R.Masciola's HDR-10 UHD Test Patterns - Page 61 - AVS Forum | Home Theater Discussions And Reviews
Forum Jump: 
 1130Likes
Reply
 
Thread Tools
post #1801 of 1833 Old 08-15-2019, 09:32 AM
Advanced Member
 
Join Date: Jul 2017
Posts: 585
Mentioned: 7 Post(s)
Tagged: 0 Thread(s)
Quoted: 404 Post(s)
Liked: 199
Quote:
Originally Posted by e010238 View Post
Hi Ryan, I am using your Dolby Vision patterns to calibrate my 2018 LG C8 OLED TV. We have the same problem in that the 21 pt white balance adjustment points and your DV patterns do not match up. The C8 IRE adjustment points vs. the patterns that respond are:

IRE100:70%
90:65
85:65
70:60
60:55
55:55
45:50
40:45
35:45
30:40
25:35
20:30
15:25
10:20
5:15

IRE 95,80,75,65,50 have no patterns that respond to RGB adjustments.

I wonder if you could talk to LG like you did on the HDR to find out correct pattern matching. Hopefully we would see solution in your next version. Thanks.
If you touched those IREs can you verify that the greyscale ramp is grey and not discoloured?

LG OLED 55 B7v | LG SK10Y Soundbar | Samsung 55KS8000 | LG Plasma 50PZ550 | i1D3 OEM | Calman Enthusiast
MrRobotoPlus is offline  
Sponsored Links
Advertisement
 
post #1802 of 1833 Old 08-15-2019, 09:59 AM
Member
 
e010238's Avatar
 
Join Date: Apr 2010
Posts: 34
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 31 Post(s)
Liked: 11
Quote:
Originally Posted by MrRobotoPlus View Post
If you touched those IREs can you verify that the greyscale ramp is grey and not discoloured?
i deleted the message because I am not sure about the accuracy of matching those patterns. There is no discoloration after adjustments ether just 2 pts or 21 pts.
e010238 is offline  
post #1803 of 1833 Old 08-16-2019, 04:32 AM
AVS Forum Special Member
 
asharma's Avatar
 
Join Date: Apr 2007
Location: Canada
Posts: 2,224
Mentioned: 14 Post(s)
Tagged: 0 Thread(s)
Quoted: 1803 Post(s)
Liked: 562
Black and White clipping values

When using my JVC N7 with Panny 820, and using 4K black and white clipping patterns, with Panny optimizer on or off, brightness is -4, contrast is +9. When using my HiMedia Q10 Pro streamer with 4K mkv’s, using same test patterns brightness is -2, contrast is 7...

Does it make sense that the 2 devices would need different brightness and contrast levels? Does it tell me that the video processor in the Panny is that much more refined?


Also, when setting white clipping using the R.Masciola white clipping patterns, is the correct approach to raise contrast until nothing past 1000 nits flashes? Thanks folks

Video:JVC N7 Sony 85” 900F Sony 55” 900F
Audio: Paradigm Prestige 95, 55C, 15b Dual JL Audio D110
Receiver/Amp: Anthem 1120, Emotiva XPA-5 Gen 2
Source: Panasonic UB820, triple black velvet batcave
Screen:120” 2.35 Elunevision Tab tensioned motorized drop down screen;
asharma is offline  
Sponsored Links
Advertisement
 
post #1804 of 1833 Old 08-16-2019, 06:51 AM
AVS Forum Special Member
 
Rolls-Royce's Avatar
 
Join Date: Apr 2003
Location: Victorville, CA
Posts: 4,110
Mentioned: 38 Post(s)
Tagged: 0 Thread(s)
Quoted: 1121 Post(s)
Liked: 1225
Quote:
Originally Posted by asharma View Post
When using my JVC N7 with Panny 820, and using 4K black and white clipping patterns, with Panny optimizer on or off, brightness is -4, contrast is +9. When using my HiMedia Q10 Pro streamer with 4K mkv’s, using same test patterns brightness is -2, contrast is 7...

Does it make sense that the 2 devices would need different brightness and contrast levels? Does it tell me that the video processor in the Panny is that much more refined?


Also, when setting white clipping using the R.Masciola white clipping patterns, is the correct approach to raise contrast until nothing past 1000 nits flashes? Thanks folks
The Panasonic, as a standalone disc player, is in nearly all cases more accurate in output (YCbCr) than streaming media players, which could be RGB or YCbCr.
With all due respect to Ryan, I don't use his clipping patterns to set contrast, merely as a check to see what my TV's tone mapping is doing. You should be setting your display's contrast and brightness to their defaults in HDR mode, for its tone mapping is based on them.
galonzo and mrtickleuk like this.

...Royce...

"I never drink...wine."
Bela Lugosi, DRACULA, 1931
Rolls-Royce is offline  
post #1805 of 1833 Old 08-16-2019, 08:14 AM
AVS Forum Special Member
 
asharma's Avatar
 
Join Date: Apr 2007
Location: Canada
Posts: 2,224
Mentioned: 14 Post(s)
Tagged: 0 Thread(s)
Quoted: 1803 Post(s)
Liked: 562
Quote:
Originally Posted by Rolls-Royce View Post
The Panasonic, as a standalone disc player, is in nearly all cases more accurate in output (YCbCr) than streaming media players, which could be RGB or YCbCr.
With all due respect to Ryan, I don't use his clipping patterns to set contrast, merely as a check to see what my TV's tone mapping is doing. You should be setting your display's contrast and brightness to their defaults in HDR mode, for its tone mapping is based on them.
Thanks, appreciate the Panny info...in regards to setting contrast for clipping, I just wonder if for projectors it is different...Given they only output 50-100 nits, Tmapping is required and thus setting contrast for clipping is required...Welcome anymore thoughts you or others may have...Perhaps Ryan could comment if this is proper usage of 1000 nit clipping pattern for projectors...

