AVS Forum banner

801 - 820 of 2005 Posts

·
Registered
Joined
·
366 Posts
Something that was mentioned got me thinking about why would LG make wide color gamut the default for HD.

Has anyone checked the horizontal/vertical ramps when using wide color gamut as their starting point for calibrating?
I got some time in this week and wanted to follow up.
The ramps have the same characteristics, relatively, as I previously described whilst uncalibrated.

Gamut option: Wide
Holy crap this is indeed WIDE. It's basically DCI-P3, or close even if using 1.9 gamma. (pictures of wide in next post)
Honestly, I don't see a reason to ever use this option.
Noticing this I had a silly thought regarding CMS controls (see one of my recent posts if interested, it's a stupid silly theory).

Gamut option: Extended
This gives my E6 a very minor saturation boost and slight hue difference.
I'm probably going to be using this to counter balance the saturation loss from using 1.9 gamma (only way to keep 0% black and problem free yet still see 0.5%).


***EDIT***
Just wanted to clarify, that this is based on MY E6, and as such it may or may not behave the same as any other E6, let alone any other model.
 

·
Registered
Joined
·
366 Posts
When you want to measure the DCI-P3 in SDR, you have to measure based as target to DCI-P3 colorspace, not DCI-P3 inside a REC.2020.

The option to select DCI-P3 target colorspace is not available to consumer colorspaces of CalMAN, you need a Business license to select it since DCI-P3 is not a consumer colorspace.

But you can create a custom colorspace using the DCI-P3 coordinates using a consumer CalMAN license if you like.

This is why your measurements are so off. Here I have posted the correct measurements, the target colorspace was the REC.709 to these pictures. ;)
I've just woken up, but I don't think I follow what you're saying.

Perhaps I'm just not saying the right words for what I want to express, so here are pictures. Hope you can correct me if what I'm trying to say is wrong (and it probably is lol).

These were taken in SDR, using your patterns on your disc, specifically the ChromaPure 09. Color Gamut (100% Saturation - 100% Intensity) set.
For what it's worth, using Masciolas' 100% Amplitude & 100% Saturation patterns, with HDR metadata stripped (or not triggered via PC) has the same result. I don't know if it should be or not, just thought I'd mention it.
The screenshots are from USB media player on the display set to:
OLED Light: 40
Contrast: 87
Brightness: 60
Color: 50
Gamut: WIDE
Gamma: 2.4
No white balance adjustments
No CMS adjustments
*Note: If PS3/PS4 is set to similar settings for LIMITED rgb, the results are very similar.
Measurements taken with HCFR.
All I'm changing is what HCFR is set to (709, DCI-P3 and P3 in 2020) and making a statement about the coverage of my measurements relative to what HCFR was set to.

[email protected]
http://www.avsforum.com/forum/attachment.php?attachmentid=1943977&d=1486242077

[email protected]
http://www.avsforum.com/forum/attachment.php?attachmentid=1943985&d=1486242357

[email protected] in 2020
http://www.avsforum.com/forum/attachment.php?attachmentid=1943993&d=1486242368


As you can see, the WIDE measurements cover alot of DCI-P3, and the vector of green primary is aligned with P3 in 2020, so I was considering this P3 in 2020.
Of course trying to measure P3 in 2020 patterns with metadata stripped doesn't have the same results (p3 green when measured like this was somewhere between yellow and green in DCI-P3 mode)
 

Attachments

·
Registered
Joined
·
8,913 Posts
I've just woken up, but I don't think I follow what you're saying.

Perhaps I'm just not saying the right words for what I want to express, so here are pictures. Hope you can correct me if what I'm trying to say is wrong (and it probably is lol).

These were taken in SDR, using your patterns on your disc, specifically the ChromaPure 09. Color Gamut (100% Saturation - 100% Intensity) set.
For what it's worth, using Masciolas' 100% Amplitude & 100% Saturation patterns, with HDR metadata stripped (or not triggered via PC) has the same result. I don't know if it should be or not, just thought I'd mention it.
The screenshots are from USB media player on the display set to:
OLED Light: 40
Contrast: 87
Brightness: 60
Color: 50
Gamut: WIDE
Gamma: 2.4
No white balance adjustments
No CMS adjustments
*Note: If PS3/PS4 is set to similar settings for LIMITED rgb, the results are very similar.
Measurements taken with HCFR.
All I'm changing is what HCFR is set to (709, DCI-P3 and P3 in 2020) and making a statement about the coverage of my measurements relative to what HCFR was set to.

