[Moonchilde]

Ok, to make some people happy, lets consolidate the input lag discussion to this thread for the ST60. For some reason, the input lag on the ST60 is testing absurdly high. So lets discuss that a bit here. Going to be quoting some stuff in no particular order.

Quote:
Originally Posted by Phatal One  /t/1450484/official-panasonic-tc-pxxst60-series-thread/1230#post_23181985


I don't think it is anywhere near a "rising blacks" level of seriousness. One night of tinkering has shown me that there are a number of settings to help reduce the issue. Hopefully there are ways to further improve performance for those who care, as the ST60 is pretty strong in other areas. Eventually I will get around to testing 3D and some other things but input lag was a bit easier to design a decent and quick test for after a long day at work than to calibrate various picture modes. Plus, until my ZT60 gets here I will not have another 3D capable display to compare it to.

Actually, I do consider this quite serious. Why is it that the TV has double the amount of input lag as last year's model the ST50, which had about 8 ms of input lag over the previous ST30? Everything else is pretty nice, the image quality is fantastic and has many improvements over the 2012 models, such as better screen uniformity and a much less aggressive ABL. Horizontal color bleed is also much improved, and the gradients are far, far nicer than the UT50 I tried out for a couple of weeks. Everything is a huge improvement, except the lag. I don't feel like these improvements are because of image processing, I think they're simply making a much nicer panel than the previous years.


As far as the input lag, on 3D games, I'm not noticing it so much. 3D games tend to have a lot of start up frames and unlimited frames of animation, making it harder to tell. It didn't feel all that much different from the UT50. But, when it came to retro 2D stuff, that's when I took notice, and the ST60 definitely has a relatively high amount of input lag vs old school gaming on a CRT monitor.

Quote:
Originally Posted by Phatal One  /t/1450484/official-panasonic-tc-pxxst60-series-thread/1200#post_23181476


Well here are my preliminary results from my lag time tests:


I ran 50 iterations of the Human Benchmark Reaction Time Test . Unless noted otherwise, all tests on the ST60 were run with MPEG NR - OFF, Motion Smoother - OFF, and Video NR - LOW.


Control Test:

Dell U3011 - Avg 275ms

Known Input lag - 24ms

My Estimated lag - 251ms


Tests

Test 1: ST60 w/ Pixel Direct On - Game Mode Off - Pixel Orbiter On - Overscan On - Avg 347.0ms - Est. Display Lag 96.0ms

Test 2: ST60 w/ Pixel Direct Off - Game Mode Off - Pixel Orbiter On - Overscan On - Avg 322.5ms - Est. Display Lag 71.5ms

Test 3: ST60 w/ Pixel Direct Off - Game Mode On - Pixel Orbiter On - Overscan On - Avg 320.0ms - Est. Display Lag 69.0ms

Test 4: ST60 w/ Pixel Direct Off - Game Mode On - Pixel Orbiter On - Overscan On - Avg 315.0ms - Est. Display Lag 64.0ms

Test 5: ST60 w/ Pixel Direct Off - Game Mode On - Pixel Orbiter Auto - Overscan Off - Avg 305.2ms - Est. Display Lag 54.2ms

Test 6: ST60 w/ Test 5 Settings - Video NR Off - Avg 300.8ms - Est. Display Lag 49.8ms


Repeat Control Test:

Dell U3011 - Avg 279.9ms


I ran the 3rd test a second time because my animals decided to provide an unneeded distraction towards the end. Since there is a human and environmental element here I would say the range on these results is at least +/- 5ms and possibly add another +/-5ms b/c of the wireless mouse on my HTPC. I will retest with a wired mouse later but surprisingly paying attention long enough to click 250 times is somewhat fatiguing at the end of the work day!

