lag is better though. All I want is a perfect display for $500 or so. Is that too much to ask? :)
 
Could you try the following to test if there's any differences to your current results: uncheck the following boxes in ATI Catalyst Control Center
-Reduce DVI frequency on high-resolution displayes
-Alternate DVI operational mode
 
A definite way to do the dvi test would be to use a dvi splitter from one dvi output. Clone mode on dual dvi output card would be the other disputable alternative.
 
Askanison said:
32ms difference


Using vga splitter to 3 monitors at 1920x1200 60Hz res
Click to expand...

if thats not proof then I give up trying to explain to people that there is lag.
 
Has any one had any luck contacting monitor manufacturers about this to see if there's any setting or mode that would improve this?

I wish some website *hint, hint* would take this up and use their connections to get to the bottom of it. It seems almost like some monitors buffer 2-4 frames to look ahead and make sure that pixels don't get changed to a color that will look too out of place in the upcoming frames.

Having gamed on a 25ms response time LCD for a while, I think I'd rather have ghosting, than this buffering/'input lag'/video processing delay.

And FWIW, This is a problem in the home theatre arena for people with advanced displays and image processors...so many audio receivers have a feature to delay audio 'x' number of milliseconds so that it syncs up.
 
Loque said:
Reviving this excellent thread to add some observations about input lag on other non-VA type panelled LCDs, from the perspective of someone who plays fast fps games and has noticed input delay while gaming on more than one LCD.

Recently I borrowed a Samsung 204B and found the lag to be a significant problem, especially when aiming precision weapon in the thick battle in UT2004 and other fast games. I researched the topic a bit and found the test methods for input lag here so I started my own informal testing. Unsurprisingly, the 204B had 35-40ms+ latency in comparison with my Samsung 900IFT CRT monitor using the clone-mode-stopwatch method. Maybe this explains why playing offline with fps games felt like I was online with a mediocre connection. Take note: this is very noticeable lag on an LCD with a fast TN panel.

I also tried and tested the Viewsonic VX922 but unfortunately my findings were inconsistent so I hesitate to say anything. Subjectively I felt there was some noticeable but minor lag (nothing like the 204B) in comparison with the CRT but in testing I measured consistently fairly low latency averaging 15-20ms. I have another few hours with this monitor so maybe my subjective impression will change, but for now I am inclined not to keep this it (I am also not thrilled with the washed out colours and difficulty finding acceptable brightness/contrast and color settings).

Finally I also tried the new LG 1970HR, also with TN panel, but which reportedly has no 'overdrive' technology according to a review at chip.de (I find this hard to believe since LG is claiming '2 ms' RT). Subjectively I detected no noticeable lag on the 1970HR whatsoever and found it almost as good as my CRT, apart from the obvious LCD limitations: lower frame rate cap with vsync on, unbearable tearing with vsync off, and some typical but non-intrusive LCD motion blur (no ghosting that I noticed). I measured low latency in tests against my CRT, in the range of 15-20ms.

Conclusions: as some have remarked, it is really unacceptable that otherwise thorough hardware review sites (tomshardware.com, prad.de, hardware.fr) reviewing 'gaming' LCD monitors are not taking input lag seriously and attempting to measure it on units they review. I may be in the minority as a player of fast fps games, but my experience with 5 or 6 recent LCDs in the past is that ghosting and motion blur is getting to be a relatively MINOR problem despite the being the preoccupation of reviewers. On the other hand I've found input lag is a real and significant problem with no official attention being paid to it and no manufacturer specifications (comparable to response time used to predict ghosting, etc.) to guide consumer purchases. In the case of my own purchasing decision, I still want to try an LCD with IPS so if I find one to test I will report back. I am not holding my breath that an IPS panel (or any panel technology) 'solves' the problem either, since I have seen complaints online of input lag on the Dell 2001FP, for example, and one web site tracking input lag using a different methodology reports a 55-76ms latency with the 2001FP: http://www.draebenstedt.de/reaction/loopback.html.