[email protected]
http://www.avsforum.com/forum/attachment.php?attachmentid=1943977&d=1486242077

[email protected]
http://www.avsforum.com/forum/attachment.php?attachmentid=1943985&d=1486242357

[email protected] in 2020
http://www.avsforum.com/forum/attachment.php?attachmentid=1943993&d=1486242368


As you can see, the WIDE measurements cover alot of DCI-P3, and the vector of green primary is aligned with P3 in 2020, so I was considering this P3 in 2020.
Of course trying to measure P3 in 2020 patterns with metadata stripped doesn't have the same results (p3 green when measured like this was somewhere between yellow and green in DCI-P3 mode)
Yeah, as I posted, the Wide is for DCI-P3 usage, because they always use 3D LUT for calibration in post-production, so they select Wide and they are doing the display characterization for DCI-P3 as target colorspace.

About your measurements, you see the same primaries when you look DCI-P3 as target colorspace vs. DCI-P3 inside REC.2020 because you are using wrong patterns to evaluate the DCI-P3 inside REC.2020.

The patterns you are using have 100% Saturation with 100% Stimulus level, for example 100% Red has RGB Triplet 235.16.16 (aka Red 100% . Green 0% . Blue 0% in video levels)

When you want to validate measuring 100% Saturation with 100% Stimulus level Red of DCI-P3 inside REC2020, you have to display a pattern with RGB Triplet 222.178.174 (I'm using CalMAN's RGB Targets REC.2020 SDR for this example)...so you de-saturate as you see a lot to reach the 100% Red of DCI inside 2020, the 235.16.16 is the 100% Red of REC.2020.

When you want to display 100% Saturation with 100% Stimulus level, for example 100% Red for DCI-P3 you have to display RGB Triplet 235.16.16.

The workflow you see at DCI-P3 inside REC.2020 is expecting 50% Stimulus pattern also, so the RGB triplet is different for 100% Red with 50% Stimulus, it's: 119.97.95
 

·
Registered
Joined
·
366 Posts
Yeah, as I posted, the Wide is for DCI-P3 usage, because they always use 3D LUT for calibration in post-production, so they select Wide and they are doing the display characterization for DCI-P3 as target colorspace.

About your measurements, you see the same primaries when you look DCI-P3 as target colorspace vs. DCI-P3 inside REC.2020 because you are using wrong patterns to evaluate the DCI-P3 inside REC.2020.

The patterns you are using have 100% Saturation with 100% Stimulus level, for example 100% Red has RGB Triplet 235.16.16 (aka Red 100% . Green 0% . Blue 0% in video levels)

When you want to validate measuring 100% Saturation with 100% Stimulus level Red of DCI-P3 inside REC2020, you have to display a pattern with RGB Triplet 222.178.174 (I'm using CalMAN's RGB Targets REC.2020 SDR for this example)...so you de-saturate as you see a lot to reach the 100% Red of DCI inside 2020, the 235.16.16 is the 100% Red of REC.2020.

When you want to display 100% Saturation with 100% Stimulus level, for example 100% Red for DCI-P3 you have to display RGB Triplet 235.16.16.

The workflow you see at DCI-P3 inside REC.2020 is expecting 50% Stimulus pattern also, so the RGB triplet is different for 100% Red with 50% Stimulus, it's: 119.97.95
I was just going by coverage. I assumed something else was at play here due to the luminance targets being way off.

I assume 2020 should also be using 50% stimulus patterns, then?

Actually, I'm interested in why less than 100% stimulus, or rather why a combination of saturation and stimulus patterns are not a typical procedure.
I'm assuming that a sweep with a mix of saturation and stimulus would be a better metric for replicating what we see in the real world. (i.e. natural, not synthetic colors ala CGI or post processing).
*making mental note for further research*

I appreciate the info.
 

