AVS Forum banner

241 - 260 of 3759 Posts

·
Registered
Joined
·
79 Posts
Interesting results, thanks for sharing. Can you also share the report for this set with hand calibration?

What's the reason you suggest it is good for post production but not for home use? Seems that if the reason not to use it is accuracy then there is even less reason to use it for post production facilities?

Did you see this from Steve @Light Illusion about the poor stability / drifting of WRGB OLED? Were you doing anything to try to counteract the panel drift / stability?
https://www.lightillusion.com/forums/index.php?action=vthread&forum=8&topic=485#msg2829

The drift chart is pretty shocking. It looks like LG must have some compensation built in which has "steps" causing the sawtooth drift chart behaviour. Pretty ugly.

@bobof check these two pdf documents attached to the previous post. There are measurements of all four modes (hand/irp/lightning/matrix).

I suggest it could be used for post production because it's normal for me to spend 6+ hours on one display in studio to make f.eg. 10 000 points IRP LUT. I can cal other displays in the facility during the process. And one picture mode (Rec 709 or DCI) is usually enought. Typical customer wants to have day / night / HDR modes). That would last forever.

I followed everything that was suggested in Calman workflow, so 30/5 pattern insertion during greyscale and 10/5 during 3D LUT. I know these displays are instable as hell. There are ways to deal with it though. I did some calibrations for movie studios before, to make the display consumer reference, and match studio grade OLED as good as possible. It worked fine, at least the colorist was satisfied. Here's drift chart of OLED E7 i calibrated for German studio:



The pdf report of this cal is attached below. It looks better than IRP in C8 and was calibrated faster. It's of course different case. I used LightSpace and AJA LUT-Box. There is minor oversaturation of cyan colors, but only the dark ones. Red's are on thei'r places and there is no banding. It matched PVM-A250 very well.

My solution to overcome drift (using LightSpace) is to put the display into PC mode, and drive it native resolution (3840x2160p60, RGB) then you can measure with 0 extra delay. I also used drift compensation (1 white frame every 100 patches). Newest LightSpace releases have option to insert black frame to cool the display down. I haven't tried that yet. Calman is different beast and has other options. I have no idea which one is better since Calman can't plot drift chart. I can only evaluate calibration results.

First of all thx for sharing your thoughts and results with us. Regarding to your reports and visual tests these aren't good news at all.

Did you use the pattern Inseration mode? And why didn't you run a 42p Grayscale run to improve the low IRE?

Furthermore did you also tested a iRP LUT with less points
 

Attachments

·
Registered
Joined
·
6,804 Posts
I've seen similar behavior calibrating the eeColor for the LG OLEDs, it's critical to have good drift compensation. Whether it's that or a combination of that and the profile fitting and interpolation, CalMAN consistently underperforms both LightSpace and ArgyllCMS, but at a level that should not be visibly noticeable. With well designed patch sets there is no need to perform more than 2500 measurements to get the eeColor correction to < 1dE00 for 99% of large random verification test sets. That should hold for 3dLUTs at coarser spacing as well. BTW, lightning LUT will work best if you calibrate the white point and primaries at 100%/100% prior to doing the LUT.
 

·
Registered
Joined
·
588 Posts
BTW, lightning LUT will work best if you calibrate the white point and primaries at 100%/100% prior to doing the LUT.

That's interesting because CalMAN suggest only to set OLED Light to the desired luminance output and the hit AutoCal without any prior 2p adjustments and pre cal for the primaries.

If I get this right then you can reach better results if you calibrate the primaries before and do a 2p adjustment?
 

·
Registered
Joined
·
6,804 Posts
That's interesting because CalMAN suggest only to set OLED Light to the desired luminance output and the hit AutoCal without any prior 2p adjustments and pre cal for the primaries.

If I get this right then you can reach better results if you calibrate the primaries before and do a 2p adjustment?
You should always calibrate the 100% white point prior to any type of LUT profiling, this will preserve your desired peak luminance if you choose to calibrate the 16-235 range which I suggest. If you let CalMAN calibrate the "whiter than white" range you'll end up with a somewhat reduced peak white. At least that is the way it works with the eeColor, I don't know if the 2018 internal LUT has control points between 235-254. The lightning LUT works best for linearizing the fully saturated luminance channels so if you pre-calibrate primary x,y locations you should end up with a better result.
 