Overall, I can say I was never able to match my times on the Dell U3011. I achieved my best times with game mode on and my own individual variability may be skewing the results of game mode a bit. At this point, I can say pixel direct appears to be quite intensive in processing requirements. It basically adds the entire input lag of the U3011 to the existing base panel lag. Game mode seems to be hit or miss in my results but I believe it is worth at least 10ms of processing improvement. I still have to finish the Guitar Hero calibration and would like to see if I can eliminate some of my own variability from the results above. The only thing I can definitively conclude from this limited data set is that the display and firmware in its current state has more than 2x the lag of the U3011 (which is not exactly a gaming monitor).


I'm willing to rerun some of these test in case I have some settings wrong or if there are other combos that may yield better results.


Updated:

Added Test #5 to account for pixel orbiter setting and updated Test 1-4 descriptions accordingly.

Reran the control test at the end of the night just to make sure Test #5 results weren't due to better response time on my part.

Added Video NR settings as suggested by Supermandlb

Added Test #6 to test the Video NR setting.

I saw this, and tried it out myself. Unfortunately, I think this has too much human error, for one, the first test you run you're still fresh from "playing" this game. I noticed that as you "played" more, you got better at it. I don't know if that actually correlates to the test results

Secondly, if you aren't 100% focused, then the results will vary too much as you get bored of looking for green. I personally couldn't get a compressed average window because I got bored after looking for green for a 10th time.

Quote:
Originally Posted by supermandlb  /t/1450484/official-panasonic-tc-pxxst60-series-thread/1200#post_23180324


Maybe I missed it, but have you tested the ST60's input lag?? I believe El Matadurr was referring to you testing the ST60 model

Yes, I did the camera and CRT method. It's not exactly scientific but it did give me a pretty good idea of the display's input lag over the CRT I tested on.


My settings were as such: Picture mode Custom; Color temp Normal; Vivid color off; CATS off; Video NR off; MPEG NR off; Game mode on; 1080p pixel direct off; HDMI content type off; HDMI/DVI RGB range non-standard; Black level light; Screen format full; Overscan off; H size Size 1 a.k.a. 100%; Pixel orbiter auto a.k.a. off. My tests yielded multiple 78 ms and even as high as 107 ms as the worst case that I had. Important to note was that there was no scaling by the display being done at the time of the tests, everything was a pixel 1:1 ratio.


I also tested the same settings above, Custom, with Game mode off. No difference, it would bounce around between the same 80 to 100 ms range as Game mode being on. Which really makes me wonder, why the hell is it bouncing from 80 to 100 with the same settings?


I ran another test with 1080p pixel direct on, and it hovered in the same amount of ms. I only tested this because there was one photo of a CRT vs ST60 floating around that claimed it was higher, but my same tests with it off would fluctuate by about 20 ms so I have to say that this mode probably doesn't really add that much to it.


I tested Vivid mode which disables the Pro Settings mode that Custom and Cinema has, and no difference.


I did not test Motion smoother, as that seemed a waste of time to me.


Hope that helps answer some of your questions.

Quote:
Originally Posted by gbmannc  /t/1450484/official-panasonic-tc-pxxst60-series-thread/1200#post_23181655


That is a head scratcher isn't it. Could the improved sharpness people have been reporting be the cause due to increased image processing?

I honestly don't see a difference between the "sharpness" of the image vs the UT50 with the sharpness turned off. Everything looked pixel clear on both sets, the only difference I notice is how reflective the screens were.

Quote:
Originally Posted by El Matadurr  /t/1450484/official-panasonic-tc-pxxst60-series-thread/1200#post_23181638


...that's easily double the lag as the ST30. Those sensitive to lag would surely notice that.


This does beg the question though--one that would be great for Panasonic to answer--with everything turned off and game mode activated, why is this year's model slower than previous years' models? Does Murphy's law have a backtracking algorithm?...

The ST30 was 32 ms of lag. Pretty good, I was hoping the ST60 was going to go back to those measurements, but alas... Anyway, having played some modern games on a UT50 and then played on a ST60, there isn't too much of a difference, at least not noticeable right away. Unfortunately, I did not get to do my retro game test on the UT50. Now I wish I did.