Questions: (1) What are the implications of using a less fast digital camera apart from possibly needing to sort through more unusable photos? I mean, if you capture legible digital stopwatch readouts on both the CRT and LCD screens the measurement is valid, is it not? (2) How would one explain a photo capturing identical stopwatch time readouts on both screens if digital lag in fact does not fluctaute? Would this be simply an incorrect reading of a botched photo or could it be explained by the 17ms or so inaccuracy due to measurement error/limitation surmised by okashira? (3) How does vsync affect perceived lag? I felt there was less lag with vsync off on my 204B but tearing was severe, and I noticed that testing at the above this site again shows higher latencies with vsync on: http://www.draebenstedt.de/reaction/loopback.html.

We really need a site to compile input lag measurements for different LCDs...

EDIT: After playing more on the VX922 I not longer subjectively perceive any input lag. I think it didn't feel right before because I first tried the VX922 right after switching video cards to a 6600GT with dual dvi to do some comparative LCD testing, and it turns out I was getting worse frame rates with that card than my normal vid card. I feel better now that subjective perception has come into line with measurements from the objective testing laid out in this thread. :cool:
Click to expand...

One thing to bear in mind is refresh rates. In order to get truly comparable / accurate results when comparing two monitors, you'd really need to have a refresh rate in line with the accuracy you are trying to achieve - ie, you'd need a refresh rate of 1,000Hz if you're trying to measure down to thousandths of a second. This explains why you get quite a bit of variation when taking multiple images. Of course, if you take several images and average the delta then you will probably get reasonably accurate results...
 
On the refresh rate thing, that's why in the picture of the 2405 v. 2407 you see a 32 ms lag between them. That's effectively 2 'frames' at 60Hz, eh?
 
Hi guys,

I got a Samsung 215TW since Saturday. Wail I wont denie that there isnt lag behind a CRT monitor its definitely not as bad as people claim here for me. On desktop when moving the mouse I really cant tell if I see the lag or not. I tried 2 different mices a MS Intelleye explorer 3.0 connected into an USB 2.0 port and a Logitech G5 connected to same USB2.0 PCI card. I really cant tell if there is lag or not on the desktop itself. Looks like none to me however moving a mouse on a CDT is smoother and you see the full cursor moving. Here you just see it as full if you move slower. Once you move it fast it disappears if thats what you meant with lag? However I see virtually no delay in reaction time or reaching it destination time compared to my hand.

However I did notice a delay with my Wheel in all the racing games I own compared to a CRT monitor and this may drive me crazy I dont know yet. If the games change the physics due to delay it will of course be horrible. Have to test this more.

I'm also downloading CSS again and will see how its ingame.
 
You can feel it really easily in CSS. I was wondering why my movements, particularly strafing, were definitely off. Most of the lag I felt came in the keyboard input, and NOT the mouse.

Which is a bit odd, but regardless, it is as someone else said, "like moving through molasses."
 
Well I dont see a big deal in CSS either. I always had a wirless keyboard so I probably wouldnt notice the delay with the keyboard anyway.
I just hope that racing games will work normal since the track we are racing today I got problems with sliding out and hope its just the weather lol.
 
I was wondering if you ever get used to this? I mean my monitor was sold as great for games, movie and all around work lol. What a freaking lie!
 
I have to agree with the other poster.

to test for millesecond lag, you need a source that is about 10x faster.

the test right now is probably ok to about a 50ms accuracy, (meaning you can tell a differeance)

you need a clock that can run even faster then the one used to test the source right now.

I don't know how toms does it, but it would be best to get it under a scope (both the source, and a pixel) and see what the delay from command to output is.

Also, you would have to figure in the panel delay. because you are taking a picture of the pixel change.

so say for example you have a 12ms display (which toms mesaures and says is about 20ms real world change).

lets just take ops numbers. if the crt shows 40ms and the lcd shows 20ms) there is no input delay, because you have to factor in the 20ms delay of the panel changing.

basically, you would take the delay measured by the camera, and the delay of the panel and subtract.

