Ways to reduce input lag on a Viera Panasonic plasma?

[Roboplodicus]

So I'm the new owner of a sweet 42" Panasonic Viera Plasma TV, and using the Rockband 2 Guitar with the light sensor and mic I Lag tested this screen which is a Th-42fz80u and it came up with 30 ms of input delay, this is a 2008 screen btw and it does have a "game" picture mode but I'm pretty sure that its just a image color balance preset and that it doesn't actually disable any signal processing. There really wasn't much to disable in the standard menu I'm curious if anybody knows if some more image processing can be disabled in the service menu. Either way 30ms isn't *TOO* terrible, or atleast it hasn't started bugging me yet and the screen looks absolutely gorgeous and the motion clarity is fantastic, even compared to my 1ms GTG 144hz G-Sync Benq monitor, although that monitor has only 10ms of latency so it is a trade off like a lot of things in life.

[opt2not]

I have the Panasonic VIERA TC-P50ST60, and unfortunately there is no way of reducing the input lag. This TV is not fit for games -- it's actually one of the worst ones.
The lag makes any type of action game uncomfortable to play on because the lag is just too severe. The only games I play on it are RPG's, or games that doesn't require twitch inputs. You are right, the Game Mode does nothing for input lag, and there aren't any settings you can adjust to improve it.

I'm glad that mine is starting to show signs of dying, because I'm looking forward to replacing it with a TV that is better suited for gaming (gonna have to do my research on that).

[Roboplodicus]

I have the Panasonic VIERA TC-P50ST60, and unfortunately there is no way of reducing the input lag. This TV is not fit for games -- it's actually one of the worst ones.
The lag makes any type of action game uncomfortable to play on because the lag is just too severe. The only games I play on it are RPG's, or games that doesn't require twitch inputs. You are right, the Game Mode does nothing for input lag, and there aren't any settings you can adjust to improve it.

I'm glad that mine is starting to show signs of dying, because I'm looking forward to replacing it with a TV that is better suited for gaming (gonna have to do my research on that).

I was looking at the database over at DisplayLag.com and all of the plasmas in the database had a minimum of 30ms of lag, do you know of any Plasmas that have a reputation for not being terribly laggy?

[opt2not]

I was looking at the database over at DisplayLag.com and all of the plasmas in the database had a minimum of 30ms of lag, do you know of any Plasmas that have a reputation for not being terribly laggy?

I do not. Someone who's more knowledgeable with this can correct me if I'm wrong, but I believe Plasma's are just not good for gaming in general. It's the nature of the technology. Great for watching movies/tv, but for gaming you're better off getting an OLED these days.

[ShadowofBob]

I was looking at the database over at DisplayLag.com and all of the plasmas in the database had a minimum of 30ms of lag, do you know of any Plasmas that have a reputation for not being terribly laggy?

I recall seeing that the ZT60(24-30ms) or the S60(~29ms) has some of the lowest lag for the final year of Panasonic Plasma. I have a 55ZT60 and don't mind gaming on all genres on it, but I'm not super lag sensitive. I've turned off as much processing as possible and have used calibration settings online to get a very nice baseline image that has the last lag possible (especially in game mode, but I don't find it cuts it down significantly). The bigger thing that annoys me is the temporary image retention. The best LG OLEDs are at ~21ms I believe so I don't think they are that significantly faster than the best Plasmas. If you are particularly sensitive to lag at that level I would just stick to newer LCD screens or computer CRTs.

[FinalBaton]

Interesting. I see the fz80u all the time in the classified here. Thanks for sharing your thoughts on it.

Regarding lag : the ST50 has "16ms" according to HDTVtest. I don't know if it's camera test, or if it's a Leo Bodnar one(and where on the screen was it checked)

[ChuChu Flamingo]

Someone correct me if Im wrong, I don't think any plasma will have less than 16 ms of lag simply due to how it processes the picture ( 600hz).The plus side is that it refreshes the pictures all at once so top, middle, and bottom will all be the same input lag.

The Plasma I currently have is a 50" G10 which was reported to have 21 ms of lag with the camera method. It is probably a little higher than that but most likely under two frames. The other monitor I have is a VH232H evo monitor at 9-10 ms. When I switch between them there is definitely a difference for games I know. Super easy to tell on fighting games and panel de pon.

[citrus3000psi]

Someone correct me if Im wrong, I don't think any plasma will have less than 16 ms of lag simply due to how it processes the picture ( 600hz).The plus side is that it refreshes the pictures all at once so top, middle, and bottom will all be the same input lag.