Quote:
Originally Posted by El Matadurr  /t/1450484/official-panasonic-tc-pxxst60-series-thread/1200#post_23181715


What??? You can't disable the pixel orbiter on this year's model? Now that's just crazy. I wonder how many other settings have ON or AUTO (a.k.a. still on).

The Auto setting is equivalent to off. I don't know why they renamed it, maybe to cut back the paranoia of burn in?

Quote:
Originally Posted by fahrenheit  /t/1450484/official-panasonic-tc-pxxst60-series-thread/1200#post_23180330


Well if we don't know what is responsible for the lag, then we don't know what will fix it.


Keep in mind that I've got nothing to base the following on and don't know whether this is occuring with the 2013 models. Owners will be able to correct me if this is off-base....


I'd speculate that it may be because of how My Smart Home is implemented.

On my GT50, the internet-based functions are not a service that is running in the background, it has to load when it is called. I also notice on my friends ST50 that this process is even slower (slower SoC). If on the 2013 models this process is always on (running in the background), then that would require additional processing power and/or buffering that could be responsible for the lag. The upper models (GT60 and beyond) have a more powerful SoC and its likely that they can handle this processing without having any net-affect on input lag.

If this were the reason for the lag, then Panasonic would need to (in firmware) disable the My Smart Home functionality when Game mode is chosen.


Just out of interest, it would be good if an owner could report what happens with game mode enabled and they call up My Smart Home. Does game mode stay enabled or not?

Could you describe My Smart Home? I would like to see what you're talking about. For what it's worth, if I hit the Home button, the set zooms the picture out and shows me a menu around the image. This is always a running process and has no loading time at all, I can switch to Full Screen TV, Lifestyle Screen, or Info Screen all on the fly. There is no longer a Viera Tools, instead those are on the Apps menu, which loads up a ton of stuff instantly. There is a paint program that will load instantly, other programs must be added to a Home screen to be launched from, and then there are internet programs which require connecting to, probably like a cloud app. The whole OS runs on top of the image, I can access Home at any point and it will be like pressing the Home button on a PS3 mid game to bring up the XMB, so I know it's always running.


I hate Smart TV features. Why can I just get a nice image with pro menu controls? I don't want all this other garbage running. Seriously, I would love a custom firmware that hacked all that crap out and left me with just the menu.

 

[El Matadurr]

A thread like this was warranted Moonchilde. Thanks for posting.

Though I can't help but think it'd be helpful to mention the S60 in the thread title as well, since input lag is a concern over there, too (it's looking much better, ala ST30/ST50 levels of input lag). Apart from smart features and a good anti-reflective screen, they seem similar in raw performance (ST60 has more gradation steps, some more user-menu calibration tweaks, etc.).

Quote:
Originally Posted by Moonchilde  /t/1467481/official-st60-input-lag-thread#post_23182168


I hate Smart TV features. Why can I just get a nice image with pro menu controls? I don't want all this other garbage running. Seriously, I would love a custom firmware that hacked all that crap out and left me with just the menu.

Now to start a "How to hack the ST60's Firmware" thread.

 

[xcloudx82]



Quote:
Originally Posted by Moonchilde  /t/1467481/official-st60-input-lag-thread#post_23182168


Actually, I do consider this quite serious. Why is it that the TV has double the amount of input lag as last year's model the ST50, which had about 8 ms of input lag over the previous ST30?

From hdtvtest.co.uk review of 42ST60 (EU) i've read:


The TX-P42ST60 doesn’t feature Panasonic’s newer 6-core “hexa processing engine” though, and instead runs off the older V-Real 3D Pro system (apparently running on a chip from Mediatek, as can clearly be seen through the vent on the back, rather than Panasonic’s own silicon)


And, some time ago, in the same forum i've read that last year's models didn't use this chip from Mediatek, but Panasonic own SOC (UniPhier MN2WS02 ARM Cortex 9, [email protected]).


