2405fpw LCD vs. CRT - 40ms signal delay / latency for LCD users ?!?

gustep12

n00b
Joined
Nov 21, 2004
Messages
31
Hi all,

Last year, I posted a review / comparison of two 24" LCD screens, the Samsung 243T and the Dell 2405fpw. One of the my main conclusions was that the 2405fpw has indeed much improved color and better pixel response time (and hence less motion blur / ghosting), but that it apparently suffers from a signal delay compared to the 243T. You can read it here:

Dell 2405fpw VS. Samsung 243T - Dell has Lag!!

This time, I compared my 2405fpw against a Sony Multiscan E210 Trinitron 17" CRT, taking a photo (1/60sec) that shows both screens at the same instant. The screens are cloned displays on my Radeon X800, showing a timer (A.F.8 Digital Clock, downloaded from http://www.fauland.com/af8.htm). Of the three digits that you see, the first is seconds, and the last digit gives 1/100 seconds:

2405fpwdelay3gk.jpg

Specifications: 2405fpw as primary display using DVI, E210 as cloned secondary display using VGA, 1280x768 pixels, 60Hz refresh rate.


As you can see, the 2405fpw trails behind the CRT by up to 40ms. In most other photographs, the difference was 30ms, but due to the timer resolution and screen refresh rate the exact time delay cannot be defined more precisely.

Unfortunately, I don't have the 243T anymore to perform this experiment, but my guess would be that the 243T is probably 15-20ms slower than a CRT.

Again, this shows that while a CRT works in "real time", LCDs can (and do) have signal latency issues, because they do some fancy digital signal processing. The situation is similar in nature to the fact that analog music amplifiers work in real time, whereas dolby digital receivers can have a small - but occasionally annoying - delay due to the digital signal processing of the audio data.

I would encourage others to test their LCDs in a similar way against a CRT and post figures. Furthermore, I would suggest we ask that LCD manufacturers specify the signal latency of their LCD's electronics, and not just the pixel speed once the signal actually arrives at the pixel.

For online gaming especially, this sucks. You'll have a competitive disadvantage that is at least equivalent to having 40ms added to your ping, or worse. I'm not entirely sure yet if 40ms extra ping on a CRT are exactly the same as playing on the 2405fpw, or if there are other effects to consider, like the fact that not even local events are displayed in real-time on your screen.

I play UT2k4 on occasion, and for fun I tried playing it on that tiny 17" CRT. My eyes started to hurt, but my performance and target accuracy seemed better indeed.

Comments welcome. :eek:
 
I"ll make this plain and simple......

For me. IceWind. I play alot of FPS and all of it online.
Had 2405 for 4 months.
Never experienced nor seen any difference in "input" lag from my Viesonic CRT to L90D+ to 2405 and shortly a VP930b in a 9 month change over i've had of 4 change of monitors and I NEVER had this experience of "input" lag.

And for the HUGE market of LCD" sales which go to non gaming work enviroments, NO one is gonna notice a 20-30ms input delay. And yet I find it ironic you bash the 2405 because of this yet continue to use one? Interesting.....

Now if you'll excuse me, I have Iraq's to kill
 
Of course the 2405fpw is an absolute outstanding display, I wouldn't trade it for any CRT or for any other 24" LCD that I can think of at the moment. It's just sad that it LCDs aren't perfect. And yes, this lag does exist, its tricky to notice, that's why you have to measure it to really see it.

The obvious solution would be to buy a 24" widescreen CRT, too, and use a dual monitor setup. :)
 
Thier is no latency/lag problem with the display this is coming from a hardcore gamer using high-end CRT's for the past 7 years. If your running in DVI at native resolution with the dell monitor driver then there isnt a problem, for those people who ignore even one of these criteria then your on your own. Also make sure your mouse drivers are upto date.
 
One question why is your display not running in its native resolution?

(your icons. start bar etc... look prety big, vertically you should be able to fit 15 icons, in your image I counted 9)

Dont try and get out of this one, noway is that image running at 1920x1200.

Theres you problem dude.
 
One thing I have learned on these forums while trying to decide what LCD to buy over the last two months is that everyone is WAY too anal about response times and paper specs.

I had a shitty shitty MAG 17" LCD display for over 3 years. You do the math....offbrand + 3 year old LCD technology. Because I never introduced myself to things such as "omg blacklight bleeding" or "omg .238974 color points offf!!!!!" I havent noticed any problems when using it to play fast paced games.
 
