or Connect
AVS › AVS Forum › Display Devices › Plasma Flat Panel Displays › The Official Plasma Input Lag Thread.
New Posts  All Forums:Forum Nav:

The Official Plasma Input Lag Thread.

post #1 of 785
Thread Starter 
So I'm not sure how many people care about this, but it was something I really wanted to see when I was researching PDP's. I looked for hard numbers for quite a while, but for the most part, couldn't really find any and just had to go by anecdotal reports or one line blurbs from reviews (usually for foreign equivalent models that may or may not be the same). The goal is for this thread to be a one-stop compilation of several displays, so please post results or PM me links to others tests so I can add them. You can read more about the testing procedure here and even try it yourself . Jvarisco has put together a guide to understanding the general problem of "input lag". Frito also runs an LCD equivalent thread in the LCD forum that goes into a lot of detail on the subject.

*Testing Procedure Instructions*

Physical Connections
1) Connect a laptop to your TV and cycle to the video input you used. HDMI is preferred, though VGA is better than nothing.
2) Do not run your tests through an Audio Receiver (AVR), use a direct connection between the TV and PC. This is very important. Some AVR's can and will introduce lag, even on 1080p "Through" connections. OSD overlays are the biggest offenders.

On your PC
1) Turn off Windows Aero (see how here). This is very important, Aero and V-sync will skew your results.
2) If you have a dedicated video card (Nvidia/ATI), it may be helpful to disable (force off) Vertical Sync. I believe disabling Aero usually takes care of this, but if you have doubts, this may be worthwhile.
3) Right click your desktop and choose "Screen Resolution". (*Note* these are Windows 7 instructions, Vista/XP may differ slightly)
4) Click the "Multiple Displays" drop down and select the "Duplicate these displays" option.
5) Your desktop should now be mirrored/cloned to your TV and laptop display simultaneously.
6) Point your browser to this link.

On your TV
1) Use the display's "Game Mode" (check the manual for instructions)
2) Turn OFF all special processing features in the TV's menu (e.g. CATS, Motion Smoother, Dynamic Contrast, Frame Interpolation, Cinesmooth, Blur Reduction, Local Contrast Enhancer, etc...)

On your camera
1) You will want to use a mode or option with a fast shutter speed so the timer numbers show up clearly.
2) On lower quality cameras, using the flash often helps with this.
3) Start the timer in your browser and take a picture so the numbers are visible on both displays simultaneously (see below for examples).


2013 Models


Panasonic S60, 34.1ms (by David Katzmaier via Lag Tester)


Panasonic ST60, 73.6ms (by David Katzmaier via Lag Tester)


Panasonic VT60, 47.9ms (by David Katmaier via Lag Tester)


Panasonic ZT60, 46.2ms (by David Katmaier via Lag Tester)


Samsung F5500, 37-38ms (by Display Lag Database via Lag Tester)


Samsung F5300, 40-53ms (by Display Lag Database via Lag Tester)


Samsung F8500, 56-68ms (by Display Lag Database via Lag Tester)


Samsung F4500, 38-55ms (by Display Lag Database via Lag Tester)




Panasonic ST60E (European), ~50-100ms via cloned timer, 75ms via Lag Testing Device (by Norvegia)



2012 Models



Panasonic U54, ~20ms* (by myself) *vs. unknown laptop LCD, additional results in post.






2011 Models


Panasonic GT30: ~24ms* (by Chudsmith) *Average out of 43 shots. Point of comparison is unknown LCD.




Panasonic ST30: ~15-32ms* (By thepoohcontinum) *Primary result was 32ms. See his post for measurements of frame interpolation and other processing features, 62-125ms (also tested by Gandu)




Panasonic S30: ~16-32ms* (By DrFrank) *Point of comparison is unknown laptop LCD. Windows Aero on, results potentially slightly skewed.




