• Welcome to AVForums running Xenforo 2.2. The upgrade from 2.1 was successful and finished in just 90 minutes, well before 8am. The most notable changes in 2.2 are listed in this post.

LG 920 and 950 input lag figures

JimBob1971

Well-known Member
There was been a lot of talk about lag in the 950 thread so I thought I'd buy a Leo Bodner tester and see for myself as I currently have both sets.

TLDR;
Both sets average low 50ms. PC mode/GAME doesn't seem to change matters, although ISF mode seemed to have least lag strangely. Clear motion takes lag up to about 128ms. Setting sound out to headphone rather than speaker didn't seem to make any difference.

I turned all processing off for both sets, and tested with the input type as PC, HDMI and HDMI that was labelled as PC (as some people thought that made a difference on 930). In each of these I tested GAME and ISF 1.

The Bodner device displayed different measurements each go, and could vary over 10ms with each run. For example, one reading for the top bar might give 43, then the next time I might see 60! I've taken 4 reading for each bar on each setting for the 920, 5 for the 950, and averaged the results (which is the last one in the row). The data below show top, medium, bottom bar data in that order.

What is the takeaway? PC/GAME mode doesn't seem to alter the reading the Leo Bodner device gets. Owners can game in ISF modes with no added lag, and possibly slightly less.

Personally I used my PS4 on Console input and used True Motion when playing Soma and I didn't feel any lag. I actually preferred strafing with True Motion on.

When I ran my PC through the 920 and 950 using a steam controller I could not feel any lag when moving the mouse. Some people have said mouse movement was 'swimmy'. I did not see this at all. However, it should be noted that the inputs were direct to the tv. No doubt there would be lag added if the signal goes through an AV receiver.

To me, the 920 and 950 felt the same when playing MGS V. I haven't tried much on the 950 yet, but Geometry Wars, and old school 2d shoot, felt perfectly playable on the 920.

920

PC

Game
47.2 46.6 56.6 52.1 50.625
51.1 50.3 60.5 56 54.475
54 54.2 64.2 59.8 58.05
ISF
47.1 55.7 58.8 57.2 54.7
50.9 59.6 62.7 61.1 58.575
54.8 63.3 66.5 64.9 62.375


HDMI

Game
50.6 52.9 46.2 51.5 50.3
54.4 56.7 50 55.2 54.075
58.1 60.4 53.7 59 57.8

ISF
53.7 46.1 46.4 56.6 50.7
57.4 49.8 50.1 60.2 54.375
61.2 53.6 54 64 58.2


HDMI as PC
Game
47 57.2 56.2 52.3 53.175
50.6 61 60 56 56.9
54.3 64.7 63.7 59.8 60.625

ISF
60.3 46.3 48.8 59 53.6
64.1 50.1 52.7 58.9 56.45
67.9 54 56.4 62.6 60.225

950
PC input settings
Game
61.1 56.7 57 49.8 60.1 56.94
65 60.5 60 53.6 63.8 60.58
68.7 64.3 64.5 57.3 67.6 64.48

ISF
48.5 54 45.8 47.6 50.2 49.22
52.1 57.9 53.4 51.3 53.8 53.7
55.9 61.6 57.1 55.1 57.6 57.46

HDMI setting
Game
49.9 48.8 49.7 54.5 51.4 50.86
53.6 52.5 53.6 58.2 55.1 54.6
57.3 56.2 62.6 63.3 61.6 60.2

ISF
48.6 46.5 55.1 55.8 54.2 52.04
52.2 50.2 58.7 59.6 57.8 55.7
56.1 54 62.6 63.3 61.6 59.52

HDMI input, but labelled PC
Game
45.7 56.2 61 47.4 43.5 50.76
49.6 60.1 62.8 51.1 47.1 54.14
53.2 63.7 68.5 54.9 50.9 58.24

ISF
49.2 46.7 56.5 52.4 55.8 52.12
53 50.4 50.2 56.1 59.5 53.84
56.7 54.1 53.9 59.7 63.3 57.54

No sound – highest was 69, lowest was 47, so seemed about the same
 

Magnesus

Standard Member
Thanks. Do you know if it delays the sound output to match the lag? 128ms difference beetween image and audio would be noticable.
 

