Understanding input lag test

B

Bun-Bun

Gawd
Joined
Sep 11, 2003
Messages
555
Two displays beside each other connected to the same computer running that stopwatch app.

Sorry for my shitty cell phone pics, the camera has issues.

YMKW8x9.jpg


That picture looks the same... so same input lag?

ODqDuA3.jpg


This one looks like the left one is 10-20ms faster?

Before I did this test it felt to me the display on the right had more input lag than the display on the left. Then I hooked them up at the same time and did this test and im a bit confused by the results.

Rest of the pics are here http://imgur.com/a/qlUcL
 
Last edited:
You can connect two to the same computer and do accurate testing only if you have a dvi-splitter, so they get the exact same signal, not with clone mode.
 
Morkai said:
You can connect two to the same computer and do accurate testing only if you have a dvi-splitter, so they get the exact same signal, not with clone mode.
Click to expand...

The two displays are connected via DVI-> HDMI -> monoprice HDMI 2x1 splitter -> each display.
 
imo the best way to measure relative input lag without special equipment is this:
http://forums.blurbusters.com/viewtopic.php?f=10&t=1162

basically take advantage of a camera's rolling scan operation to allow sub-frame accuracy

as for the stopwatch method... its pretty much useless unless you want to characterize the lag of something that is really laggy (like tvs with 80ms lag)
 
I got out my Nikon D600 and did some more pics at 4000th of sec shutter speed.

http://imgur.com/a/mcvje

All pretty much the same. On a couple the circle animation is one tick ahead on the left display.

So for all intents and purposes I'm going to say these displays have very similar input lag and that it shouldn't be a deciding factor between the two.

Now that I have them both set up and I switch between them I notice no difference and if I focus in the center above them and only have them in my peripheral vision I notice zero difference.

Interestingly if I put the nvidia driver in Y Pb/Cb Pr/Cr mode I perceive less input lag. Could be in my head though. I have no way to prove this.
 
Last edited:
did you read the links provided before? it doesn't matter what camera you use, or the shutter speed, using the single stopwatch method like that is unreliable and will give you errors.
 
Drags said:
did you read the links provided before? it doesn't matter what camera you use, or the shutter speed, using the single stopwatch method like that is unreliable and will give you errors.
Click to expand...

Right, but how much error are we talking? the error of 1 frame? ~16ms. If at any given point in time within 4000th of a second the displays are showing the same information then for all intents and purposes they have the same input lag. Obviously the display on the left is faster but within 16ms faster.

When I was originally comparing them separately I was perceiving a significant increase in input lag when switching to the display on the right. This was a major deal breaker to me. But it was obviously in my head with the results I am getting with them side by side.

I was going to go get the SMTT 2.0 but can't find a free download and the maker is no longer selling it.
 
we're talking about an error which for various reasons makes using the single stopwatch very difficult to trust. have a read of that article at TFTCentral which explains it well. that might also explain why you're feeling one screen is faster than the other (probably not in your head) even though that erroneous stopwatch method says otherwise. it's just too prone to error to use reliably at all.

in the absence of being able to test it any more reliably, you could probably base it on how each of them feels to you i would think
 
Drags said:
we're talking about an error which for various reasons makes using the single stopwatch very difficult to trust. have a read of that article at TFTCentral which explains it well. that might also explain why you're feeling one screen is faster than the other (probably not in your head) even though that erroneous stopwatch method says otherwise. it's just too prone to error to use reliably at all.

in the absence of being able to test it any more reliably, you could probably base it on how each of them feels to you i would think
Click to expand...

I read that article and I realize there is error and without other methods impossible to get within 16ms of accuracy using the stop watch method. What I was initially perceiving was a difference like between my Samsung 930BF and 244T monitors (difference of 30-50ms input lag). That is definitely not the case here.

I did more than the shots I linked as well. Tried a couple different FPS @ 4000th shutter speed over 100 shots. Same results.

Plus now that I do have them split side by side I have spent 6 hours watching video content and playing games and no longer perceive a difference.
 
Last edited:
It's extremely doubtful you'll perceive a difference if both are within 1-2 frames of input lag, unless you're playing extremely fast paced games.

Is the TV on the right an F8500?
 
Nikyo said:
It's extremely doubtful you'll perceive a difference if both are within 1-2 frames of input lag, unless you're playing extremely fast paced games.

Is the TV on the right an F8500?
Click to expand...

PN58C8000 on the left, PN60F8500 on the right.
 
Bun-Bun said:
PN58C8000 on the left, PN60F8500 on the right.
Click to expand...

The F8500 has around 62-64 ms input lag according to hdtvtest and http://www.displaylag.com/display-database/; both use the Leo Bodnar method. That's around four frames of input lag.

If you're not perceiving a difference, then I don't really suspect that the C8000 is any higher/lower, but I also can't find any good measurements of it, given it's age. Overall, I'd keep the F8500 since it's one of the only plasmas that can actually be used in a well-lit room without destroying the black levels.
 
Last edited:
Yes that's the whole problem, no good measurements of the C8000.

Now that I have had the F8500 directly beside my C8000 (and calibrated it to match my C8000) I am going to keep the F8500. Once I make the warranty exchange I will have two F8500's! :)

Its really too bad these plasma's are no longer being made. Wouldn't mind trying out a Panasonic ST60.
 
You must log in or register to reply here.
Back
Top