·
Registered
Joined
·
366 Posts
But I'm curious...
How do you all feel about HDR shadow detail?
Making 0.5-1.5% visible?
Go for the "official" calibration of setting control point 127 to 0.14 nits? (a 7.2% pattern)
Adjust a normal 5% pattern to 0.061 nits?
Adjust a normal 10% pattern to 0.328 nits?
I'm just kind of going by raising as low as possible raising black and without crushing peak luminance (~700 nits IIRC)
Adding on to this, has anyone tried forgoing "gamma" all together by opting to use L* or some similar logarithmic grayscale (HLG?)
Obviously for movie content it wouldn't be correct, but what about video games?

I may try one of those for a comparison because why not, but I don't quite understand the intentional differences (one is purely relative, the other being log-scalable for HDR).
 

·
Registered
Joined
·
8,913 Posts
I assume 2020 should also be using 50% stimulus patterns, then?

Actually, I'm interested in why less than 100% stimulus, or rather why a combination of saturation and stimulus patterns are not a typical procedure.
I'm assuming that a sweep with a mix of saturation and stimulus would be a better metric for replicating what we see in the real world. (i.e. natural, not synthetic colors ala CGI or post processing).
*making mental note for further research*

I appreciate the info.
REC.2020 HDR in HCFR requires 50% Stimulus, if you see, it says 126.16.16 for Red @ the pattern window.

Using ST.2084 PQ (REC.2020 HDR):

100% Stimulus, for White is 10000 nits
50% Stimulus, for White is 92.24 nits
 

·
Registered
Joined
·
380 Posts
Procedure to create a PC input

Over in the HCFR thread, @harlekin said if you rename an HDMI input as PC, you'd get extra ISF memories. I can't get it to work.

Can someone please give me the procedure?

Thanks in advance...
Ron
 

·
Registered
Joined
·
792 Posts
Over in the HCFR thread, @harlekin said if you rename an HDMI input as PC, you'd get extra ISF memories. I can't get it to work.

Can someone please give me the procedure?

Thanks in advance...
Ron
The procedure? So you give no error report ("It doesn't work."), then assert something that I probably haven't said exactly that way, then you want a step by step - for basically the most blatent tip there is - as a reward -

Here's your step by setp - figure out how it works.
There are two different ways of "renaming" an input (on the same menu screen) use the one where you select input names from a drop down list - PC is already there.
This gets you an entire new set of memory banks for that input.

Also in "PC mode" a few settings like Motion compensation (afair) are greyed out.

If you have used "apply to all inputs" in the past - don't be surprised, that it did its job.

Oh, and btw - I really don't like doing other peoples basic logic for them. It feels like they don't care - and would rather have all questions answered to them on their smartphone vie a textmessage service. WHich is exactly how most people use forums these days.

Also - this is a RTFM question to a certain extent (of course there is no manual for LG TVs anymore - because people in mass would rather flock to the internet), thats a support question to a certain extent, so why don't you ask them - their phone number is in your TV, I think that step by step guide requests should be banned on most forums in general, I don't like to do a manufacturers support work for them... I also don't like that for all the work I put into this thread - the freaking "settings tip" is what I get remembered for. It feels like people just want to support and consume these days - and have no interest in contributing even the slightest cerebral effort even for their own problem solving.

And still - you win - because the vague hint that "what he said doesn't work" is enough of a distribution of fake facts - that I have to intervene - if I care about that.

Think about that a little - there is some sound logic in here about what "support requests" should be and what they really turned out to on the internet today.

"Can someone plese write me a short manual for my issue?!"

Please?
 

·
Registered
Joined
·
380 Posts
The procedure? So you give no error report ("It doesn't work."), then assert something that I probably haven't said exactly that way, then you want a step by step - for basically the most blatent tip there is - as a reward -

Here's your step by setp - figure out how it works.
There are two different ways of "renaming" an input (on the same menu screen) use the one where you select input names from a drop down list - PC is already there.
This gets you an entire new set of memory banks for that input.