The Pioneer plasma PDPV402, I recorded 2ms with the ossc. It has since died, may it rest in piece.

[Roboplodicus]

I've been going back an forth between my 10ms monitor and the 30ms plasma and I sort of feel like atleast for FPS games where you have the entire screen moving constantly as opposed to shmups where you have alot more of the screen static often, the motion clarity, contrast and color accuracy of the plasma sort of makes up for the input delay in how much I enjoy playing these games although I'm just playing one player campaigns right now, multiplayer could be a bit different. I can track where things are going to be even when the field of view is swinging around pretty quickly on the plasma, and my gaming monitor actually has a 1ms grey to grey theoretically(and it definitely has a lower grey to grey because its much less blurry than my other monitor that is billed as having 5ms grey to grey) but when I start doing quick 90 degree or 180 degree turns the screen is pretty much all blur.

The Pioneer plasma PDPV402, I recorded 2ms with the ossc. It has since died, may it rest in piece.

I just looked up that screen and its pretty cool I'd only seen 16:9 480p plasmas before, that was a professional monitor not a consumer one correct? Do you have any idea if there are any 1080p/i plasmas with ultra low latency like the PDPV402?

[citrus3000psi]

I just looked up that screen and its pretty cool I'd only seen 16:9 480p plasmas before, that was a professional monitor not a consumer one correct? Do you have any idea if there are any 1080p/i plasmas with ultra low latency like the PDPV402?

Yes it was considered a professional monitor. I was just pointing out the technology is capable. I don't know of any ULL 16:9 plasma like the V402, but maybe there is one out there.

[Roboplodicus]

I just looked up that screen and its pretty cool I'd only seen 16:9 480p plasmas before, that was a professional monitor not a consumer one correct? Do you have any idea if there are any 1080p/i plasmas with ultra low latency like the PDPV402?

Yes it was considered a professional monitor. I was just pointing out the technology is capable. I don't know of any ULL 16:9 plasma like the V402, but maybe there is one out there.

I'm curious sort of if professional monitors are more likely to have low latency because in a professional setting you would want the screen to display exactly what it was receiving rather than putting each frame through a bunch of filters?

[citrus3000psi]

I'm curious sort of if professional monitors are more likely to have low latency because in a professional setting you would want the screen to display exactly what it was receiving rather than putting each frame through a bunch of filters?

I want to say there is a thread somewhere on this site where a user picked up a pro plasma. It had RGB inputs and accepted 240p. I don't remember if lag was ever tested. But maybe try reaching out to that person

[Galgomite]

First thing I want to say is that if your tv is a legit ~30ms, you have a good tv— and the amount of lag you could even theoretically eliminate is awfully low. If I were you I would experiment with your sources— different resolutions (my panny is lightning quick at 1080/720 but not 480), wired controllers, and of course scalers.

[maxtherabbit]

First thing I want to say is that if your tv is a legit ~30ms, you have a good tv— and the amount of lag you could even theoretically eliminate is awfully low. If I were you I would experiment with your sources— different resolutions (my panny is lightning quick at 1080/720 but not 480), wired controllers, and of course scalers.

my panasonic is markedly faster with digital inputs vs. analog

[tacoguy64]

I'm not sure how much lag there is on my vt60 but it doesn't feel that bad. I've been pairing it with both the FM and the OSSC and results have been great on both. Maybe I need to test on games that require even more precision? I know some of the pioneer elite pro models like the pdp151 have great input lag on them too. Heck my g10 panny before this one felt good too and I did play guitar hero on that one. And my cousin noticed better COD performance playing on that tv.

But yeah I have heard some plasmas having a bad time with lag.

[maxtherabbit]

worst case on my tc-p50s2 was running 240p directly into the component inputs

it runs it through the deinterlacer and ends up with 4-5 frames

[Roboplodicus]

So I just tested my asus mg24uq monitor with the rockband 2 lag test and I received a value of ~1ms but according to displaylag.com I should have received a value of ~10ms so I think this method might be reading about 10ms less lag than the display actually has. That would mean this Viera actually has about 40ms of lag, which is then 2 frames more than I'm used to on my LCD with 2/3 of a frame of lame. I feel it much more with a mouse than I do with a controller though, so I'm sort of curious how much input lag my mouse/pc/the-fps-games-I'm-playing have. I'm sort of curious what result I'd get from a fancier lag tester.