Samsung D7000: ~34ms* (by Varkeast) *Game Mode (also tested by notext)




Samsung D550: ~20ms PC Mode, ~32-50ms Game Mode* (by badmeng and DrFrank) *Point of comparison is unknown laptop LCD. Windows Aero on in DrFrank's testing, results potentially skewed.






LG PZ550: ~100-117ms* (By DrFrank) *Point of comparison is unknown laptop LCD. Windows Aero on, results potentially slightly skewed.




2010 Models


Panasonic VT25: ~15ms* (by Jrod0802) *Note this is a VGA test, HDMI could yield different results. Point of comparison is also an unknown laptop LCD panel.




Panasonic VT20: ~32ms* (by mechashiva2k) *Note test source is unknown laptop panel.




Panasonic GT25: ~15ms* (by toolman02) *Note test source is unknown laptop panel. (Also tested by Zodduska)




Panasonic G25: 16-32ms* (by NOAMattd) *Primary result seemed to be 32ms.




Panasonic S20E: 18-30ms* (by AABBCCAA) *Note this is the nearest European equivalent of the North American S2, there could be differences between territories.




Panasonic U2: 28-43ms* (by Drapers) *Shutter speed makes it difficult to call. Also note test source is an unknown laptop LCD panel, his supplementary tests suggest adding at least a full frame (16ms) to the values.




Panasonic C2: 30-50ms* (by Nerv_001) *Shutter speed seems too slow to call it definitively. Also note test source is an unknown LCD panel so actual lag is likely higher.




LG PX950N: ~50ms* (by Graco) *Note this does not include lag from the LCD panel tested against. It is also the Scandinavian model, could be territorial differences.

Game Mode



LG PK550: ~86ms* (by shaddix) *VGA, presumably a "PC Mode", measured ~46ms. HDMI labeled PC does NOT seem to reduce lag.

Game Mode



2009 Models


Panasonic G15: ~20ms (by Mr. Deap*) *480i measures included in his post.




Panasonic G10: ~15ms* (by myself and re-tested) *Note my laptop seems to lag 4-7ms based on comparisons to Deap's CRT findings, so it should actually be ~20ms as well.




Panasonic S1: ~30ms* (by myself) *This could be a difference b/t the S1 and G10 in general, or the sizes classes (e.g. 50" Vs. 65"). See my comments in link.




Panasonic X1: ~35ms (by Frito)




Samsung B650 PDP (by cleh19)

Standard Mode: 80-110ms


Game Mode: 30-50ms



2008 Models


Pioneer Kuro 5020: ~45ms* (by Video313) *Mode is unknown, but this seems to line up with other comments I have seen with Game Mode and Game Pref on.




Samsung A650 PDP: 60-80ms (by tqn)

Standard Mode



2007 Models


Pioneer Kuro 5010: ~60ms* (by tqn) *Mode is unknown




Pioneer Kuro 5080: ~20ms* (by Mathesar) *PC mode tested. Possible normal or game modes could yield different results.

Edited by Orta - 6/3/13 at 7:44pm
post #2 of 785
Thread Starter 
Reserved
post #3 of 785
This is a great idea for a thread. If nothing else, we can centralize all of the lag data into one place. Ideally we could come up with some standard tests, although I realize that may be impossible due to the need of having the proper test equipment (namely a CRT). I think there's typically 2 types of lag people are worried about:

- 480i sources - EG older game consoles.
- native sources - sources that match the native resolution of the plasma, EG modern HD game consoles

To test with 480i we can use a CRT as the reference point. I'm not sure how we can properly test lag at the native resolution other than using something like Rock Band 2 with the camera built into the guitar.
post #4 of 785
1+ on the thread, although did you confirm that your laptop has zero lag?
post #5 of 785
Great post Orta! Glad you started this (esp. because this is my primary gaming display as well). It would seem your test between standard and game display profiles on the G10 would support the theory that these are only menu setting presets. Odd that the IFC or smoothing in Standard would not create any measurable lag though