Also in "PC mode" a few settings like Motion compensation (afair) are greyed out.

If you have used "apply to all inputs" in the past - don't be surprised, that it did its job.

Oh, and btw - I really don't like doing other peoples basic logic for them. It feels like they don't care - and would rather have all questions answered to them on their smartphone vie a textmessage service. WHich is exactly how most people use forums these days.

Also - this is a RTFM question to a certain extent (of course there is no manual for LG TVs anymore - because people in mass would rather flock to the internet), thats a support question to a certain extent, so why don't you ask them - their phone number is in your TV, I think that step by step guide requests should be banned on most forums in general, I don't like to do a manufacturers support work for them... I also don't like that for all the work I put into this thread - the freaking "settings tip" is what I get remembered for. It feels like people just want to support and consume these days - and have no interest in contributing even the slightest cerebral effort even for their own problem solving.

And still - you win - because the vague hint that "what he said doesn't work" is enough of a distribution of fake facts - that I have to intervene - if I care about that.

Think about that a little - there is some sound logic in here about what "support requests" should be and what they really turned out to on the internet today.

"Can someone plese write me a short manual for my issue?!"

Please?

@harlekin,

I meant no disrespect in my post, nor did I mean to imply that your suggestion in the other thread was incorrect: I put your name in there because I wanted you to get the credit for a valuable suggestion to me. I also didn't intend that anyone should have to spend a lot of time in a response.

When I got the set, most of the modes had default settings, so I was assuming that changing the mode to 'PC' would bring along some defaults. I had done the steps you wrote in your response, but I didn't notice that anything had changed. When I didn't notice any changes, I assumed that I was doing something wrong. Looking back now, this would have all been useful diagnostics when I asked for help: this is all on me.

I will continue to ask for help, but hopefully I will do a better job.
Ron
 

·
Registered
Joined
·
366 Posts
I will continue to ask for help, but hopefully I will do a better job.
Ron
Yes, please do. This is a community of very helpful and friendly people. This includes Harlekin too, although a lot of his (her?) posts do rub some people the wrong way. I mean if he (she?) didn't want to help in some way, the response would have been quite different (and probably shorter ;)). I won't say more about this, and I hope no one responds directly to this part of my post.


To add to what Harlekin said: another, and IMO better, way you can have more ISF setting slots is by switching HDMI cables as each input slot has their own storage for separate settings.
For each input slot, you can assign an "Icon mode". These are in the "All Inputs" screen, then you hit the "Edit Icon" button, and not by editing the "Input Label" (you can rename it after changing the Icon).
Each Icon mode stores it's Picture mode setting separately. For example, you can customize the Cinema preset differently for each Icon mode.
ISF settings are an exception, as they are shared for all Icon modes.
HOWEVER, the sole exception to this is the PC Icon mode, which is isolated. This also means the "Apply to All Inputs" button will NOT transfer from between PC and all other Icon modes.
What makes this Icon mode special is that it's the only way to get 4:4:4 chroma, no other Icon mode supports it. The rest are all 4:2:2 or 4:2:0. I don't remember since I don't care about it. Chroma is not the only difference, you loose some options like motion interpolation (TruMotion), Real Cinema, Noise Reduction, no "HDR Effect" picture mode (fake HDR). There may be different signal processing like RGB to/from YCbCr, sharpness algorithm, framerate cadance, and possibly others, but I don't have the tools and experience to investigate.
Aside from chroma, I don't think there would be any perceptible differences, so if you need 4:4:4 chroma, you're stuck with two ISF settings if you don't want to swap cables all the time. If your sources are only 4:2:0 chroma, like movies and not PC/Consoles, then no worries.

For gaming, every picture mode in the PC Icon mode has low latency, so you can game while retaining your calibration settings via ISF modes. The difference between the "Gaming" picture mode (on all Icon modes) and all picture modes while using the PC Icon mode is minimal. The former is lower, but IMO it's a insignificant (old school twitch shooter gamer ala quake/team fortress, so I'm sensitive to latency).
 

