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. - Page 3

post #61 of 687
5/8/10 at 1:30pm
The other night I tested my 54G25 off a PC using SMTT and RefreshRateMultiTool.

tl;dr: At 1080p most tests showed 31ms or 2 frames, with a few results of 16ms or 1 frame. RB2 tests on Wii gave me 23ms audio lag and between 16 and 32ms video lag.

I cloned my PC to a CRT monitor and the plasma: my video card is a Radeon HD 5770 with the signal cloned; DVI->VGA to the CRT and HDMI to the Panny, both set to 1920x1080 output. Take these with a small grain of salt, as I'm not convinced the cloning's perfectly in sync. I would have preferred to do the 60fps test but my camera only does video at 30fps.

One thing to note is that the blue phosphors react much faster than the red and greens, which is why I did blue numbers on black for SMTT and why you'll see blue, white and yellow frames on RRMT.

Ignore the flashlight, that was my ghetto way of decreasing the shutter speed on my camera :P

Oh and FWIW I did the RB2 test with RB2 Wii (480p) a few times:

Audio Lag = 23ms
Video Lag = Between 16 and 32 ms, which fits with the results below:





Reply
Reply

AVS Top Picks

post #62 of 687
5/8/10 at 2:01pm
Very cool. I will use that program to do more thorough tests on my pk550 now.
Reply
Reply
post #63 of 687
5/8/10 at 2:44pm
Thread Starter 
Thanks for the results NOA, I'll update the OP.
Reply
Reply
post #64 of 687
5/9/10 at 5:42pm
Thank you, NOAMattD! I had been waiting for someone on the forums here to corroborate the hard to find input lag tests for the G25, especially since I broke down and got a one this weekend :P Glad to see that there appears to be a general consensus on the 16-32ms mark for the input lag on these sets.

One question about this, though: does this 16-32ms variance have more to do with the testing methods used or with the input lag on the set actually fluctuating? Being a gamer, I would certainly prefer the former, as the latter could cause some slight problems in competitive situations.
Reply
Reply
post #65 of 687
5/10/10 at 9:08am
Orta - Note that my set is a G25, not a G20. You might also want to include sizes in the OP so we can track if there are any variances based on size.

Phasemaster - The tests themselves definitely have something to do with it, but I'm definitely not qualified to say if lag is variable within a set. I don't think lag is a static figure like say pixel response time or refresh rate.

My RB2 tests were manual calibrations - alone they're only good for a very general idea bordering on subjective "it feels" anecdotes (it felt lageless to me BTW ). The stopwatches are better but still not quite perfect. Even the shoryuken.com 60fps test, accepted as the most accurate method among fighting game players, produces at least two results (e.g. 1-2 frames).
Reply
Reply
post #66 of 687
5/10/10 at 10:20am
Thread Starter 
Quote:
Originally Posted by NOAMattD View Post

Orta - Note that my set is a G25, not a G20. You might also want to include sizes in the OP so we can track if there are any variances based on size.

Phasemaster - The tests themselves definitely have something to do with it, but I'm definitely not qualified to say if lag is variable within a set. I don't think lag is a static figure like say pixel response time or refresh rate.

My RB2 tests were manual calibrations - alone they're only good for a very general idea bordering on subjective "it feels" anecdotes (it felt lageless to me BTW ). The stopwatches are better but still not quite perfect. Even the shoryuken.com 60fps test, accepted as the most accurate method among fighting game players, produces at least two results (e.g. 1-2 frames).

Ah yea thanks, I didn't catch that. I just replaced apfhex's preliminary G20 results with yours, which is why it's mislabeled. I'll be sure to include size differences if we get multiple results in the same model, but with only 1 measure, I think the prefix makes the model numbers a little more confusing to find or follow. I'll keep note of them though.
Reply
Reply
post #67 of 687
5/10/10 at 10:22am
Awesome thread! I hope to see someone post results for a Samsung C8000. Once my set is delivered I will test out if noone else has by then.
Reply
Reply
post #68 of 687
5/18/10 at 10:40pm
Could anyone test the PN50C6500? I'd really appreciate it.
Reply
Reply
post #69 of 687
5/20/10 at 6:58am
60pk550 hdmi input:
winblows xp, both crt and display are set to 1920x1080@60hz


.
.
.
.
.
.