JimBob1971

Well-known Member
Can't say at the moment I'm afraid. All anyone has to do is turn on True Motion and see if the to speakers sound matches the pic. This assumes no AV receiver in the chain. Any takers?
 

ashenfie

Well-known Member
It a menu option
I couldn't see how to disable sound. I tried by sending it to headphone rather than speakers. Anyone know how to fully disable sound?
Its a menu option on the TV.
 

Rogmeister

Active Member
got to be honest mate but i cant game on ISF mode. Something on it causes lag on my PS4. The simple check for me is the Playstation menu bar. flicking it left and right has noticible lag on ISF mode that i dont get on game mode.

wonder why you are getting different results.....
 

vaktmestern

Banned
I feel the isf lag vs game mode ...
 

JimBob1971

Well-known Member
Very strange. Are you guys running through a receiver? (Not that it should affect isf but not game) Definitely have just scan enabled , and everything turned off? I had the Bodner device running and went through all the picture modes and the Isf modes did not change real time lag at all (the number would be roughly the same as I cycled through the picture modes even though the picture obviously changed). I then altered one of the modes to have true motion and the lag increased to 120+. I then switched on a single picture mode and did numerous tests, only switching modes after all tests, and the figures were, on average, the same as was seen above. Enabling true motion absolutely crippled the lag figures though.

Perhaps the Leo Bodner device just doesn't work with these sets. Also the sample data I collected is quite small as you can see. It was pretty consistent though - nothing faster than 43ms and nothing slower that 69. I triple checked the results, making sure the Tv was in the mode I thought it was.

I wonder if lag on these sets is also dependant on what is in the screen? That would make the Bodner device useless for them.

I bought a hdmi splitter 1 input, 2 outputs to test this sort of thing on the 950 v 920 but unfortunately the 2nd output is about 100 ms behind the first output so isn't much use.

Anyhow, I'm giving Leo a shout to see if his device is accurate with 4K oleds. I've seen a number of sites using it, and if it isn't accurate for these sets then all the figures are off.
 
Last edited:

NBrice

Well-known Member
Very strange. Are you guys running through a receiver? (Not that it should affect isf but not game) Definitely have just scan enabled , and everything turned off? I had the Bodner device running and went through all the picture modes and the Isf modes did not change real time lag at all (the number would be roughly the same as I cycled through the picture modes even though the picture obviously changed). I then altered one of the modes to have true motion and the lag increased to 120+. I then switched on a single picture mode and did numerous tests, only switching modes after all tests, and the figures were, on average, the same as was seen above. Enabling true motion absolutely crippled the lag figures though.

Perhaps the Leo Bodner device just doesn't work with these sets. Also the sample data I collected is quite small as you can see. It was pretty consistent though - nothing faster than 43ms and nothing slower that 69. I triple checked the results, making sure the Tv was in the mode I thought it was.

I wonder if lag on these sets is also dependant on what is in the screen? That would make the Bodner device useless for them.

I bought a hdmi splitter 1 input, 2 outputs to test this sort of thing on the 950 v 920 but unfortunately the 2nd output is about 100 ms behind the first output so isn't much use.

Anyhow, I'm giving Leo a shout to see if his device is accurate with 4K oleds. I've seen a number of sites using it, and if it isn't accurate for these sets then all the figures are off.
Will be interesting to see the results of the 4k device when release. Wonder if bypassing the upscaler will lower lag, could be better for 4k PC gaming.
 

Rogmeister

Active Member
Very strange. Are you guys running through a receiver? (Not that it should affect isf but not game) Definitely have just scan enabled , and everything turned off? I had the Bodner device running and went through all the picture modes and the Isf modes did not change real time lag at all (the number would be roughly the same as I cycled through the picture modes even though the picture obviously changed). I then altered one of the modes to have true motion and the lag increased to 120+. I then switched on a single picture mode and did numerous tests, only switching modes after all tests, and the figures were, on average, the same as was seen above. Enabling true motion absolutely crippled the lag figures though.

Perhaps the Leo Bodner device just doesn't work with these sets. Also the sample data I collected is quite small as you can see. It was pretty consistent though - nothing faster than 43ms and nothing slower that 69. I triple checked the results, making sure the Tv was in the mode I thought it was.