·
Registered
Joined
·
792 Posts
His.:)

Everyone is invited and even encouraged to ask for help. Spoken as someone with no real influence in this forum whatsoever.. :) Just be at least a little mindful when asking rtFm (capitalized F for emphasis) questions and try not to ask for step by steps if you can help it.

How someone chooses to help you, is in their control - not in yours. If they spend 5 minutes, they spend five minutes. If they spend 20 minutes, they spend 20. If you need to ask for step by steps - there will be far less people willing to attend to your needs - for a very good reason, there is a difference between asking or information and asking for a service.

If you "bait" an answer to a service request with a suggestion that something you have read could have been wrong (even if it is implied) - and then set an @mention -- understand that this can very well be read as creating work for someone else - or even as adding uncertainty and doubt to a statement or tip, so others are less likely to try it.

So - when I get notified an log in, my mood already isn't exactly at a high point. Then I look at the question thats asked and it is essentially "where do I change the input names on my TV" - respectively how to use the extra memory slots - but thats unclear because of how openly the question was asked.

Then add to that the notion that I am carrying with me - that following procedures blindly is responsible for at least three major issues this scene currently faces (- away from ;) )- and approximate the amount of joy it brings me to add to that and do the product support job of explaining to you where to find a setting I've outlined before in your TV.

Information that gets collected, processed, shared, tried out and used by different people is a good thing. Asking for personalized help services is not.

The line isnt always as clear - and different people will certainly not react as harshly as I tend to - but, there is something said for having to earn your spurs, by showing that you care more about the stuff in here than only up to the conclusion of your immediate issue. And that includes trying to figure stuff out on your own. If you fail - explain how, and people will want to help.

But "it doesn't work" is about as inciteful as a sentient potatoe skin.

Once you can show that you've done the basic trial an error stuff on your own - people tend to be more forthcoming and helpful - especially if you can present them with interesting "problems" to solve.

Thats how this stuff works on a social level. Regardless how forums are set up to attract users "this is a place where you get helped" - doesn't work for self maintaining communities.

(It only works for places - that have PR people roaming around as part of their job description, because they can do product support that scales a little better and they attract new customers that way -- which is how most platforms work today. "Hey, I am X, I have a problem, can you solve it for me - I bought product. Thanks. xoxo" And I have to say I dont react all that well to that approach of "community". Maybe because I've experienced the days where the internet wasn't primarily driven by simply generating ad impressions.)

Or in short - people that grew up on facebook never learned "netiquette". On facebook theres only polite conversation, likes, personas, "packaged outrage", and every news item is "interesting", because it is from a "friend". On the rest of the internet people tend to operate a little differently.

When a geek reacts harshly to a certain request - its not necessarily because we are socially inept or bad people - its because people tend to abuse us to help them understand how "that thing they wan't to do works" - an then buzz off - with only a thank you falling from their lips.

If you have experienced this 10, 20 times you tend to wan't to shield your communities against that behavior.

On reddit or facebook you are usually encouraged to give your opinion or ask questions right away - but thats usually not where you find expertise if you think about it. (Because channels being flooded by support requests and long form answers being pushed away by more support requests, from people that can't even be bothered to read stickies, or faqs or manuals, or use search, makes the whole experience an echochamber, where only the most baseline hearsay knowledge gets shared between people that will always tend to tell you what they have heard - fast.)
 

·
Registered
Joined
·
366 Posts
No offense harlekin, but let's get back on topic before mods come back ;)

I'd like to share more about some quite interesting behavior that I've observed with my E6.

Recap:
We know HDR 20-point white balance calibration controls are funky right out of the box. The white end of the controls don't align with their matching patterns. This is nothing knew, there are other displays that have this problem so this is not a "well HDR is new" issue.
We know CMS controls are funky. Control adjustments have less coverage, becoming weaker the farther a color is from the primary/secondary vectors. Too much adjustment causes bit overflow which makes artifacts (ex: if the bit for +10 luminance was the max possible for a single pattern swatch, and going beyond that to +11 would cause it to go from 100% to 20% (total=120%). +13 might be 30% (total=130%). On another pattern the same +11 bit might have already exceeded 100% and now back up to only 70% (total=170%), +29 might even get back to 100% Numbers are for illustration to explain what I mean by overflow).