60pk550 vga input
winblows xp, both 1920x1080@60hz


those last two are messed up because I decided to meter off of the overhead light, so they might be more accurate? Dunno how this stuff works. I don't have an actual camera. Using the sony camcorder to take these.
Reply
Reply
post #70 of 687
5/20/10 at 10:37am
Thread Starter 
I'm not familiar with that test you're using shad, is the white square advancing 1 block per frame? The multi-colored phosphor tailing on the PDP also seems to be making it even harder to read, do you ignore the blue/yellow trails and just count to the solid whites? Were you not able to get a stopwatch timer to show up?
Reply
Reply
post #71 of 687
5/20/10 at 12:17pm
Yeah, the program turns one block white for every frame of the display. It sequences left to right then goes down a row. So it flashes 60 blocks per second in sequence when displays are set to 60hz. So 16.67ms per block.

So on hdmi pictures it seems to be 2-3 so ~50ms? And almost always 2 blocks behind on vga so 33ms

This is faster than what I measured before with the stopwatch, I was getting 80 or so on hdmi and 40 on vga.
And I am skeptical that the difference in the VGA and the HDMI is that little. I notice a large change in mouse responsiveness between the two inputs.

What is that timer software that shows multiple clocks up and down the screen?
Reply
Reply
post #72 of 687
5/20/10 at 12:52pm
The multi-stopwatch is Small Monitor Test Tool, or SMTT. I got it from frito, pm me and I'll send you a copy.

Orta - The "blocks" test I linked to in my G25 post is RefreshRate MultiTool, or RRMT. It runs at 60fps and draws one full block per frame, so you count the difference between the CRT and HDTV.

In the case of plasmas the blue phosphors react quicker and the red/green decay slower, so we see in some photos a sequence of blue-white-yellow, where the input's been received and the phosphors are starting to react but the images in the previous frames haven't yet decayed. I counted the difference starting with the blue frames if they were there.

shaddix, your tv has about 3-4 frames of lag in both modes. 0 frames of lag would be if both blocks were white (or blue/white) at the same time. Most of the time the difference two black blocks between where the plasma is and where the CRT is, meaning 3 frames.
Reply
Reply
post #73 of 687
5/21/10 at 4:57am
Quote:
Originally Posted by NOAMattD View Post

shaddix, your tv has about 3-4 frames of lag in both modes. 0 frames of lag would be if both blocks were white (or blue/white) at the same time. Most of the time the difference two black blocks between where the plasma is and where the CRT is, meaning 3 frames.

Thanks for interpreting those results for me. That makes more sense and is more in line with what I had tested earlier with a stopwatch. Except now I'm thinking perhaps I had a setting on the TV wrong because VGA is quite noticeably faster than hdmi in mouse responsiveness. But the pictures don't seem to verify that.

With hdmi it feels more like a wiimote. But using vga its much more like a mouse on an lcd monitor.
Reply
Reply
post #74 of 687
5/22/10 at 12:04pm
new to this topic... it seems like the most discrete unit used is a single frame/16ms. Why is that? Are plasmas and CRTs displaying at the same frequency?

If not, then it should be able to measure more precisely using sequential images and higher shutter speeds.

also for future tests can non-native resolutions be included?
Reply
Reply
post #75 of 687
5/23/10 at 3:45am
Quote:
Originally Posted by nanothis View Post

new to this topic... it seems like the most discrete unit used is a single frame/16ms. Why is that? Are plasmas and CRTs displaying at the same frequency?

If not, then it should be able to measure more precisely using sequential images and higher shutter speeds.

also for future tests can non-native resolutions be included?

input lag is most relevant for games, and most games that critically depend on input lag such as fighting games and shooters run at 60fps on consoles. So it's easy to think about how much lag a TV has based on how many frames the display is running behind the console or PC.

For these tests, yes, the plasmas and CRTs are both running at 60hz
Reply
Reply
post #76 of 687
5/25/10 at 5:14pm
I had issues performing this testing due to what I suspect is an EDID issue (ATi HD5670, Windows 7 64bit). I wasn't able to clone the desktop in 1080p despite being able to set CRT to 1080p in 'extended' mode.

I was able to test at 720p though. My 50" V10 was consistently 50ms (3 frames) behind the CRT. I tested THX, Normal (Custom mode in the US I believe?) and Game mode. There was no decernible difference between them.

What I cannot deduce from this is whether or not scaling has affected the results and whether running at native (1080p) would shave off a frame?

I run my 360 at 1080p and the PS3 is set to select 720p if the game deems it, but I'm thinking I'd now be better off forcing the PS3 to upscale by deselecting 720p when I can.
Reply
Reply
post #77 of 687
5/25/10 at 7:16pm
Quote:
Originally Posted by fahrenheit View Post

I had issues performing this testing due to what I suspect is an EDID issue (ATi HD5670, Windows 7 64bit). I wasn't able to clone the desktop in 1080p despite being able to set CRT to 1080p in 'extended' mode.