[orange808]

So I just tested my asus mg24uq monitor with the rockband 2 lag test and I received a value of ~1ms but according to displaylag.com I should have received a value of ~10ms so I think this method might be reading about 10ms less lag than the display actually has. That would mean this Viera actually has about 40ms of lag, which is then 2 frames more than I'm used to on my LCD with 2/3 of a frame of lame. I feel it much more with a mouse than I do with a controller though, so I'm sort of curious how much input lag my mouse/pc/the-fps-games-I'm-playing have. I'm sort of curious what result I'd get from a fancier lag tester.

I apologise in advance for opening this can of worms, but what I'm saying is accurate. So, I apologise for any possible confusion.

-----

I don't believe the Rock Band lag tester is a good test and I certainly would never use it for anything. With that said, there may be an explanation:

I believe your actual display latency on that Asus mg24uq monitor is what the website says; however, *in comparison to a CRT* the latency would be around one and a half milliseconds (~1.5ms)--with some slight variation depending on the vertical position of the screen we are using as a measurement sample.

The *true* display latency *is* what the lag test from the website says, but the "latency versus a CRT" is going to be different. It may be that the Rock Band tester is giving you an estimation of the latency you are getting versus a CRT.

The Leo Badnar test display latency test numbers from a CRT (with no additional processing lag) will be:
Top = Zero (0ms)
Middle = 8 1/3 ms
Bottom = 16 2/3 ms

Also, your top number on that monitor is ~1.7ms, so that's how much time the monitor waits before it starts "drawing" the frame. So, there's another way to look at it.

[Roboplodicus]

I apologise in advance for opening this can of worms, but what I'm saying is accurate. So, I apologise for any possible confusion.

-----

I don't believe the Rock Band lag tester is a good test and I certainly would never use it for anything. With that said, there may be an explanation:

I believe your actual display latency on that Asus mg24uq monitor is what the website says; however, *in comparison to a CRT* the latency would be around one and a half milliseconds (~1.5ms)--with some slight variation depending on the vertical position of the screen we are using as a measurement sample.

The *true* display latency *is* what the lag test from the website says, but the "latency versus a CRT" is going to be different. It may be that the Rock Band tester is giving you an estimation of the latency you are getting versus a CRT.

The Leo Badnar test display latency test numbers from a CRT (with no additional processing lag) will be:
Top = Zero (0ms)
Middle = 8 1/3 ms
Bottom = 16 2/3 ms

Also, your top number on that monitor is ~1.7ms, so that's how much time the monitor waits before it starts "drawing" the frame. So, there's another way to look at it.

I'm writing this late at night guys so some of my thoughts are little disorganized, but I tried my best to relay what I think I've figured out, feel free to correct my/add your opinion to this discussion because I think its a pretty interesting one.

These are my thoughts as of right now:

Your not opening any can of worms. I understand that the way a crt draws a frame is left to right top to bottom and the whole process takes 16.7 ms for a whole frame, so the top of the frame is will measure a lower latency with a device like the leo bodnar because the sensor doesn't detect light from the whole screen it only detects it from a few rows, hence at the beginning of the 16.7ms the crt will be drawing the tops rows, at the middle of the 16.7ms(at ~8.3ms) the 3 electron guns will be aimed at and firing their beams at the middle of the screen and at a little before 16.7ms the 3 electron guns will be aimed at and firing their beams lighting up the phosphors at the very bottom of the screen. Thus, the different times for the different parts of the screen. I was hoping to get a rough estimate of relative lag on my different screens, using the lag database and a crt as reference points, although thinking about it now I realized I'm not even using the same outputs of the 360, I used component for my crt reference out of the xbox 360 and HDMI for the benq monitor and the plasma. And unlike the Leo Bodnar the guitar doesn't have any way to block out light from other parts of the screen so I'm not even sure at what point during the frame being "drawn" its picking it up, nor do I know how much latency the wireless connection it uses adds, or how much latency the rock band game engine adds(although I guess I assumed that the makers of rockband would have taken into account the latency added from the game engine and the wireless connection although I know that atleast the latency from the wireless connection of an xbox 360 controller is actually variable, and I because the xbox 360 can have other stuff running in the background while its running rockband 2 I know the game engine latency is also variable. It would be nice to know by how much these are variable by though). Though if I were not a idiot and actually used the same output for the different displays, I think I might be able to get an *approximate* and *relative* measure of comparative display lag between my different displays, although very possibly there would be a whole frame margin of error. Also up until today I was getting consistent results (within 1 ms) for each display with the rockband guitar it was only today that the LCD I tested actually gave me significantly different results between tests. Also now that I'm thinking too that the different brightnesses that my displays are outputting also probably affect when the guitar's light sensor registers the white frame its "looking" for.