First, set 0% black so pixels are off in a dark room. Stick your eye on the screen while blocking all light except from the oled screen, like look through a cardboard toilet paper roll or something, and give your eyes a few minutes to adapt until you finally have true black.
Using the 20-point white balance controls, I have set the last control point to +50 red, -50 green, -50 blue. Second to last is set to -50 red, -50 green, +50 blue. Third from last, -50/-+50/-50, etc. You get the idea. Also, videos at the bottom show how brightness/contrast sliders effect this coverage for HDR (so far SDR doesn't seem to be as bad).
Then go to a grayscale ramp with steps. Take a picture or put a sticky note on the screen marking the center of the coverage area of each of these control points. I used sticky notes, so I'll just continue explaining it like this.
Next, go look at a pattern, like 80% (or HDR code value 606).
Is the pattern tinted with the +50 color you set that control point to?
Adjust contrast up and down, find the "middle" of the coverage area where it's most saturated with that color and write down what the contrast value is.
**NOTE: If at any point you wind up with a center point that lifts black, you need to start over and drop brightness by 1.
Go back to the grayscale ramp and you might find your control points have moved.
I left my original marks alone, and made a new one.
Now go to a different pattern (like 90%/653).
Check pattern for matching tint, adjust contrast to find center and write it down.
Once again, go back to the grayscale ramp.
Check what happened to the PREVIOUSLY marked pattern. For me it was neither at the starting point or my last adjusted point.
Continue doing this for all control points, not just the ones giving you trouble.
The position of the control points change as well, so even if you crushed/lifted enough to have the same colors at the same marks on the grayscale, checking the pattern for a mark like this will show you something different. :eek:

Once done, you now have a list of Contrast values, and all you need to do is find the average, but only the average of all troublesome points.
Remember to manually check each individual pattern to make sure it's being tinted to match the controls.

With all the sticky notes marking the original, ideal and best fit average for the control points, you can then have a better idea of how to weigh each measurement to compensate for each control point. If the current pattern position is below the ideal center, then measurements from that pattern need to be a little higher than the target your software says it should measure at. Further away = more compensation.




Regarding HDR controls, specifically:

Service menu settings, the InStart ones at least, are worth playing with.
Tthe "White Balance" controls are 2 point controls that work for HDR mode (yet effect all modes), while the 20point controls only works for SDR (they stack with the normal 20-point controls!). I've only done a quick check but these 20-point controls appear to have a better alignment; matching the patterns at various kinds of brightness/contrast values.
But the most important big winners are Sub Contrast and Sub Brightness.

I've mentioned many times that you can lower contrast, instead of oled light, to better follow the "official" HDR calibration procedures for the displays in the other calibration thread.
The big exception being the range of nits possible between the last two controls, codes 653 (480.85 nits) and 668 (540 nits). It's impossible on MY E6 to get more than a 20 nits difference. I've tried a few weird alternatives, like ignoring one of them, or setting absolute peak at 540 nits via a 100% 1k nits pattern (668 is not "peak"), or shifting the targets up one step, or even just letting peak go as high as possible and just balance r/g/b. None of them satisfied me, I thought they made the tone mapping look wrong because the peak roll off starts at control point 668. It's really noticeable in LG's HDR demo clips: NASA (crests of moon craters and solar flares blending near peak white) and Colors of Journey (lantern and campfire, driving out of the tunnel, clouds/sky behind ornate glass windows at 0:39 and 0:49, light bloom from candles and sparklers, glossy highlights and train lights).