IceWind said:
I"ll make this plain and simple......

For me. IceWind. I play alot of FPS and all of it online.
Had 2405 for 4 months.
Never experienced nor seen any difference in "input" lag from my Viesonic CRT to L90D+ to 2405 and shortly a VP930b in a 9 month change over i've had of 4 change of monitors and I NEVER had this experience of "input" lag.

And for the HUGE market of LCD" sales which go to non gaming work enviroments, NO one is gonna notice a 20-30ms input delay. And yet I find it ironic you bash the 2405 because of this yet continue to use one? Interesting.....

Now if you'll excuse me, I have Iraq's to kill

Hi IceWind! :)
Are you saying that you cannot reproduce the example that he showed on your 2405fpw?

Edit: 40ms is much for FPS players (can throw the game totally off) and if he is correct, it means that you must press fire 40ms BEFORE the target gets into aim. It has nothing to do with the response time. If this is true, he has done the community a big favor in discovering and documenting this and its not about bashing the 2405 even though you don't like the information.
I read his previous post a long time ago, and there they only tried to make excuses about faulty VGA port. This time he produced it with the DVI port as well and it would be interesting to know if the same thing happens on other 2405fpw's as well! He has gone to great lenght to prove his points and validate his claims and that should we be greatful for. If this is not fact, then prove him wrong!:D
 
kleox64 said:
One question why is your display not running in its native resolution?

(your icons. start bar etc... look prety big, vertically you should be able to fit 15 icons, in your image I counted 9)

Dont try and get out of this one, noway is that image running at 1920x1200.

Theres you problem dude.

I guess hes running it at the same res as the CRT (1024x768) to avoid people saying, "your running the Dell at larger res, of course its going to take longer to update!". However i can't believe the image scaling to non native res is going to introduce 40ms of lag, if it does, then that in itself is pretty wierd. Might be worth trying running the comparison at native res on the Dell and see what happens...
 
Ummagumma said:
Hi IceWind! :)
Are you saying that you cannot reproduce the example that he showed on your 2405fpw?

Edit: 40ms is much for FPS players (can throw the game totally off) and if he is correct, it means that you must press fire 40ms BEFORE the target gets into aim. It has nothing to do with the response time. If this is true, he has done the community a big favor in discovering and documenting this and its not about bashing the 2405 even though you don't like the information.

What do you mean discovering this? I posted my results (using CoolEdit) almost
almost 2 months ago.
 
Ummagumma said:
Sorry, I didn't see yours... :( But he did make his first post about this in May, so I think he was first...
http://www.hardforum.com/showthread.php?t=900484

Now we have at least 1 confirmation. It would be interesting to see how this problem effects different range of monitors (different models and brands). :)

It would turn into something of a scandal if it affected a great many displays, seeing as there being marketed towards the gamer these days ... it might explain why my aim sucks recently tho :D
 
I'm curious what else can affect the delay.

Does resolution play a part?
Does swapping connections show the same result?
Is the lag similar between nVidia and Ati?
Do cables play any part?
I've noticed a loss of performance (5%-10%) in games when running 2 monitors cloned? Can that play a factor at all?
What about games? 3dmark has a frame counter (not frames per second). Is the same amount of lag demonstated by the benchmark?

Also, to clear things up, we are talking about 2/5 of a second delay, right?
 
Da Frechman said:
Also, to clear things up, we are talking about 2/5 of a second delay, right?

No, less than 1/20th of a second which is tiny. But tell a gamer their super amazing LCD is showing stuff from 3 frames ago and they'll be lighting their torches, grabbing thier pitchforks and marching on *INSERT LCD MANUFACTURER HERE* offices :D
 
Roceh said:
No, less than 1/20th of a second which is tiny. But tell a gamer their super amazing LCD is showing stuff from 3 frames ago and they'll be lighting their torches, grabbing thier pitchforks and marching on *INSERT LCD MANUFACTURER HERE* offices :D

Hmmm, wonder why I keep my score in the top 5 standings in HL2, BF2, and UK2k4 then?

Honestly, this is like saying your new corvette is manufactured poorly as a result of clear overcoat not being "smooth" enough so it effects your top speed :rolleyes:
 
I also wonder that... Maybe you fire off prematurely? ;) lol
Such a delay can do much when I play at least...
2405fpw243tplasma6oe.jpg
 
IceWind said:
Hmmm, wonder why I keep my score in the top 5 standings in HL2, BF2, and UK2k4 then?

