HLG Grey Scale calibration? - AVS Forum | Home Theater Discussions And Reviews
Forum Jump: 
 
Thread Tools
post #1 of 10 Old 10-10-2018, 07:51 AM - Thread Starter
Member
 
Join Date: Mar 2018
Posts: 107
Mentioned: 5 Post(s)
Tagged: 0 Thread(s)
Quoted: 84 Post(s)
Liked: 79
HLG Grey Scale calibration?

I'm using HCFR, and I'm starting to wonder what the correct process is for Hybrid Log Gamma (HLG) grey scale measurement using the Masciola test patterns.

I have three TVs that can display HLG, and I'd sort of assumed that I should override the Target MaxL to 1000, and that shows a very accurate EOTF for one of the TVs (a Panasonic FZ952 OLED), but pretty bad results for the other two Samsung TVs which have a much lower maximum NITS:
Click image for larger version

Name:	a.jpg
Views:	35
Size:	127.3 KB
ID:	2466626

That actually correlates to my viewing experience, where HLG on the Panasonic OLED looks perfect, matching very well to the equivalent PQ image. Both the Samsung TVs look very dull with HLG material.

If, however, I don't override the Target MaxL and let it be set by the 100% reading, then the EOTF curves look all wrong:
Click image for larger version

Name:	b.jpg
Views:	23
Size:	128.7 KB
ID:	2466628

So, what should a TV do with HLG when the TV has a maximum luminance that's lower than 1000 cd/m²? Should it try to preserve the luminance for content below diffuse white (75%), or just match the HLG curve exactly and sacrifice the luminance for content below diffuse white?
tarbat is offline  
Sponsored Links
Advertisement
 
post #2 of 10 Old 10-11-2018, 10:53 AM
AVS Forum Special Member
 
Join Date: Dec 2014
Location: Mississauga, ON, Canada
Posts: 6,617
Mentioned: 124 Post(s)
Tagged: 0 Thread(s)
Quoted: 4634 Post(s)
Liked: 1717
Quote:
Originally Posted by tarbat View Post
I'm using HCFR, and I'm starting to wonder what the correct process is for Hybrid Log Gamma (HLG) grey scale measurement using the Masciola test patterns.
My understanding is that HLG behaves more like SDR, in the sense that the EOTF is relative rather than absolute. Thus there shouldn’t be any reason to override the measured peak luminance.
Dominic Chan is online now  
post #3 of 10 Old 10-11-2018, 11:36 AM - Thread Starter
Member
 
Join Date: Mar 2018
Posts: 107
Mentioned: 5 Post(s)
Tagged: 0 Thread(s)
Quoted: 84 Post(s)
Liked: 79
Quote:
Originally Posted by Dominic Chan View Post
My understanding is that HLG behaves more like SDR, in the sense that the EOTF is relative rather than absolute. Thus there shouldn’t be any reason to override the measured peak luminance.
Thanks. Yes, that's now my understanding after researching on various resources.

So, it looks like every TV I've tested has a very inaccurate HLG EOTF luminance response. Is that normal? I haven't seen much said on review sites and by calibrators about inaccurate HLG response, has anyone else tested this?
Click image for larger version

Name:	1.jpg
Views:	23
Size:	49.5 KB
ID:	2467166 Click image for larger version

Name:	2.jpg
Views:	20
Size:	48.9 KB
ID:	2467168 Click image for larger version

Name:	3.jpg
Views:	19
Size:	47.7 KB
ID:	2467170

It's strange, because the same 3 TVs all have excellent SDR and HDR-PQ luminance response curves, it's just HLG which is so far out. Or is there maybe something wrong with the HLG reference values calculation in HCFR?

Last edited by tarbat; 10-11-2018 at 11:47 AM.
tarbat is offline  
Sponsored Links
Advertisement
 
post #4 of 10 Old 10-11-2018, 02:16 PM
AVS Forum Special Member
 
Join Date: Dec 2014
Location: Mississauga, ON, Canada
Posts: 6,617
Mentioned: 124 Post(s)
Tagged: 0 Thread(s)
Quoted: 4634 Post(s)
Liked: 1717
Quote:
Originally Posted by tarbat View Post
Or is there maybe something wrong with the HLG reference values calculation in HCFR?
That’s quite possible. HLG implementation in HCFR may not have been fully tested. It gives me a very different reference curve when “Override” is selected, even if the peak luminance remains the same.
Dominic Chan is online now  
post #5 of 10 Old 12-12-2018, 11:56 AM - Thread Starter
Member
 
Join Date: Mar 2018
Posts: 107
Mentioned: 5 Post(s)
Tagged: 0 Thread(s)
Quoted: 84 Post(s)
Liked: 79
So, my research led me to this really useful document from the ITU - http://www.itu.int/dms_pub/itu-r/opb...2017-PDF-E.pdf

This chart in particular:
Click image for larger version

Name:	1.jpg
Views:	19
Size:	22.0 KB
ID:	2495956

Now, for HLG, Reference White is at 75%. So, looking at my TV's performance, at 75% my TV delivers 208 nits. So, according to this chart, my TV is setup to assume a peak luminance of 1000 nits, even though it actually only produces a maximum of 820 nits.