I wonder if lag on these sets is also dependant on what is in the screen? That would make the Bodner device useless for them.

I bought a hdmi splitter 1 input, 2 outputs to test this sort of thing on the 950 v 920 but unfortunately the 2nd output is about 100 ms behind the first output so isn't much use.

Anyhow, I'm giving Leo a shout to see if his device is accurate with 4K oleds. I've seen a number of sites using it, and if it isn't accurate for these sets then all the figures are off.

No matey my PS4 is direct into the TV. From reading HDTv test, he had a massive difference between the two inputs on the 960. I doubt this would have changed much for the new TV's


Gaming responsiveness wasn’t the quickest on the LG 55EG960V, but it’s not the slowest either: our Leo Bodnar input lag tester returned a figure of 49ms in [Game] mode. We suspect that for most users, it’ll be perfectly playable especially when the mesmerising contrast, vibrant colours and clean motion (LCD panels are prone to black smearing and overdrive errors) are taken into account.

http://www.*/image/uhdtv/LG-55EG960V/lag.jpg

Labelling the HDMI input as [PC] did not reduce input lag any further in [Game] mode; instead this renaming increased lag slightly to 55ms, which is still an improvement over the 108ms measured in [ISF Expert] mode.
 

JimBob1971

Well-known Member
Hi Rog,

Yeah, I'd read that review prior to doing the readings. I'm just not seeing those figures. Incidentally I receive a reply from the Bodner folk. I'll retest in low light when I get a chance:-

--------
Thank you for your enquiry.

Yes the device will provide accurate results for a 4K TV that is being given a 1080p signal.

Fluctuating results can be caused by a timing difference in the back light and screen refresh rate. In this instance we recommend using the lowest value. For best results we also recommend setting the screen to the highest brightness and reduce ambient lighting where possible to ensure the most accurate readings.

The device detects the signal being emitted from the screen exactly as the human eye would see it. If each of the picture modes gave the same result then it shows the lag is not effected by this on the screens you are using. The lag increase caused by the TrueMotion feature is what I would expect to see.
--------

So taking the middle bar's lowest figure, we are talking 49 or 50.

I tried to reproduce lag on the PS4 last night by flicking the games bar as you said. Maybe I'm getting old, or there is something amiss with the telly, but Game and ISF modes performed so close that I couldn't feel any lag. However, with Truemotion on it felt a tiny bit slower, but not so much that I'd notice. Maybe I'm just not susceptible to it, or I'm being 44 I'm slowing down. I haven't played CS in a number of years. I used to be a demon, I suspect I'd get my bum kicked these days!

Out of interest, what are you guys getting for this?
Human Benchmark - Reaction Time Test
:D
I'm getting 209 on my Mac, so my brain maybe runs too slow to notice!

Anyway, as mentioned I'll test again soon in a dark environment.
 
Last edited:

JimBob1971

Well-known Member
For comparisons sake, here are my input lag measurements of the EG960V:
LG EG960V owners and discussion thread
LG EG960V owners and discussion thread
Similar to the 920 and 950, isf1 doesn't really do any worse than game mode.
Cheers Yappa. Yes those look very close to my figures.

Our figures don't really explain why some people are experiencing greater lag in ISF mode. (Unless they have true motion on). I'm beginning to wonder if how 'busy' the picture is affects the lag in certain modes. The tester works by displaying a few boxes and text on an otherwise empty screen. Perhaps the processing in the tv can keep up no sweat with this, but can't when a lot is going on. Speculation on my part but I'm intrigued. I have no idea how these sets synchronise the picture. I sometimes get the feeling I'm seeing strange tearing effects on certain scenes in films. I wonder if this is a clue, or whether I'm just misinterpreting normal judder.

Of course, it might just be a case of Occams razer and people reporting more lag in ISF are wrong. (Or that the test doesn't work on these sets). Beats me.

When I get a chance I'll turn on all processing except true motion and see how that effects measurements.
 
Last edited:

The latest video from AVForums

Podcast: Q Acoustics Q3030i, Humax Aura, Roku Streambar & WandaVision Reviews and more...
Top Bottom