Honestly, this is like saying your new corvette is manufactured poorly as a result of clear overcoat not being "smooth" enough so it effects your top speed :rolleyes:

Your probably right, in practice it probably makes no difference, which explains why if this effect does exist it has hardly been noticed. Like a previous posted mentioned its something to blame instead of "OMG thiz netcode sux!", becomes "OMG thiz 40ms signal delay that causes me to visualise data from 3 frames ago sux!", hmm maybe not as catchy tho....
 
I made a better timer in Java that's better than 0.1 ms (1/10000th of a second).

My digi-cam isn't around yet, but if anybody wants to try it now here's the code:

Code:
public class ConsoleTimer {
    public static void main(String[] args) {
        long start = System.nanoTime();
        
        while(true) {
            System.out.println(System.nanoTime() - start);
            System.out.flush();
        }
    }
}

To run it use:
Code:
javac ConsoleTimer.java
java ConsoleTimer

You'll have to have the Java 1.5 SDK installed of course.



And if your gonna do the test make sure you:
Set 2405FPW to 1600x1200, non-scaling 1:1
Set CRT to 1600x1200 @ 60Hz
Set your camera to less than 1/60
 
Thanks for the info.

As per quality vs price the 2405fpw still seems to be a reasonable purchase. Unless something unexpectately pops up in the next 2 months, I should be yet another proud owner of one.
 
Well, I did my test using my timer!

timedelta2pd.png


This test shows the delay to be 17 ms...

I took a few more pictures, and the delay was about 10 to 40 ms...

What I found most suprising is the delay isn't constant...it varies...

Btw, my timer was going too fast, so I slowed it down. Here's the new code.
Code:
public class ConsoleTimer {
    public static void main(String[] args) {
        long start = System.nanoTime();
        long last = start;
        long next = start;
        
        while(true) {
            if(next - last > 100000) {
                last = next;
                System.out.println(next - start);
                System.out.flush();
            }
            next = System.nanoTime(); 
        }
    }
}
 
excellent work with this test.

I think the main insight here is that LCD manufactures may be sacraficing signal time in favor of digitally tweaking the image to make it look good in tests. This isn't a direction any gammer would want manufactures to go in. This should be brought to the attention of the big review sites and signal delay should be added to the basic tech stats of a monitor in those reviews.

ps. i have really been hoping that lcd has matured enough to be ready for prime time, this tells me that maybe the technology is still well behind
 
For me the lag was noticable without using timers etc. ,For example when I would open and close windows you would see the window respond a split second sooner on the CRT side vs. LCD,It was a very short delay but noticable if you were actualy looking at both displays at once, this was with a Sony G520P CRT & Dell 2001FP LCD.
 
1) The 2405 has to be run using DVIat its native resolution (1920 x 1200) for minimum lag/delay.

This is where your delays are coming from, also make sure your using the Dell monitor driver.
Come back and report your results only when ALL the above is true, otherwise your results are just c**p.

2) The clone feature on Nvidia cards introduces a delay anyway when two displays are connected.
 
kleox64 said:
1)
2) The clone feature on Nvidia cards introduces a delay anyway when two displays are connected.

I thought that might be the case until I conected two CRTs via clone mode, (my CRT HDTV has a DVI input) ,there was no lag noticable with both CRTs going. So in my case the lag was being caused by the LCD( 2001FP) itself. Cant argue the facts there.
 
kleox64 said:
1) The 2405 has to be run using DVIat its native resolution (1920 x 1200) for minimum lag/delay.

This is where your delays are coming from, also make sure your using the Dell monitor driver.
Come back and report your results only when ALL the above is true, otherwise your results are just c**p.

2) The clone feature on Nvidia cards introduces a delay anyway when two displays are connected.

I think kleox's first point maybe the culprit (I hope!). When not running at the native 1920x1200, the 2405 will have to do interpolation calculations to scale the res to the screen. Hopefully, that's the source of the delays.

If not...ugh. I wonder if the same behaviour would happen on the component inputs?

Navy
 
kleox64 said:
1) The 2405 has to be run using DVIat its native resolution (1920 x 1200) for minimum lag/delay.

This is where your delays are coming from, also make sure your using the Dell monitor driver.
Come back and report your results only when ALL the above is true, otherwise your results are just c**p.