Maybe this mediatek chip is causing more input-lag? It's not present in the reviewed GT (in fact it has the new hexa processing engine) and input lag is way much better...what do you think?

 

[Moonchilde]

Possible, if the chip is slow then that could be the cause, if it can't handle processing the image and the OS on top of it. I wonder what chip the S60 uses, if it uses the Mediatek chip or Panasonic's own SOC.

 

[Stump909]

Overall, it' starting to look like a process bloated system. In some regards that's a good thing, since it's more software related than hardware, but it also means a solution will have to come via firmware or a service menu setting.

 

[Stump909]

Overall, it' starting to look like a process bloated system. In some regards that's a good thing, since it's more software related than hardware, but it also means a solution will have to come via firmware or a service menu setting.

 

[CSP2]

Midkay posted a really good video showing a visualization of the ST60's lag: https://www.youtube.com/watch?feature=player_detailpage&v=VXMsmQqV6zs#t=28s


The explanatory post is here: http://www.avsforum.com/t/1461177/official-panasonic-s60-series-discussion-thread/390#post_23182137


I take it Game Mode is on in this video, but I'm not certain.

 

[xcloudx82]

great video...so much input lag...for me this is unbearable for PC gaming

 

[gregleg]

Wow, that video certainly makes me hesitate when it comes to this TV. IF the settings can be tweaked so that the lag can be reduced to a more reasonable level without killing the great image quality, or IF a firmware update can help and it's not because Panny cheaped-out on the processor, then MAYBE this TV is still in the running to replace my old SXRD. I was all set to do a "shut up and take my money!" on a 65ST60, but now I'm waiting...

 

[Moonchilde]

I've disabled just about everything I can in the user menu and nothing really helps it get lower than 78 ms (the lowest I was able to get) and service menu is extremely barebones. There is pretty much no option you can change to help get the set lower. This is most definitely a bad combination of too much software and not enough processing power. I refuse to believe the panel itself is this bad. Even service menu takes about 2x as long to load up as the UT50, there is definitely something fishy going on. I'm not really sure a firmware update will fix it, because I doubt Panasonic is going to be willing to spend the money to remove features that people may complain about missing after seeing them supported on the set. Something like this is mostly going to end up a firmware hack, IF someone is actually willing to hack it. Sadly, it's far beyond my knowhow or else I'd have removed everything from the firmware other than the regular user menu. Shame because the panel looks so damn nice, even if I can't stand the trailing of fast moving horizontal objects.


I miss the ultra complicated service menu for my Sony XBR910, you could change just about ANYTHING via service menu, it was insanely cool. I also miss how damn reliable CRT is.


At this point, I dunno what to do.

 

[Looke89]

I have a question: Is the input lag less noticeable in games running at 30 frames per second? When the game is running at that speed, it takes about 33.3 ms to generate one frame. Therefore, if people usually only start to feel the difference if the lag lasts for longer than 3 frames of picture, tv with 99.9 ms input lag should be ok for games running at 30 fps. Is my thinking process right?


Cheers, Luke

 

[Moonchilde]

No, because games that run at 30 fps have about 133 ms of input lag naturally. So a game that runs at 60 fps natively has 4 frames of input lag, plus your TV lag. That means with the ST60 we're look at about 10 frames of total lag from the time you press the button in a best case scenario. When you play a game that is 30 fps, it has 8 frames of input lag natively, and the ST60 has about 6 frames so you're looking at 14 frames of lag from when you press a button. This isn't even taking into consideration games that have extra post processing that adds additional lag.


So, lets say I play a game at 60 fps, well it's going to feel more like 30 fps and maybe that's why it feels "ok" for games like Darksiders on PC. Games on consoles that run at 30 fps and even sometimes struggle to maintain that, aren't going to feel good, although IMO for modern 3D gaming it doesn't feel that detrimental to the games. Retro 2D games like old NES titles and arcade games you'll definitely feel it. Maybe you can adjust, I'm sure I could, but honestly, why should we when it's expected that other sets are up to a 1/3 faster.