I'm curious what you guys' thoughts are on my above thinking so do share with me please.

[maxtherabbit]

I need to correct myself - the CRT I was testing the plasma against (sony KV-32FV27) apparently had some lag

I got another CRT and ran more tests - the panny is 2-3 frames behind the CRT now on HDMI through the OSSC and 4-5 over analog

the internal deinterlacer is still adding 2 frames of lag, which is consistent with my previous testing
"game mode" has no appreciable effect on the results, it just makes the picture look worse lol

I can only attribute the lag on the sony set to the PiP function (even though it was off) - it's a regular SD set

[Roboplodicus]

I need to correct myself - the CRT I was testing the plasma against (sony KV-32FV27) apparently had some lag

I got another CRT and ran more tests - the panny is 2-3 frames behind the CRT now on HDMI through the OSSC and 4-5 over analog

the internal deinterlacer is still adding 2 frames of lag, which is consistent with my previous testing
"game mode" has no appreciable effect on the results, it just makes the picture look worse lol

I can only attribute the lag on the sony set to the PiP function (even though it was off) - it's a regular SD set

What inputs were you using on the 32FV27? I think the composite and possibly the s-video inputs will be put through the 3d line filter which will cause latency. `8

[gray117]

Someone correct me if Im wrong, I don't think any plasma will have less than 16 ms of lag simply due to how it processes the picture ( 600hz).The plus side is that it refreshes the pictures all at once so top, middle, and bottom will all be the same input lag.

Incorrect I'm afraid... They suffer input lag in the same way as any image processing unit would. Further more they update by flashing colours (+and different brightness) in quick succession when updating an image... rather than displaying the whole rgb frame at once permanently - Which can be noticeable especially in high contrast images to those sensitive to it - often seeing green flashes/flicker, for other people (the majority ?) this can actually have the reverse effect - particularly on moving elements - apparently smoothing image transition.

Although the plasma can update at 600hz you do not in effect get 600fps ... in actuality some of this time will be needed to achieve the right brightness of your colours - but the effect of how the screen is updated can result (for most people) in an apparent smoother interpolation between frames. To claim a plasma updates the image at 600hz/fps is a bit like counting black frame insertion on lcds as another fps. What a plasma tv does to update a frame:
https://www.youtube.com/watch?v=FMlpl7AmvNo

Many plasmas of course suffer from image retention and rely on noise/orbiting to reduce this. Relatively minor for most big screen displays, but would always prevent it from being used up close/applications (also combined with a typically noticeable operating hum - nothing too off-putting, but still, generally of note should you want to tout such panels in a professional environment, especially anything related to severe colour accuracy). However, playing to their strengths - as an affordable big screen display did, I think, help keep a focus on keeping input lag down on plasma products (you didn't want something getting out of sync with movement/sound).

The real quality advantage *at the time* was the black level that was achievable on a plasma was great, and angles were great too (another plus for bigger displays). And it was cost effective; for the larger screens in particular.

There were also some nice low input lag plasmas - right around the time that 100+ ms wasn't uncommon on lcd tvs and gamers in particular started to kick up a fuss, whilst of course av enthusiasts were already familiar with delaying their audio on their receivers and for them input lag never mattered as long as it was accounted for... This was also right around the time different inputs could be found to be treated differently - vga if available often being a goto test for the input means with the *likely* (but not guaranteed) way to test for lowest possible input lag on a set, before this tended to get a bit more unified.

There were still plenty of other plasmas about that were 50-100ms range though. And you would do well to find a set that was less than 30 so even though a lot performed well it'll still be a bit of a hunt to find the ones that people have let go (and not broken) that perform under 1 frame/30ms.

These days there are no plasmas being fabricated. And I'd only consider one used if you're getting a complete bargain price wise and/or can guarentee it's in good nick. The only exceptional one imho would be that pioneer pdp 402(?) that was 4:3 because I still like having a proper edge to my 4:3 games...