2) The clone feature on Nvidia cards introduces a delay anyway when two displays are connected.


Well, for this test I had my 2405fpw set to 1600x1200, with no scaling. IE. It is using native resolution, expect with two black bars.

I don't think that would add any lag as it doesn't do any interpolation.

And I did use DVI for my LCD and DSUB for my CRT.
 
Ugh, that's pretty damning. :( Now I'm grasping at straws... What about the dell monitor driver?

Navy
 
NavyH16 said:
Ugh, that's pretty damning. :( Now I'm grasping at straws... What about the dell monitor driver?

Navy

You have a unconditional 21 day return policy through Dell. If it HONESTLY has lag to you, you can return to Dell for a full refund, no questions asked.
 
it requires all the components I mentioned.

1) DVI
2) Native resolution or 1:1 scaling.
3) Monitor driver.
4) Mouse driver.

If you have all of the above there is no lag/delay, if there is then its a faulty monitor. What do you expect after dell sells so many of these and we have just 2 guys complaining about this problem.
 
kleox64 said:
it requires all the components I mentioned.

1) DVI
2) Native resolution or 1:1 scaling.
3) Monitor driver.
4) Mouse driver.

If you have all of the above there is no lag/delay, if there is then its a faulty monitor. What do you expect after dell sells so many of these and we have just 2 guys complaining about this problem.


Ding ding ding! We have a winner, you get the cookie!

hellsyeah7qi0kw.gif
 
When yuo refer to "1:1 scaling," do you mean setting the image size in the monitor menu to 1:1, or does using a multiple of the native resolution count as well (for instance, I play BF2 @ 1680x1050).
 
kleox64 said:
it requires all the components I mentioned.

1) DVI
2) Native resolution or 1:1 scaling.
3) Monitor driver.
4) Mouse driver.

If you have all of the above there is no lag/delay, if there is then its a faulty monitor. What do you expect after dell sells so many of these and we have just 2 guys complaining about this problem.

Kleox, when using DVI there is no 1:1 scaling, the image is always stretched to fill the screen. 1:1 scaling is typically only available when using VGA, at least that is how it works on the Dell 2405fpw.

Unfortunately the only CRT that I was willing to haul all the way to my apartment and back was not able to display the 2405fpw's native resolution. But yeah, I agree that having a 1920x1200-capable CRT would be the best way to convince naysayers.

However, I don't agree with your suggestion that mouse drivers factor into this - during this "photo" test, the mouse isn't even being used, it could be unplugged or uninstalled entirely.

And the display drivers probably don't matter either, since display cloning is most likely done in hardware and at the very end of the GPU architecture, and thus not affected by display drivers. Even so, I was using the CRT as the *secondary* display, so if there was a delay from copying the primary to the secondary display, then the CRT should have been delayed - not the LCD on DVI and primary.
 
Nice stuff, thanks for the input from Kevlarman, who shows 41ms delay http://www.hardforum.com/showpost.php?p=1028282387&postcount=8, and from Pyrolistical, who reports 17-40ms delay.

I think the varying figures when using the "take-a-picture" method result from the fact that the displays are not switching the frames simultaneously. This can be illustrated by the following table, where a 60Hz refresh is assumed (i.e., one new frame every 0.016.. seconds). Of course you don't have to go in 0.01 second steps, this is just to show that depending on the exact moment, the difference between the two monitors can vary:

Seconds - CRT - LCD (assuming 60Hz and 40ms delay) - Frame difference in photo

0.00 - Frame1 - Nothing - 1 frame difference
0.01 - Frame1 - Nothing - 1 frame difference
0.02 - Frame2 - Nothing - 2 frames difference
0.03 - Frame2 - Nothing - 2 frames difference
0.04 - Frame3 - Frame1 - 2 frames difference
0.05 - Frame4 - Frame1 - 3 frames difference
0.06 - Frame4 - Frame2 - 2 frames difference
0.07 - Frame5 - Frame2 - 3 frames difference
0.08 - Frame5 - Frame3 - 2 frames difference
0.09 - Frame6 - Frame4 - 2 frames difference
0.1 - Frame7 - Frame4 - 3 frames difference

As you see, the max difference observed in this case is 3 frames, and it is observable less often than a 2 frame difference. You would therefore measure at least two different delays in this scenario, namely 2*1/60 sec = 33ms and 3*1/60 sec = 50ms, depending on which moment you happen to photograph.