Not that you have time to take requests, but if you do perform more lag tests, could you try the game test with CATS on? I've always been curious about the effect of dynamic contrast based on ambient lighting. Contrast is the one setting I can never get dialed in to my satisfaction...would love to set it at 50 and just leave CATS on if the effect on input lag was negligible.
post #6 of 785
Quote:
Originally Posted by slyderulz View Post

Great post Orta! Glad you started this (esp. because this is my primary gaming display as well). It would seem your test between standard and game display profiles on the G10 would support the theory that these are only menu setting presets. Odd that the IFC or smoothing in Standard would not create any measurable lag though


I was shocked that the frame creation on the g10 didn't cause any extra lag as well. It seems that Panasonic is the only TV maker that can do it right.

I've confirmed that my 54g10 also only has 1 frame of lag on THX mode (my prefered mode for everything) so these Panasonic sets continue the tradition of Panasonic's being almost lag free.
post #7 of 785
Has anyone tested the B650 or any of the 2009 Samsungs?
post #8 of 785
A long overdue thread to be sure! Can't wait to see more results from the current 2009 plasma crop. Orta, you may want to put a link to the test procedure in the first post so others who may not know how to do it can contribute.

Here is one video I found of someone testing a G10 with frame creation on, outputting from a laptop to CRT and the G10 simultaneously (as per the test on http://hdtvlag.googlepages.com/ourtest)

http://vimeo.com/4913119

I think videos are better over pictures personally, they're less effort to take and upload and all you have to do is pause in a bunch of spots to get as many "samples" as you like.
post #9 of 785
This is a much needed thread. Input lag is rarely talked about in reviews.

However, the comparisons need to be made to CRT. If the G10/V10 have about 1 frame difference to an LCD, then the total lag on the input is around 2 frames.
post #10 of 785
pete2s has it right. Comparing against CRT shows you the lag for 480i material, important for last-gen game systems and SD material.
post #11 of 785
Thread Starter 
Quote:
Originally Posted by NOAMattD View Post

pete2s has it right. Comparing against CRT shows you the lag for 480i material, important for last-gen game systems and SD material.

Good news, Mr. Deap has done just this. I included the link in my second post.
post #12 of 785
ok i'm confused (sorry) is the Panasonic better than the samsung LCD then?

Was the 650 not widely reported as having awful lag?
post #13 of 785
Great idea for a thread. My only concern is that there is probably some input lag on your computer monitor too. Especially if you have one of the newer ones with dynamic contrast or other image processing. So that might skew the results.
post #14 of 785
Quote:
Originally Posted by astonbilla View Post

ok i'm confused (sorry) is the Panasonic better than the samsung LCD then?

Was the 650 not widely reported as having awful lag?

The A650 has bad lag in the standard modes (~60ms), but had fairly low lag in PC or Game mode (15-30ms). The G10 has lower input lag than the A650 does in PC/Game mode.
post #15 of 785
Thread Starter 
I updated post #2 with A650 Plasma measurements, and conjecture for the B's.
post #16 of 785
Does anyone know how the Pioneer 600M or 500M do in this department?
post #17 of 785
Thread Starter 
Quote:
Originally Posted by Megasabin View Post

Does anyone know how the Pioneer 600M or 500M do in this department?

I would imagine they're going to be the same as the XX20's, which seem to be ~45ms avg. with Game Pref on.
post #18 of 785
Guys, it's just baffles my mind why these review websites fail to include how any HDTV handles HD gaming(both 360 and PS3..heck even Wii and the last generation, which were rendered in 480i/p), especially when the gaming industry itself is a huge reason why HDTV's have been sellling like hotcakes since the Xbox360 arrived in late 2005'.

These sites(ie..Cnet,HDGuru and others) need to do tests for input lag, phosper trails, and moving line resolution.

It's comical how Cnet says the they do not notice the difference of a 300-400 measured (1080p) HDTV compared to a full 1080p measured model.