I was able to test at 720p though. My 50" V10 was consistently 50ms (3 frames) behind the CRT. I tested THX, Normal (Custom mode in the US I believe?) and Game mode. There was no decernible difference between them.

What I cannot deduce from this is whether or not scaling has affected the results and whether running at native (1080p) would shave off a frame?

I run my 360 at 1080p and the PS3 is set to select 720p if the game deems it, but I'm thinking I'd now be better off forcing the PS3 to upscale by deselecting 720p when I can.

1080 will most definitely decrease the input lag, the g10 is extremely fast so unless something crazy went on with the v series then it should be fast too. But sadly the ps3 can't upscale. If you deselect 720p output it will just output 480p for 720p games
Reply
Reply
post #78 of 687
5/25/10 at 7:22pm
Quote:
Originally Posted by shaddix View Post

But sadly the ps3 can't upscale. If you deselect 720p output it will just output 480p for 720p games

I not sure that applies to titles these days. I know it was the case at launch, but I believe something changed a couple of years back and upscaling became possible.

Will test it later to be certain.
Reply
Reply
post #79 of 687
5/25/10 at 7:39pm
Quote:
Originally Posted by fahrenheit View Post

I not sure that applies to titles these days. I know it was the case at launch, but I believe something changed a couple of years back and upscaling became possible.

Will test it later to be certain.

Games have to be made with upscaling built into the code. (FFXIII, MW2 both do this) Sadly not all games do this.

(imagine FFXIII with it's 1080p CG and 720p ingame switching back and forth on TVs that take a while to switch resolutions, it would have been a major issue)
Reply
Reply
post #80 of 687
5/25/10 at 7:52pm
Thats a bummer, I guess I'll be keeping 720p ticked then.
Reply
Reply
post #81 of 687
5/25/10 at 8:48pm
I was looking at some 720P plasmas for my kids' game room (trying to keep the price reasonable). Does anyone have any idea on input lag associated with these sets?

I was looking at Panasonic, Samsung and LG. Looks like LG is probably a bad choice, but I am curious about Panasonic and Samsung input lag.

They will be playing Wii, PS2 and PS3 (no war games except Call of Duty), but GH, RB. Shawn White, Mario, etc.)
Reply
Reply
post #82 of 687
5/25/10 at 9:04pm
I have no real lag issues with my Samsung 50c450. Music games can be a minor issue but it isn't that bad. (Panasonic is better with lag, but with the black level issues, I just can't recomend them until they actually fix their built in defect)
Reply
Reply
post #83 of 687
5/25/10 at 11:43pm
Thread Starter 
Quote:
Originally Posted by fahrenheit View Post

I had issues performing this testing due to what I suspect is an EDID issue (ATi HD5670, Windows 7 64bit). I wasn't able to clone the desktop in 1080p despite being able to set CRT to 1080p in 'extended' mode.

I was able to test at 720p though. My 50" V10 was consistently 50ms (3 frames) behind the CRT. I tested THX, Normal (Custom mode in the US I believe?) and Game mode. There was no decernible difference between them.

What I cannot deduce from this is whether or not scaling has affected the results and whether running at native (1080p) would shave off a frame?

I run my 360 at 1080p and the PS3 is set to select 720p if the game deems it, but I'm thinking I'd now be better off forcing the PS3 to upscale by deselecting 720p when I can.

Did you remember to turn off Windows Aero to make your measurements?

Quote:


On PS3 scaling...

Right, there was no global fix, it still varies from title to title. Even among games that support it, there is big differencing in scaling quality. There is presumably some crappy implementation included in the SDK or something (that scales to 960x1080 IIRC), but some developers apparently opt to improve on it or create their own. Titles with quality 1080p scaling will usually default to it over 720p (e.g. FF13, MGS4). Titles that do support 1080p scaling, but employ the low quality version (e.g. Uncharted 2) will default to 720p. First party developers are usually aware which resolution looks best and will default to the better of the two, but you have to be careful of ports and multiplatform fare.
Reply
Reply
post #84 of 687
5/26/10 at 1:32am
Quote:
Originally Posted by Orta View Post

Did you remember to turn off Windows Aero to make your measurements?

Facepalm moment!

I got so hung-up on not being able to get 1080p working that I forgot to disable it.

Retesting shows inputlag.exe running at 40ms behind on the plasma and RRMT showing 2 frames (32ms).