 

[Looke89]

Thanks for a thorough answer

That's quite weird because 14 frames of lag should be quite noticeable (especially as I'm quite sensitive to such things), yet it feels exactly the same like my previous Philips LCD screen. Maybe EU model differs to the US one a little bit?

 

[Moonchilde]

It really depends on the game. Some games, like Killzone 2, have native 12 frames of input lag before you add on your display lag. So that means, playing that game on a ST60 would net you 18 frames of input lag.

So far, both the NA and EU models are coming in at about the same amount of lag. HDTVTest had it at 73 ms of input lag, yet their camera tests were at 47 which is far lower than my camera tests. I'm not really sure what's going on.

 

[Looke89]

I found camera test results for P50ST60 on one of the Polish HDTV sites (performed by a professional), and the results were 100 ms with Game Mode off and 50 ms with Game Mode on. Interesting.

 

[El Matadurr]



Quote:
Originally Posted by Moonchilde  /t/1467481/official-st60-input-lag-thread#post_23183506


They obviously read the email proper.

Honestly, pretty typical of e-mail customer service.

 

[Phatal One]



Quote:
Originally Posted by Moonchilde  /t/1467481/official-st60-input-lag-thread#post_23182168


Unfortunately, I think this has too much human error, for one, the first test you run you're still fresh from "playing" this game. I noticed that as you "played" more, you got better at it. I don't know if that actually correlates to the test results

Secondly, if you aren't 100% focused, then the results will vary too much as you get bored of looking for green. I personally couldn't get a compressed average window because I got bored after looking for green for a 10th time.

That would be why I repeated the test 50 times in each iteration... statistically the odds of being perfect or stupidly distracted 50 times is infinitesimally small. Running the test this many times minimizes the impact of outliers such as a lucky click or the user being distracted as the user's performance will revert to the mean over the long term. (Kind of like flipping a coin... you can get lucky and get 5 heads in a row... but by the time you do it 50 times you are likely close to 25 heads - 25 tails) Additionally, I provided estimated margins of error for the data points. Had the data been in a format that I could analyze I could have provided specific margins of error. In fact, I would have run a full statistical analysis on the data for each of the tests. Lastly, I did not tend to get better at it. If you notice, I purposefully reran the control test at the end of the night to determine if I had improved during the test. I actually performed slightly worse than when I originally ran the test.


I will agree that there are some variables that I would like to eliminate. If I could eliminate myself from the equation, that would be perfect. If I could have repeated each test 1000 times, that would have been even better. If I could use a USB light/color meter or something like that to automate that task that would be preferred. In that instance I could easily program a simple program to do the test without need for a website. I am actually thinking about whether I care enough about the challenge to actually do it!

I have been very open about the test conditions, factors that may have limited the accuracy, and provided some conservative confidence intervals up front. All of that said, despite your reservations, you will find my results are within the range of what others are reporting in regards to input lag measurements on the ST60.

 

[Phatal One]

delete

 

[MarkAELSU]

Man, I have had this preordered since the beginning of February. Now I'm thinking about cancelling it. I use my tv to game about 65%. So this MAY make me choose something else, but as a plasma fan boy, and with a $2,000 max spending budget on such a set, what else should I choose? 55" or larger

 

[Stump909]



Quote:
Originally Posted by MarkAELSU  /t/1467481/official-st60-input-lag-thread#post_23183877


Man, I have had this preordered since the beginning of February. Now I'm thinking about cancelling it. I use my tv to game about 65%. So this MAY make me choose something else, but as a plasma fan boy, and with a $2,000 max spending budget on such a set, what else should I choose? 55" or larger

Being in a similar boat...I may hold out for the Costco S64's. They're the S series with the ST filters. Still I haven't fully made up my mind yet. I really want to see an US review.