THANK YOU, to the guys who did the measurments of the Panny G10 and Samsung A650(plasma) because while i was on the fence about the color accuracy of my Panasonic S1(hopefully for a return for a G10 next week), i'd easily take less then perfect colors, compared to worse lag.

Ironically the Samsung B550 claims it has a .001 ms response time in it's specs, but it's basically the same model as the A650, add to that it measured only 800-900 lines of resolution, and it's IR prevention is not as good as the Panasonic line(S1/G10).

Be it sticking with the S1(if i can push the green level a bit lower in the user controls) or the G10, i'm sticking with the Panasonic Plasma as my gaming HDTV.

One more thing, i went from a Sony 34" XBR960(CRT), and i noticed no visible lag on both a 42" U1 or the 50" S1 models.

And i notice lag on every 240hz LCD during just a preseason live broadcast.

Hopefully these review sites will eventually figure out how many gamers are buying HDTVs, and the're main reason, is for gaming.

Phil
post #19 of 785
I've seen a lot of info on this thread related to the G10, but is it safe to assume that the same can be said for the V10?
post #20 of 785
Quote:
Originally Posted by JBgobears View Post

I've seen a lot of info on this thread related to the G10, but is it safe to assume that the same can be said for the V10?

I hope so as I'm getting mine delivered this Tuesday. Depending on how you look at it, I'm upgrading from the XBR9 to the V10. Love the XBR9 but the avid gamer in me wants the best.

Correct me if I'm wrong but I believe the G10 and the V10 use the same panel so in theory, they should perform the same when it comes to input lag.
post #21 of 785
Here's a post from a user who measured 0ms on his V10 - a hopeful sign.
post #22 of 785
The best test for input lag is the counter. Games themselves have lag and some more so than others. From my experience, Guitar Hero has less lag than Rock Band.

But to answer the question above, the V10 most likely has similar lag to the G10.
post #23 of 785
I would expect the V10 and G10 to have similar lag as well
post #24 of 785
I don't want to sound like a dick, but I think you guys should come up with a better test.

This method is good for comparing the delay between two screens, but it's not good for an actual account of delay.

For true delay, you'd have to have a high-speed video recording measuring say, a scrolling line going across the screen, and making a noise when it hits a certain point. This is what would actually matter when playing a video game. Although a first-person-shooter game would be noticeable, a "musical-rhythm" game would probably be even more noticeable. An example would be Dance-Dance-Revolution, or any of those guitar/band games.

Not to mention that although the delay might be one or two frames by your test, if it crosses the line from imperceptible to perceptible, that's the only thing that counts.
post #25 of 785
The counter compares a display that has no input lag to a display with lag, with a counter that shows a difference in the ms... how can you do a better test? As I said, using gaming consoles to test using a musical rhythm game is not a good test because gaming consoles (especially those with wireless controllers) may have lag within the game itself. Even if that wasn't the case, a musical rhythm game would only confirm findings from the counter test, but it would be less accurate because you're going by ear/eye vs a snapshot showing precisely how much delay there is. Case in point, tests done with the A650 LCD show that game mode has 25-35ms or ~60ms without game mode, and that's almost exactly the results I get using the RB2 manual calibrator.
post #26 of 785
Quote:
Originally Posted by Meigus View Post

I don't want to sound like a dick, but I think you guys should come up with a better test.

This method is good for comparing the delay between two screens, but it's not good for an actual account of delay.

For true delay, you'd have to have a high-speed video recording measuring say, a scrolling line going across the screen, and making a noise when it hits a certain point. This is what would actually matter when playing a video game. Although a first-person-shooter game would be noticeable, a "musical-rhythm" game would probably be even more noticeable. An example would be Dance-Dance-Revolution, or any of those guitar/band games.

Not to mention that although the delay might be one or two frames by your test, if it crosses the line from imperceptible to perceptible, that's the only thing that counts.