·
Registered
Joined
·
588 Posts
You should always calibrate the 100% white point prior to any type of LUT profiling, this will preserve your desired peak luminance if you choose to calibrate the 16-235 range which I suggest. If you let CalMAN calibrate the "whiter than white" range you'll end up with a somewhat reduced peak white. At least that is the way it works with the eeColor, I don't know if the 2018 internal LUT has control points between 235-254. The lightning LUT works best for linearizing the fully saturated luminance channels so if you pre-calibrate primary x,y locations you should end up with a better result.


Thx zoyd I Just read that John for example didnt run a 2p pre adjustment and in the new LG workflow it's not recommended as well but I will give it a shot. But that makes absolutely sense what you say because after the 1D Cal CalMAN gives the opportunity to readjust the OLED light because it can happen what you mentioned the reduced light output.

Also I will try to pre adjust the primary colors is it possible to change the xy values via DDC?
@D-Nice do you release your review like you said this weekend?
 

·
Registered
Joined
·
6,577 Posts
Discussion Starter #247
^^^OK so you guys need to help me here.

1) I don't see a 2pt layout in the Calman autocal workflow. Yes you can manually adjust the 30 and 100% using the DDC controls but it looks like the minute you run autocal, it starts all over. Also last i looked, the first thing autocal does is set the white balance.

2) I also thought, and Tyler confirmed this, there is no DDC interface for the CMS or maybe the controls don't do anything?
Were you referencing Calman or another product?
 

·
Registered
Joined
·
2,243 Posts
^^^OK so you guys need to help me here.



1) I don't see a 2pt layout in the Calman autocal workflow. Yes you can manually adjust the 30 and 100% using the DDC controls but it looks like the minute you run autocal, it starts all over. Also last i looked, the first thing autocal does is set the white balance.



2) I also thought, and Tyler confirmed this, there is no DDC interface for the CMS or maybe the controls don't do anything?

Were you referencing Calman or another product?


I’m pretty sure he’s referencing an external lut like the eecolor box and not the 2018 oled. Because from what I remember that was the first step with using the eecolor, was to set white balance


Sent from my iPhone using Tapatalk
 

·
Registered
Joined
·
6,804 Posts
^^^OK so you guys need to help me here.

1) I don't see a 2pt layout in the Calman autocal workflow. Yes you can manually adjust the 30 and 100% using the DDC controls but it looks like the minute you run autocal, it starts all over. Also last i looked, the first thing autocal does is set the white balance.

2) I also thought, and Tyler confirmed this, there is no DDC interface for the CMS or maybe the controls don't do anything?
Were you referencing Calman or another product?
I was speaking in general terms that the 100% white point should be set manually rather than through the 3dLUT. If there is a 1dLUT that CalMAN uses to do this then that is probably fine and losing some peak luminance to balance the high end shouldn't be an issue since you've got lots of head room with the OLED light. It still might be worthwhile if you are looking for a quick calibration that beats a full manual one to do WRGB at 100%+lightning LUT if this is possible.
 

·
Registered
Joined
·
79 Posts
Don't do a 2pt adjustment, just set 100% white to D65 using the gain control.
It makes no sense IMO since Calman does 1D LUT first, so white balance and gamma is aligned before 3D LUT measurements and calculations.

I wouldn't adjust anything in the TV. This most probably will mess things up. I think 1D and 3D LUTs are processed by display's CPU before OSD settings. Pay attention that contrast and brightness must be set to default before any LUT measurements. This is to make 109% input signal level match 1023 code value on panel. If you f.eg. set contrast to lower value, it will misalign and pick different measurements.

Things are easy when we use external LUT device. Then we know the processing chain. In that case we don't know that. Somebody would have to upload engineering documents which are probably confidential.
 

·
Registered
Joined
·
6,577 Posts
Discussion Starter #251
So far i find for SDR the 42 put 1DLUT and the lightening LUT works best if you want a fairly fast calibration. I'm finding the extra low IRE calibration seems to work well to get the near black in line. Below is a screen shot of my A1E on top and the C8 on the bottom. This picture was taken with my iPhone so a lot of the detail is missing but it illustrates the point.