[maxtherabbit]

I need to correct myself - the CRT I was testing the plasma against (sony KV-32FV27) apparently had some lag

I got another CRT and ran more tests - the panny is 2-3 frames behind the CRT now on HDMI through the OSSC and 4-5 over analog

the internal deinterlacer is still adding 2 frames of lag, which is consistent with my previous testing
"game mode" has no appreciable effect on the results, it just makes the picture look worse lol

I can only attribute the lag on the sony set to the PiP function (even though it was off) - it's a regular SD set

What inputs were you using on the 32FV27? I think the composite and possibly the s-video inputs will be put through the 3d line filter which will cause latency. `8

I was only using YPbPr

[ChuChu Flamingo]

Although the plasma can update at 600hz you do not in effect get 600fps ... in actuality some of this time will be needed to achieve the right brightness of your colours - but the effect of how the screen is updated can result (for most people) in an apparent smoother interpolation between frames. To claim a plasma updates the image at 600hz/fps is a bit like counting black frame insertion on lcds as another fps. What a plasma tv does to update a frame:

I know 600hz isn't 600 fps. Like you said it is drawing each frame ten times to achieve the proper brightness for colors if im understanding for its sub field drive rate.

There were still plenty of other plasmas about that were 50-100ms range though. And you would do well to find a set that was less than 30 so even though a lot performed well it'll still be a bit of a hunt to find the ones that people have let go (and not broken) that perform under 1 frame/30ms.

I mean 1 frame of lag at 60 hz is 16.67 ms at the bottom of the screen. As I understand plasmas refresh the whole screen frame by frame vs a crt lcd that goes from top to bottom. If the plasma input lag results from avsforums are anything to go by, the lowest amount I've seen for a plasma is 20-21ms. That one being a Panasonic G10 50" which I own.

citrus3000psi however stated his PDPV402 has 2 ms. I assume total lag would be 18.67ms if we factor in 60 hz scanout or 2 ms over a crt?

[citrus3000psi]

citrus3000psi however stated his PDPV402 has 2 ms. I assume total lag would be 18.67ms if we factor in 60 hz scanout or 2 ms over a crt?

That was the OSSC latency tester, so not sure how accurate that is with plasmas. Fudoh basically calls it 'lag-free'

viewtopic.php?f=6&t=43256

[Roboplodicus]

Many plasmas of course suffer from image retention and rely on noise/orbiting to reduce this.

So the plasma will actually just use what looks like old fashioned noise to prevent permanent image retention? I've been playing about 3-4 feet from the screen and I definitely see some noise, is there any way I can discern if its supposed to be there?

Edit: Am I understanding it right that a plasma displays a entire green image, then whole blue image then whole red image successively many times a frame, rather than drawing left to right top to bottom once a frame?

[gray117]

So the plasma will actually just use what looks like old fashioned noise to prevent permanent image retention? I've been playing about 3-4 feet from the screen and I definitely see some noise, is there any way I can discern if its supposed to be there?

Almost certainly it is supposed to be there. A pixel orbiter will actually shift the image (exact implementations do vary either subpixels/complete pixel hop) to try and help prevent burn-in but I believe the noise is a similar technique. You may have options to turn off one/both but this is not generally advisable though - especially with kind of game content we have with huds/ui/text etc. which can typically cause retention...

Edit: Am I understanding it right that a plasma displays a entire green image, then whole blue image then whole red image successively many times a frame, rather than drawing left to right top to bottom once a frame?

Not really but implementations do vary. Some more effective than others - but basically the entire screen is pulsing several times a frame to build colours and brightness. Poor/early implementations might literally do this or come close to it depending on colours to be shown and method used... but I think it does vary with timing techniques either to mix colour/brightness - unless it's just an overall variance in the 600-400hz (less?) setup of the image pulsing on/off on some panels.

Upshot - unless you see/feel it, it doesn't really matter... if you do, plasma may not be for you.

[Roboplodicus]

Almost certainly it is supposed to be there. A pixel orbiter will actually shift the image (exact implementations do vary either subpixels/complete pixel hop) to try and help prevent burn-in but I believe the noise is a similar technique. You may have options to turn off one/both but this is not generally advisable though - especially with kind of game content we have with huds/ui/text etc. which can typically cause retention...

Ok that's sort of a bummer I play pretty close and the noise is a little bit distracting, although I only got the screen a week ago so Ill see how much it ends up bothering me over time.

Not really but implementations do vary. Some more effective than others - but basically the entire screen is pulsing several times a frame to build colours and brightness. Poor/early implementations might literally do this or come close to it depending on colours to be shown and method used... but I think it does vary with timing techniques either to mix colour/brightness - unless it's just an overall variance in the 600-400hz (less?) setup of the image pulsing on/off on some panels.

Upshot - unless you see/feel it, it doesn't really matter... if you do, plasma may not be for you.

I don't notice any flickering. I was curious because I was wondering at what point a latency testing device would detect a new frame. If the whole image was drawn at the same time then obviously there woudn't be a different latency time for the top bottom or middle of the screen.