Also, when setting HCFR to "override targets" for HLG, you need to enter both the Target MaxL, and the HLG System Gamma. This table from the same document defines what gamma to use for each maximum nits:
Click image for larger version

Name:	2.jpg
Views:	17
Size:	20.5 KB
ID:	2495962

So for my TV, I needed to enter a MaxL of 1000, and a Gamma of 1.20. This gave a close match to the TV's actual performance, with a drop-off above 75% because of the limitations of the TV's 820 nits maximum luminance with a 10% window.
Click image for larger version

Name:	3.jpg
Views:	15
Size:	137.0 KB
ID:	2495966

EDIT: If the TV has an Ambient Light sensor and it's enabled for HLG, then you also need to adjust the gamma in "override targets" to match the ambient lighting. In my case, that reduces gamma from 1.20 to 1.18 for calibration in my dimly lit nightime room.
Click image for larger version

Name:	4.jpg
Views:	7
Size:	48.0 KB
ID:	2496200

Last edited by tarbat; 12-13-2018 at 05:29 AM.
tarbat is offline  
post #6 of 10 Old 12-12-2018, 12:54 PM
AVS Forum Special Member
 
Join Date: Dec 2014
Location: Mississauga, ON, Canada
Posts: 6,617
Mentioned: 124 Post(s)
Tagged: 0 Thread(s)
Quoted: 4634 Post(s)
Liked: 1717
Quote:
Originally Posted by tarbat View Post
So, my research led me to this really useful document from the ITU - http://www.itu.int/dms_pub/itu-r/opb...2017-PDF-E.pdf

This chart in particular:
Attachment 2495956

Now, for HLG, Reference White is at 75%. So, looking at my TV's performance, at 75% my TV delivers 208 nits. So, according to this chart, my TV is setup to assume a peak luminance of 1000 nits, even though it actually only produces a maximum of 820 nits.

Also, when setting HCFR to "override targets" for HLG, you need to enter both the Target MaxL, and the HLG System Gamma. This table from the same document defines what gamma to use for each maximum nits:
Attachment 2495962

So for my TV, I needed to enter a MaxL of 1000, and a Gamma of 1.20. This gave a close match to the TV's actual performance, with a drop-off above 75% because of the limitations of the TV's 820 nits maximum luminance with a 10% window.
Attachment 2495966
There’s an updated version:

https://www.itu.int/dms_pub/itu-r/op...2018-PDF-E.pdf
Dominic Chan is online now  
post #7 of 10 Old 12-12-2018, 12:56 PM
AVS Forum Special Member
 
Light Illusion's Avatar
 
Join Date: Aug 2010
Posts: 1,781
Mentioned: 27 Post(s)
Tagged: 0 Thread(s)
Quoted: 754 Post(s)
Liked: 1095
You can also check with the free LightSpace DPS.
Compare the different results with those from LightSpace?

Steve

Steve Shaw
LIGHT ILLUSION

Light Illusion is offline  
post #8 of 10 Old 12-12-2018, 01:00 PM - Thread Starter
Member
 
Join Date: Mar 2018
Posts: 107
Mentioned: 5 Post(s)
Tagged: 0 Thread(s)
Quoted: 84 Post(s)
Liked: 79
Quote:
Originally Posted by Dominic Chan View Post
Thanks for the link. Doesn’t seem to change anything concerning HLG calibration.
tarbat is offline  
post #9 of 10 Old 07-19-2019, 09:51 PM
Advanced Member
 
Anderegg's Avatar
 
Join Date: Jul 2012
Posts: 705
Mentioned: 27 Post(s)
Tagged: 0 Thread(s)
Quoted: 475 Post(s)
Liked: 270
I am preparing to create an HLG calibration for my Sony X900F, which produces a peak HDR10 luminance of 800nit with all the boosters enabled, but 550nit with them turned off for calibration. My black level at max brightness is typically around 0.15nit. Could anyone give some advice on how to set up HCFR and my signal path for calibrating? I am proficient in limited-full-limited RBG 8 bit SDR and limited-limited-limited Ycbcr 10 bit HDR10 calibrating, but HLG setup is not something I have attempted before.

Looking for an HLG calibration to view and grade my Sony camcorder HLG files

Paul

Sony X900F SDR and HDR Calibrations https://www.avsforum.com/forum/166-l...l#post57551552
Anderegg is offline  
post #10 of 10 Old 07-21-2019, 01:59 AM - Thread Starter
Member
 
Join Date: Mar 2018
Posts: 107
Mentioned: 5 Post(s)
Tagged: 0 Thread(s)
Quoted: 84 Post(s)
Liked: 79
The only advice I can give it to:
1. Start by measuring the luminance of the TV with a 75% grey scale window.
2. Use chart 1 in my post #5 to determine what maximum luminance your TV thinks it can do, and then override the HLG targets in HCFR with that value.
3. Also override HLG System Gamma (from chart 2).
4. If your TV is using an inbuilt ambient sensor, adjust for viewing environment using chart 3.

So for example, on my FZ952, I measure 75% grey scale as 208 cd/m². From the chart, that tells me my TV's HLG EOTF curve is setup for 1000 cd/m² maximum luminance. and HLG System Gamma is 1.2. Then I adjust for my viewing environment (75 lux) and subtract 0.04. That gives me overrides of Target MaxL = 1000 and HLG System Gamma = 1.16.

Click image for larger version

Name:	1.jpg
Views:	6
Size:	93.2 KB
ID:	2592986Click image for larger version

Name:	2.jpg
Views:	7
Size:	119.2 KB
ID:	2592988
tarbat is offline  
Sponsored Links
Advertisement
 
Reply Display Calibration

Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page


Forum Jump: 

Posting Rules  
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off