Sorry for the confusion and thanks for the PS3 explaination.
Reply
Reply
post #85 of 687
5/26/10 at 7:07am
Would love a 58V10 lag test. LG PK950 too. Someone take their nettop to best buy. Chances are they'll have no idea what you're doing. Looks like the G series is the leader of the pack, hopefully the V10 follows suit. I'm planning on picking mine up within the next two weeks.
Reply
Reply
post #86 of 687
5/26/10 at 4:12pm
posted my current issues in the lag wars thread but probably fits better here so i'll just quote it:

Quote:
Originally Posted by schwuppe View Post

Hey guys, first post here.

Read up on the last couple of pages and from what I understood you're all pretty hyped about the S1 (which should be the S10 in Europe I guess? Me from Ger).
I'm just researching on what best Gaming TV to get and I've come across the S20 (S2 in US?!) which has had very good test results as far as gaming is concerned (check review here http://www.hdtvtest.co.uk/news/panas...0100511647.htm) vs the GW10 (G10 in US).

First thing is, the S20 from this year is probably on par with the G10 from last year (although the latter has been pricier originally) through improvements and what not. So I guess it would be the wisest choice to get the S20 (S2? Is it really that in the US?) because it's cheaper and has 2010 standard unlike the G10 (G20 has had pretty bad results, so it's not S20 vs G20).

So yeah, to get this clear: I wanted to buy the S20 but now I'm reading people in 2010 are so excited about the S10 (or S1)... why is that? Was the S20 also a step down in comparison to the S10 (or S1) as far as responsiveness/input lag goes?

Greets

*PS* and we're talking about the plasma tv here, of course. (just saying because there is also the LxxS2)

The first part doesn't apply in this topic I see as it referred to the other lag wars thread where I thought they were talking about the PxxS1 and not the LxxS1. With the different types I mention (S1, S2, G10, G20) I mean the plasma versions of course.
Reply
Reply
post #87 of 687
5/26/10 at 4:29pm
Quote:
Originally Posted by fahrenheit View Post

I not sure that applies to titles these days. I know it was the case at launch, but I believe something changed a couple of years back and upscaling became possible.

Will test it later to be certain.

Still applies, does vary by game but many new games are still 720p. Why I opt for the 360 version when buying games so that I can take advantage of the A650's PC-Mode with is essentially lag free.
Reply
Reply
post #88 of 687
5/26/10 at 4:58pm
Thread Starter 
Quote:
Originally Posted by schwuppe View Post

posted my current issues in the lag wars thread but probably fits better here so i'll just quote it:



The first part doesn't apply in this topic I see as it referred to the other lag wars thread where I thought they were talking about the PxxS1 and not the LxxS1. With the different types I mention (S1, S2, G10, G20) I mean the plasma versions of course.

I think you would have to ask other Europeans about that. There clearly seems to be major territorial differences in the performance of these models. The NA S2's color accuracy is quite a bit off compared to what that review shows of the EU S20.
Reply
Reply
post #89 of 687
5/26/10 at 5:34pm
You're right.. and I don't even know what the equivalent to the EU S20 in NA is... I guess it's not the S2 because in the other lag wars thread it said that the S1 (NA) equals the V10 (EU). I guess the same categories apply for the Plasma TVs.

The V20 (EU) Plasma is the most expensive series (only the VT20 is more expensive - but that's 3D).
Question would if the S2 is also the most expensive Plasma in NA (minus 3D).

And then it would be cool to know what the EU S20 equivalent is because this one got a very good review as can be seen in the link from my last post but I can't really be sure if it's valid and if this forum would recommend it as well.

tbh, I have looked for and researched for like 4 days straight now and compared, compared, compared and the only thing I'd like to know now is what's the best pick in regard to Input Lag / price / picture quality atm. I can spend 1100-1200$ at max and the most important thing would probably be Input Lag because that's why I'm here.
Reply
Reply
post #90 of 687
5/26/10 at 7:53pm
Does anyone have numerical performance data on 720p plasma?

PENDRAG0ON seems to have no issues on a Samsung 50C450 (a little lag on GH/RB games). It would seem to me that since the games are not the same native resolution, the scaling from 720p to 768 would cause additional delay.

Does any one play Wii or PS-2 games? How bad is the lag on 480i games?

Is it that much different? I am trying to understand how lag is affected by the scaling required by the TV. I am not a gamer, so I do not have specific experience.

I just want my kids to be able to play games as well as the old CRT that this panel is replacing.
Reply
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Plasma Flat Panel Displays

AVS Top Picks

AVS › AVS Forum › Display Devices › Plasma Flat Panel Displays › The Official Plasma Input Lag Thread.