This would make sense. A faster display, would hide input delay better.

a display that for some colors is slower then others, would show a more random time.

A slowerish display would make the problem worse.

and a slow display would blur everything enough that you could not notice.
 
Well I took the Samsung 215TW back to the shop and got my money back. This monitor is simple crap! Not good for games and this 60ms delay or maybe more is so bad I couldnt race my Nascar 2003 game on the limit! 3D shooters were somehow ok I guess you get used to it with time and also desktop use but driving sims when driving on the limit were horrible. Where you need instant respons if you dont wanna spin or you will have to drive slower and loose lots of time. NOWAY!!! down with S-PVA
 
Seems strange, the 2407WFP and 244T share the same samsung panel yet the 2407 has less input lag over the 2405 and the 244T has more... :(
 
Shocky said:
Seems strange, the 2407WFP and 244T share the same samsung panel yet the 2407 has less input lag over the 2405 and the 244T has more... :(
Click to expand...

They may have the same panel and the same level of ghosting or color shift due to the panels performance in switching from one color or brightness to the next. This is different than the input lag that is added buy the driver/scaler chipset and input options added to the panel telling it to change from one color or brightness to the next. The chipset is buffering too many or enough frames for some gamers to notice. Each branded display may be using the same panel but be using a different driver/scaler chipset and or input/output options causing more or less input lag. The driver/scaler chipset can exacerbate the input lag on any display technology, LCD, Plasma, DLP, CRT...
 
Yeah, that difference is significant. I know the 2405 are using the L1/M1's while the 2407 and 244t are on the L2/M2's (with the 244T's being on the M2's specifically).

Got that from here: http://www.flatpanels.dk/panels.php (search for LTM240M)

Too bad that site doesn't know exactly what chipset each specific LCD uses...
 
so is there a LCD without input lag at all or does this come with the package?
 
Huxxx said:
so is there a LCD without input lag at all or does this come with the package?
Click to expand...

Any digital display will require some signal processing. It’s all about how much and what kind of money you are going to put into it. If you get into some of the high end home theater stuff you’ll see stuff like this: http://www.dvdo.com/pro/pro_acc_abt-102.php

“And for Home Theater Gaming applications, the Precision Deinterlacing card can be set to minimize the real-time response with maximum performance.”

"Precision AV LipSync™
ABT's Precision AV LipSync technology automatically delays the audio signal to match the video processing delay in DVDO products. As a result, the AV Lipsync problem that is caused by video processing delay is eliminated. The audio delay can also be further adjusted to correct the AV Lipsync problem that may be already be present in the AV source."

My Denon reciver has this kind of LipSync fix where you can buffer or delay the audio so it will match the delay introduced buy the video signal processing either from the scaler in the display or an external on like this: http://www.dvdo.com/pro/pro_isvp30.php

Dell and others are trying to get the most out of the cheapest parts they can and with some of the new LCDs have pushed a little too hard. They are still fine for most everyone and it looks like they have even tweaked the 2407 so it has a little less input lag but could still be better.

You’ll just have to try it and see if it bothers you and keep an eye out in the forums to stay away from the real stinkers. I do notice it a little on the 2405 and wish it were not there but after playing SC: Source for a time I forget about it.
 
I cant try it otherwise then if I take it home. They wont let me run my racing game lol and even if I had to have the whole setup with wheel there to test it fine. Like I said before Its maybe fine for office use and it may be even fine for 3D shooters at some point since you can get used to the delay but this games are not dependant to control input that much. But its unapcetable for Sim Racing games where you are always on the edge of a spin and if the input there is slow you will be either slow or out of control. Imagine the game runs with high physics and you drive on the edge. Now with this monitor all what is displayed is delayed and when the physics tells you left you still see right on the screen and react so. 60ms is alot.
 