Also, I don't think mouse drivers or monitor drivers factor into this type of measurement, since we aren't even yet trying to measure what additional lag the mouse produces. This is strictly measuring what happens after (presumably cloned and hence simultaneous) signals leave your video card's DVI / VGA connectors.

The most advanced type of measurement I could think of to test out gaming rigs would work as follows, but unfortunately I haven't been able to pull it off:

1.) Tape your optical mouse to your monitor
2.) Write some software that displays and then suddenly moves a pattern on the screen
3.) The mouse, being fooled by the pattern, will then move the cursor
4.) So in total, the program moves the pattern on the screen and measure how much later the mouse reports back with a Mouse.Move signal.

5.) Voila: This would be the total input delay of your rig, and it would account for ALL delays, mouse driver lag, display driver lag, monitor lag, and whatever else.

The catch is that I couldn't dupe any optical mouse into registering a moving pattern on my monitor yet. I guess the DPI mismatch is just too great, something like 72dpi for the monitor and 3000dpi for the optical mouse... But who knows, with a special lens, or a dedicated optical mouse this should work. And THEN we'll be talking business... ;)

 
A question: how much does this delay affect your FPS gaming?

For me, none at all. I'm still near the top of the list in BF2 and DOD:S.

YMMV though.
 
It's hard to say how it affects my gaming... I play UT2k4 which is really fast-paced, and I wouldn't play anything above 100 ping online to begin with. Also, the few times I asked the top scoring player after a game what he used, it's always been a CRT.

I simply haven't had a good CRT set up to make an in-depth extended comparison, all I had was this 17" Sony for an afternoon. However, I scored really well when I used this tiny CRT, and my impression was that hitscan moves and "aiming while dodging & jumping" in particular were a lot more successful than usual. On the CRT everything felt refreshingly snappy.

Oh well...

UT2k4 rules, by the way. I get ~60 fps (never lower than 40) at 1920x1200, with 4xAA 8xAF and every in-game eyecandy maxed out on my X800XT / 1 gig ram / 2000Mhz (3000+ rating) Athlon XP . That's a pretty darn efficient graphics engine. Also, my favorite is Vehicle Capture the Flag with Rosebum RPG mod. Check it out sometime:

ut2004://72.36.155.148:7777

Here's a 1920x1200 screenshot (~1MB jpeg):

http://img297.imageshack.us/img297/1518/shot001670ls.jpg
 
I don't have any technical input on the argument, but I do have my opinion on what I've felt from the moment I started using my 2405. Before going to the 2405 I had been using CRT's at 100hz+ for the last 3 years, so I'm probably a little more sensitive to any sort of lag. Anyways, ever since I got the 2405 I've felt this exact lag that the OP and others have mentioned. I now find online FPS's where quick reflexes are needed to be completely unbarable. In fact I've switched to WoW as it's the only game where this lag doesn't affect my game play. I DO, run these games at less than the native resolution, however this lag is still very apparent when using the monitor in windows @ 1920 res. I'm using the dell monitor drivers, I have updated drivers for my mx500, and my display drivers are updated as well.

I'm simply posting to back up what the others have documented as they seem to be getting flack from the disbelievers. At this point it seems uncertain whether all users of the monitor are affected, but without a doubt there are many of us out there that are experiencing the issue.

Lets keep this thread alive and try to find the root of the problem, or even better, a solution.
 
is this lag issue prevalent in other LCD's? Is the lag worse with certain panel technology (PVA/IPS/TN)? Is this just a dell 2405 issue? Does the lag exist with other brands and if so is the effect more extreme with the 2405?

I thought i was ready to switch to lcd. Hope this is a 2405 thing.
 
1.those that are disputing that there is minimal impact on FPS gaming are casual gamers.
a CounterStrike player myself ~40ms is freaking huge - do u guys know wat a flash snipe is?

basically all elite CS players zoom and shoot in almost one action - called a flash coz the screen flickers between 'scope' and shot.

2. when i first got my 2405 i had input lag - fixed with new ATI drivers and fiddled with drivers.

3. native res exhibited the worst input lag. i currently play CS @ 13xx*768(?)

4. lets face it DELL prob went cheap on a few things to get this awesome display out so competitively priced.

5. new tech just aroung the corner...

still ove this unit.
 
how about we just all face that LCD suck for things like games? if you dont think so, your just used to it so its ok i guess.
 
Back
Top