For HDR, 20pt native and Matrix LUT works best. I found artifacts and posterization doing the lightening LUT with HDR. Same for DV.

I'm hoping D-Nice has some recommendations using the iRP LUT. Last i heard he said 1100 points gave fairly good results. I tried it for SDR and while better, i wasn't impressed with the amount of time it took to complete.

But all this discussion is going to be for nothing if we don't get a backup process.
 

Attachments

·
Registered
Joined
·
588 Posts
For HDR, 20pt native and Matrix LUT works best. I found artifacts and posterization doing the lightening LUT with HDR. Same for DV.

Do you see any side effects when using 26pt or 42pt AutoCal for HDR?

John did you use any kind of pattern delay and low light handler settings?
 

·
Registered
Joined
·
6,577 Posts
Discussion Starter #253
They both work fine, just the 42 pt takes slightly longer and is probably slightly better for the small extra time.
I use a Klein K10A so i don't use the low light handler and i use the default 30/5/10% for SDR and 5/5/15% for HDR and DV.
 

·
Registered
Joined
·
4,291 Posts
So far i find for SDR the 42 put 1DLUT and the lightening LUT works best if you want a fairly fast calibration. I'm finding the extra low IRE calibration seems to work well to get the near black in line. Below is a screen shot of my A1E on top and the C8 on the bottom. This picture was taken with my iPhone so a lot of the detail is missing but it illustrates the point.



For HDR, 20pt native and Matrix LUT works best. I found artifacts and posterization doing the lightening LUT with HDR. Same for DV.



I'm hoping D-Nice has some recommendations using the iRP LUT. Last i heard he said 1100 points gave fairly good results. I tried it for SDR and while better, i wasn't impressed with the amount of time it took to complete.



But all this discussion is going to be for nothing if we don't get a backup process.


Keep in mind that the target audience for most of these calibration features was Hollywood and post production facilities. LG wanted to give the same access to consumers as well. But fundamentally the LG implementation was designed around Hollywood’s needs.
 

·
Registered
Joined
·
588 Posts
I'm hoping D-Nice has some recommendations using the iRP LUT. Last i heard he said 1100 points gave fairly good results. I tried it for SDR and while better, i wasn't impressed with the amount of time it took to complete.

Do you have charts of the skin tones comparing lightning vs. 1100 iRP? Because the results of the lightning LUT from HDTV Polska didn't satisfied me.
 

·
Registered
Joined
·
6,577 Posts
Discussion Starter #256 (Edited)
^^^
No i don't but as i explained to D-Nice and Tyler, after i calibrated my C8 with Autocal with 42pt and Lightening LUT, in viewing content, to my eyes which are good but not those "magic" ones, the gray scale, colors, skin tones, etc, look essentially the same as my A1E that i calibrated manually and has a very good CMS.

Normally when you compare any two calibrated sets side by side sometimes you can see the secondaries like Cyan and or Magenta having issues and be different but the only thing i can see when looking very hard are some shades of Blue being very slightly different and that may be the Sony since you can't adjust the CMS.

That said, if we had a backup process, in my opinion, autocal is probably good enough for a consumer set calibration since if you set it up and run it properly, you will get pretty consistent results once you get experienced with it. The pro in my opinion is the 42 pt grayscale calibration which you can't do manually without Autocal and or the DDC controls, but I agree with some that you can probably do the same or better calibration manually if you are experienced. It's really a toss up if you look at it soley for a consumer calibration and you are not going to stress over every minute detail that may or may not be visible in content.

The other thing that's very apparent, if you are a newbie and or a home enthusiast and have all of the equipment and don't want to spend endless hours experimenting and calibrating your set, Autocal, once you understand how to set it up and run it, will make it very easy to "essentially" get "professional" or near "professional" results. Not refernce monitor or other tweaks that an experienced Pro might implement when calibrating to get an even a better result.

Another thing that came to mind, since autocal calibrates more of the lower IRE's, while a meter like a Klein and or other professional meter will give your superior results, i'm not sure that a meter like a C6 or other types of meters in that price range will do the same. Someone will have to test this as well. For sure it will take longer with a less capable meter but the Klein is the "king" of low light level calibration which may be another reason why the Autocal grayscale calibration looks so good to me.
Again i don't want to get into an arguement, and i don't believe Autocal replaces the experience Pro for many reasons, but these are my opinions based on my testing and what i observed.