If you do the above procedure to align white balance controls and patterns, you might notice the coverage areas of each control point change in width at various points on a grayscale ramp. Sometimes super thin so you can't see adjustments having an effect, and some times so wide a single control may effect multiple patterns!
On my E6, I could not get around this using only the brightness/contrast sliders. From my experience with my LM7600, I knew LG's Sub Brightness/Contrast service menu options changed how coarse or fine the brightness and contrast sliders are (and can be used to deal with a single primary color crushing earlier than others.
After adjusting Sub Brightness/Contrast, I was finally able to hit all targets outlined in the "official" HDR calibration procedure easily (and true peak remains >700 nits! :cool:)
Not only did this help HDR picture quality (matching "official" calibration targets let's the tone mapping work as intended), but is also greatly reduced the noise between 0-5% black, and slightly improve the color banding (the random differences in luminance for all colors not caused by grayscale banding).
Your experience may be different than mine, so if you try this be sure to share your opinion on it

Now to show how Sub Brightness/Contrast helps, I present three videos taken with my oh-so-amazingly-bestest-camera potato camera!
I recorded these originally to explain how to adjust the HDR 20-point white balance controls so they align with and adjust their appropriate patterns, only to find out that this process is applicable to SDR as well.


Adjusting Brightness & Contrast sliders does this:

How Sub Brightness/Contrast changes things (factory "HDR Standard" settings, [email protected], [email protected], [email protected]):

How Brightness & Contrast sliders look after adjusting Sub B/C:

Direct before/after comparison. Note where the controls coverage is (reference: the "1000 nits" step). Also their widths (My potato doesn't show it well, but each pair of "blue and green" have similar widths, where before one par was more blue and the other pair more green).
http://screenshotcomparison.com/comparison/199867

---
Default values of Sub B/C are both 128 (and is usually the same for all "smart" LG displays)
If you adjust White Balance service menu controls, the "Warm" color temperature is Warm2 in the normal options. You can customize Normal and Cool in the service menu, effectively giving you three kinds of Warm2 if you wanted :)
 

·
Registered
Joined
·
39 Posts
HDR calibration:
In recent days I had calibrated HDR, and I had arrived at a good point, the problem in the ED was but 'a luminance that uciva targhet from 45 to 65 ire, this TV does not have a luminance control, doing more tests I noticed that the command CONTRAST also acts on LUMINANCE curve, lowering it from 100 to 90 I improved all errors leading them to 3 DE or less.
I confirm that the CMS does not respond to the command, saturation, hue, luminance, they have no effect. CONTRAST command instead of the more or less saturated colors.
I used the Calman HDR workflow with its grayscale, I did a reading of the whole staircase, I did correction looking at the chart and based on LG checkpoints, made-scale reading, and corrected again, and so on until arrive at a good result .:

DEFAULT GREYSCALE WARM 2 :



CALIBRATED GREYSCALE , WARM 2 , OLEDLIGHT 100 , CONTRAST 100



CALIBRATED GREYSCALE, WARM 2 , OLEDLIGHT 100 , CONTRAST 90



So much for the gray scale, which as you can 'see you come to a good result, now I'll try' the best solution on the color.
 

·
Registered
Joined
·
11,696 Posts
  • Like
Reactions: jrref

·
Registered
Joined
·
380 Posts
Does PC mode only work on HDMI 1?

Following the tips in previous posts, I was successful in changing the icon to 'PC' on HDMI 1. I did see that several settings (RealCinema, TruMotion, etc) were greyed out, and the ISF memories were unique.

When I tried to replicate this on HDMI 2, I was unable to duplicate the results. After changing the icon to the PC, this time I saw nothing change. I saw no settings greyed out in Picture Settings. When I changed OLED Light to a different value, it persisted when I changed the icon back to the HDMI icon. Cinema settings are not unique, either.

I could not find anything in the LG support data base to address this. Is the unique PC mode only effective on HDMI 1?

Thank you in advance...
Ron
 

·
Registered
Joined
·
366 Posts
Following the tips in previous posts, I was successful in changing the icon to 'PC' on HDMI 1. I did see that several settings (RealCinema, TruMotion, etc) were greyed out, and the ISF memories were unique.

When I tried to replicate this on HDMI 2, I was unable to duplicate the results. After changing the icon to the PC, this time I saw nothing change. I saw no settings greyed out in Picture Settings. When I changed OLED Light to a different value, it persisted when I changed the icon back to the HDMI icon. Cinema settings are not unique, either.