Huxxx said:
...Its maybe fine for office use and it may be even fine for 3D shooters at some point since you can get used to the delay but this games are not dependant to control input that much. But its unapcetable for Sim Racing games...
Click to expand...
Not defending the display manufactures. Everyone should send all the nasty emails they can to them. This should be fixed. In the meantime…

I have to totally disagree, when there are three or more guys running around coming at you it is very important that you get that head shot and you have to get the crosshair just in the right spot. I only used to play CS on CRTs and it’s still just as hard and I still think net lag can be just as if not more painful. It is important for any game that requires critical timing. I’ve played many racing games and have a racing wheel and all that. It’s not the greatest one but it works and the same kind of stuff applies. Humans are not that fast in the fist place, many animals and insects have much better response times. When racing I still have to anticipate the turn coming up no matter what. I’m not trying to say that the input lag will never have some effect but I really think that you can compensate for all most all of it. Humans are the best at that, if you can get them to try. :)

“I cant try it otherwise then if I take it home. They wont let me run my racing game lol and even if I had to have the whole setup with wheel there to test it fine.”

What?

Some places have return policies. Dell has a no questions asked 30 day return policy. Try the 2407 for a few weeks and then you will know. Only you can tell if there is too much input lag for you.
 
That sounds like a good idea but dell take a long time to collect monitors, I asked for a refund about two weeks ago and i've still got three 2407 A02's boxed ready to go... Contacting them every week but there not collecting them.,..maybe they don't want their faulty monitors back. :p
 
Tanquen, seeing the lag is enough bad. Also why choose to try to adapt when there's other options available.
 
Cuthalu said:
Tanquen, seeing the lag is enough bad. Also why choose to try to adapt when there's other options available.
Click to expand...

Sorry, for me simply knowing of an issue is not bad enough. I went through 4 2405 LCDs to try and get one with no or fewer stuck on pixels. I ended up keeping the first one that they sent me as it had only two or three and I can not and have never seen them in daily use of the display. I have to turn off the lights put up a solid black image and slowly scan the screen to see the little dark red dots. The 2405 and the 2005 are the first LCD displays that I have used that had stuck on pixels and this really upset me at the time but I got over it. That being said, a cluster of two or three stuck off pixels in the middle of the screen would be unacceptable regardless of there seven (or is it five) dead pixel standard.

LCD Ghosting really bothered me at first too but the other positives of a digital display where enough for me to move to LCD from CRT. I still see ghosting from time to time and I hate it but I can forgive it and hopefully soon ghosting will be a thing of the past along with input lag and black levels.

Where is the alternative 24” LCD with the same features less the bugs for $700 (if you can wait for a sell) delivered?

Again, this needs to be fixed and the 2407 is a little better but we all need to bug them and let them know that this is important to gamers. We may even see a rating for input lag in the future but at the moment it’s not affecting there sales and there is not enough backlash from the enough gamers. We can barely get review sites to mention it if at all.
 
It will come with time when you can no longer get a CRT monitor. Here in our shops they only sell LCDS now since some time so this 19 Samsung is my last one and in the future I will go for TN panel which one tested had about max 11ms lag with average of 2ms and minimum of 0ms. Thats really not that bad. But not every monitor with NT is that fast either. So right now from the few tested the Viewsonic VX922 was the best. Some are similar slow as IPS panels and some IPS panels are even slower then the already slow MVAs.

What we must do to get an effect out of the manufacturers is to write every hardware LCD reviewer out there to test this input lag on each monitor he gets. After time manufacturer's will get minus points and some will have more then others. Thats the only way I see how to influence something. At the moment only one review on the whole www reviews this problem really. Sad but true :confused:
 
Yeah well thats so opposite from what I was told here on the forum and elsewhere. People always claimed that response time has nothing to do with input lag.
But look at this table now. The fastest monitor is still the 2ms response one followed by a 2ms and 5ms and so on. So a Samsung 2ms one should be ok.
 
im bumping this to see if any more to add.
any monitors free from satan's grasp?

shit i might as well cancel my lcd order and stick with this old crt??
 
You must log in or register to reply here.
Back
Top