Here are a couple of screen shots showing what i mean. Top is my A1E bottom is my Autocaled C8 with 1080P BlueRay content. I used my iPhone camera so some of the detail is lost but the colors and grayscale looks like what i saw by eye. If you zoom in and look closely you will see the Sony has a sharper image and more detail, again in person this is very noticeable vs the iPhone images below.
 

Attachments

·
Registered
Joined
·
588 Posts
^^^

No i don't but as i explained to D-Nice and Tyler, after i calibrated my C8 with Autocal with 42pt and Lightening LUT, in viewing content, to my eyes which are good but not those "magic" ones, the gray scale, colors, skin tones, etc, look essentially the same as my A1E that i calibrated manually and has a very good CMS.



Normally when you compare any two calibrated sets side by side sometimes you can see the secondaries like Cyan and or Magenta having issues and be different but the only thing i can see when looking very hard are some shades of Blue being very slightly different and that may be the Sony since you can't adjust the CMS.



That said, if we had a backup process, in my opinion, autocal is probably good enough for a consumer set calibration since if you set it up and run it properly, you will get pretty consistent results once you get experienced with it. The pro in my opinion is the 42 pt grayscale calibration which you can't do manually without Autocal and or the DDC controls, but I agree with some that you can probably do the same or better calibration manually if you are experienced. It's really a toss up if you look at it soley for a consumer calibration and you are not going to stress over every minute detail that may or may not be visible in content.



The other thing that's very apparent, if you are a newbie and or a home enthusiast and have all of the equipment and don't want to spend endless hours experimenting and calibrating your set, Autocal, once you understand how to set it up and run it, will make it very easy to "essentially" get "professional" or near "professional" results. Not refernce monitor or other tweaks that an experienced Pro might implement when calibrating to get an even a better result.



Again i don't want to get into an arguement, and i don't believe Autocal replaces the experience Pro for many reasons, but these are my opinions based on my testing and what i observed.



Here are a couple of screen shots showing what i mean. Top is my A1E bottom is my Autocaled C8 with 1080P BlueRay content. I used my iPhone camera so some of the detail is lost but the colors and grayscale looks like what i saw by eye. If you zoom in and look closely you will see the Sony has a sharper image and more detail, again in person this is very noticeable vs the iPhone images below.

That sounds much more optimistic thanks for sharing that with us and also the comparison between your A1 and the C8.

I hope I get my C8 this week so I can run some tests also.
 

·
Registered
Joined
·
6,577 Posts
Discussion Starter #258
^^^
You are welcome and remember, there will be other reports with more detail and more comparisons to what we call the "reference" monitor and or reference results. Clearly, my report is from what we call the "casual observers" point of view, meaning what does the average and or discerning consumer or enthusiast see when looking at content at the end of the day with this technology. :)

And as Tyler indicated, if you have the time and execute Autocal with the large iRP LUT process, you may get closer to "reference" quality but i'll leave it up to the pros who calibrate in that space make that determination. :)

Please keep this in mind. :)
 

·
Registered
Joined
·
4,897 Posts
...That said, if we had a backup process, in my opinion, autocal is probably good enough for a consumer set calibration since if you set it up and run it properly, you will get pretty consistent results once you get experienced with it...
John, if by "backup process" you mean a copy of the LUT that is created, CalMAN does that already, at least when working with external LUT boxes. If you mean a backup of the existing pre-calibration settings in the TV, no. Not even when working with DDC-only sets like the Samsungs.
 

·
Registered
Joined
·
6,577 Posts
Discussion Starter #260
Right so since you can't give the calibration settings to a customer like you can with the regular user controls, i mean being able to Auto-calibrate a set for a customer and store the calibration on a thumb drive that is encrypted so they can recover the calibration if the set was intentionally or un-intentionally set to factory default.

The problem is, most of the time, if you call LG support and they are unsure of the state of the TV, they will want to reset it to factory default and troubleshoot from there. Once that's done your autocal is gone.
 
241 - 260 of 3759 Posts
Top