I could not find anything in the LG support data base to address this. Is the unique PC mode only effective on HDMI 1?

Thank you in advance...
Ron
On my E6 I don't have that problem.

Perhaps deep color on\off may change this for you?

I seem to recall deep color not being a feature on all HDMI ports on the B6 or C6.
 

·
Registered
Joined
·
8,913 Posts
Setting 1.9 can expose "glowing black", someone of us don't care and use 1.9 without problem, other fix this glow lowering the 2 point cut.
For me, in dark room, with eye adjusted, lowering the Blue and the Green cut to -10 adjust this "glow", and a full field 16,16,16 appear the same of a power-off tv.
So, if exist 16 level on movie, this is pure inky black, if can't see this pure inky black, simple don't exist on the movie.

The 1.9 setting required to lowering luminance at 5ire, because at default it is very extreme.

Near black related gamma with 1.9 setting is the best near-black choice, because 2ire, 3ire,4ire,5ire (i don't have a Klein) measured near the same gamma point (with right adjustment).

Instead, gamma 2.2, 2.4 and bt1886 measure more closed on black in the first level (with the same 5ire gamma point).
Gamma 1.9 will boost the near black levels (1-5%) to be brighter from Gamma 2.4 targets, if you have uncalibrated the 5% Gray.

I have measured all these 3 modes near black performance with my Klein K-10A in a totally black environment (all devices LED's had been covered with Light-Dims stickers, notebook screen was off during measurements)

If you precalibrate the 5% Gray (validating using 2.4 as gamma target) for each of the available options (1.9/2.2/2.4) and then measure the near black performance (I used the Near Black layout page of my workflow and my calibration disk 6-Point Near Black Chapter which measures 100/5/4/3/2/1/0.5%), you will see that the differences are very small, the 1.9 provides the worst results, the 2.2/2.4 are very close, but seems there unit-to-unit variations there also, I don't have measurements of more different 65E6 units to know more, but these are my findings from the unit I measured only.
 

·
Registered
Joined
·
380 Posts
On my E6 I don't have that problem.

Perhaps deep color on\off may change this for you?

I seem to recall deep color not being a feature on all HDMI ports on the B6 or C6.
When I changed the input to HDMI 2 from interlaced to progressive, the PC mode worked as all of you have explained. When I switched it back to an interlaced signal, PC mode no longer worked.

Thank you all for the help...
Ron
 

·
Registered
Joined
·
216 Posts
A couple of interesting changes to the 2017 models:

"HDR now includes both 2-point and 20-point calibration menus, and the ‘brightness’ setting option has more granular controls, meaning that even though the steps closest to ‘50’ are still named 48, 49 as well as 51 and 52 – and so on – they actually represent 0.5 increments compared to last year’s “brightness” setting option. A new “Active HDR” system can analyze incoming HDR signals and apply dynamic metadata to the picture on a frame-by-frame basis, thus optimizing HDR picture quality, says LG."

source:
http://www.flatpanelshd.com/news.php?subaction=showfull&id=1484640539
 

·
Registered
Joined
·
72 Posts
My calibrator returned last night to attempt using the artificially defined black level of 0.0034 cd/m2 and gamma 2.35 with 170 cd/m2 peak brightness as suggested here on my B6. We couldn't work out how to produce the custom curve in Calman - can anyone advise how this is done?

Instead we used targets posted in this thread and there is definitely more detail in the black, but I'm not sure the overall picture is preferable for all material to a straight 2.4 calibration. It looks like it's best to have both configured and choose depending on content.

He also attempted HDR calibration again, and as recommended in this thread, the green emphasis was removed in low to mid tones until the controls stop working. OLED/Contrast were left at 100/100. Even though he had a Murideo, we used the Masciloa LG OLED images to calibrate as it was getting late and he wasnt' sure how to make the Murideo output the required patterns with LG metadata. With his Klein K-10A, brightness was around 770 nits with a 668 pattern, and the same at 1000nits. This seems some 230 nits brighter than LG state at 668. So there is no roll-off above 668 - just ~770 nits up to 1000. Does this sound right?
 
801 - 820 of 2005 Posts
Top