It's fine when compared to a CRT which is an analog device and not capable of digitally holding and delaying the signal the way digital processors do. I agree that comparing it to a laptop is not such a good idea as you have no idea how to interpret the results.
post #27 of 785
Thread Starter 
Quote:
Originally Posted by Meigus View Post

I don't want to sound like a dick, but I think you guys should come up with a better test.

This method is good for comparing the delay between two screens, but it's not good for an actual account of delay.

For true delay, you'd have to have a high-speed video recording measuring say, a scrolling line going across the screen, and making a noise when it hits a certain point. This is what would actually matter when playing a video game. Although a first-person-shooter game would be noticeable, a "musical-rhythm" game would probably be even more noticeable. An example would be Dance-Dance-Revolution, or any of those guitar/band games.

Not to mention that although the delay might be one or two frames by your test, if it crosses the line from imperceptible to perceptible, that's the only thing that counts.

It sounds like he's wanting to include a test for audio lag as well as account for rendering lag, controller lag, etc... (absolute delay). It's a great idea, but probably beyond the scope of this thread. It also wouldn't be very practical as anything beyond the display is really out of your control. There is actually quite a bit of inherent lag introduced by (mostly) rendering complexities and bufferings. Here is a great write up on it that tests a lot of games. Really makes it more important to have the fastest display possible because the additive effect is what kills you. As you can see, playing a particularly unresponsive game like Killzone2 or GTA4 on a Samsung can easily reach and exceed noticeable levels (200ms+). It also explains away all the arguments about how meaningless display input lag is because average human reaction time is 150ms (made up numbers), it takes 200ms to blink, or whatever--I've seen those brought up a lot.
post #28 of 785
Quote:
Originally Posted by Orta View Post

It sounds like he's wanting to include a test for audio lag as well as account for rendering lag, controller lag, etc... (absolute delay).

No. What I said was that you have it timed to an "external" source, cause unless you have a device that is hyper-responsive, you can't precisely measure absolute input lag. A non-digital audio output should, in theory, be "a lot" faster than the video output. Not to mention because it modulates on a higher frequency, you can have more precise and accurate timing (almost all operating systems use a minimum 44100 hz). This would be more effective than the "guess-the-frames" mode you guys are using now.


Quote:
Originally Posted by sethk View Post

It's fine when compared to a CRT which is an analog device and not capable of digitally holding and delaying the signal the way digital processors do.

Even if the analog was 100% analog (no menus, fixed resolution), it will still take time to get through the circuitry, which by nature could be slower than a digital screen. By nature I mean it's not designed to necessarily BE ultra fast at all, it's designed to show the picture "soon enough". This would probably be even worse if the CRT was NOT designed for computers/gaming, the same reason some people're raising a fuss over high-def tv lag!


Third, all these tests you guys run HAVE to be from the same cable, otherwise they're almost useless. Although a fast graphic card could easily display cloned images, unless it's processing and displaying them at precisely the same time, the "dominant" port will be more responsive than the "alternate" port. I won't say this always happens, but it could obviously happen if the graphic card (or software) was designed poorly.

Using the same cable, like someone suggested with the component out and a splitter, would make the tests far more trustworthy.
post #29 of 785
I did a little test at a store to see what the reaction times are through the coax feed in the store. Here is the video that I posted on youtube. http://www.youtube.com/watch?v=eDBtLQKQHzc

I was slightly surprised by the results, but the slight delay in plasma has not made me want lcd at all.
post #30 of 785
Quote:
Originally Posted by rahzel View Post

The best test for input lag is the counter. Games themselves have lag and some more so than others. From my experience, Guitar Hero has less lag than Rock Band.

But to answer the question above, the V10 most likely has similar lag to the G10.

What are you talking about? Games do not have input lag. I've been playing Rock band for a couple years on CRT TVs with the audio directly connected to speakers and I can guarantee you there is no lag whatsoever.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Plasma Flat Panel Displays
AVS › AVS Forum › Display Devices › Plasma Flat Panel Displays › The Official Plasma Input Lag Thread.