Video:JVC N7 Sony 85” 900F Sony 55” 900F
Audio: Paradigm Prestige 95, 55C, 15b Dual JL Audio D110
Receiver/Amp: Anthem 1120, Emotiva XPA-5 Gen 2
Source: Panasonic UB820, triple black velvet batcave
Screen:120” 2.35 Elunevision Tab tensioned motorized drop down screen;
asharma is offline  
post #1806 of 1833 Old 08-16-2019, 08:26 AM
AVS Forum Special Member
 
Join Date: Dec 2014
Location: Mississauga, ON, Canada
Posts: 6,781
Mentioned: 126 Post(s)
Tagged: 0 Thread(s)
Quoted: 4744 Post(s)
Liked: 1779
Quote:
Originally Posted by asharma View Post
I just wonder if for projectors it is different...Given they only output 50-100 nits, Tmapping is required and thus setting contrast for clipping is required...
The JVC NX-7 knows it’s a projector so its tone-mapping is designed for that. It also adjusts the tone mapping according to the metadata. Unless you know what you’re doing (not implying you don’t ) there’s no reason to adjust the Contrast for HDR.
Rolls-Royce and galonzo like this.
Dominic Chan is online now  
post #1807 of 1833 Old 08-16-2019, 09:30 AM
AVS Forum Special Member
 
asharma's Avatar
 
Join Date: Apr 2007
Location: Canada
Posts: 2,224
Mentioned: 14 Post(s)
Tagged: 0 Thread(s)
Quoted: 1803 Post(s)
Liked: 562
Quote:
Originally Posted by Dominic Chan View Post
The JVC NX-7 knows it’s a projector so its tone-mapping is designed for that. It also adjusts the tone mapping according to the metadata. Unless you know what you’re doing (not implying you don’t ) there’s no reason to adjust the Contrast for HDR.
Thanks Dom...trust me, I don’t know what I’m doing...is there a downside to setting my contrast at +9 which is where it clips at 1000 nits...? Brightness at -4 (juuuust see bar 68) contrast at +9, DRA on Panny +3, any downsides...John Wick 2 looked totally amaaaaaazing with tons of POP!

Video:JVC N7 Sony 85” 900F Sony 55” 900F
Audio: Paradigm Prestige 95, 55C, 15b Dual JL Audio D110
Receiver/Amp: Anthem 1120, Emotiva XPA-5 Gen 2
Source: Panasonic UB820, triple black velvet batcave
Screen:120” 2.35 Elunevision Tab tensioned motorized drop down screen;
asharma is offline  
post #1808 of 1833 Old 08-16-2019, 10:32 AM
AVS Forum Special Member
 
Join Date: Dec 2014
Location: Mississauga, ON, Canada
Posts: 6,781
Mentioned: 126 Post(s)
Tagged: 0 Thread(s)
Quoted: 4744 Post(s)
Liked: 1779
Quote:
Originally Posted by asharma View Post
Thanks Dom...trust me, I don’t know what I’m doing...is there a downside to setting my contrast at +9 which is where it clips at 1000 nits...? Brightness at -4 (juuuust see bar 68) contrast at +9, DRA on Panny +3, any downsides...John Wick 2 looked totally amaaaaaazing with tons of POP!
John Wick 2 has MaxCLL of 1000 nits. How well does this setting work with discs mastered at higher nits?

Ultimately, if you’re doing adjustments primarily based on visual feedback then there’s really no “harm” in that.
galonzo likes this.
Dominic Chan is online now  
post #1809 of 1833 Old 08-16-2019, 10:38 AM
AVS Forum Special Member
 
asharma's Avatar
 
Join Date: Apr 2007
Location: Canada
Posts: 2,224
Mentioned: 14 Post(s)
Tagged: 0 Thread(s)
Quoted: 1803 Post(s)
Liked: 562
Quote:
Originally Posted by Dominic Chan View Post
John Wick 2 has MaxCLL of 1000 nits. How well does this setting work with discs mastered at higher nits?

Ultimately, if you’re doing adjustments primarily based on visual feedback then there’s really no “harm” in that.
Thanks again Dom...that was my next test...I’ll test MadMax tonight...based on -4, +9, any predictions for what I’ll see with a 4000 nit title?

Video:JVC N7 Sony 85” 900F Sony 55” 900F
Audio: Paradigm Prestige 95, 55C, 15b Dual JL Audio D110
Receiver/Amp: Anthem 1120, Emotiva XPA-5 Gen 2
Source: Panasonic UB820, triple black velvet batcave
Screen:120” 2.35 Elunevision Tab tensioned motorized drop down screen;
asharma is offline  
post #1810 of 1833 Old 08-16-2019, 01:08 PM
AVS Forum Special Member
 
Join Date: Jul 2016
Posts: 1,859
Mentioned: 11 Post(s)
Tagged: 0 Thread(s)
Quoted: 1882 Post(s)
Liked: 1171
I have a Panasonic UB-820 player and the Masciola UHD HDR10 disc set that I transferred to USB thumb drive.

When using Calman to manually calibrate with this setup, does anyone have recommended settings that I should be using on the Panasonic regarding the HDMI bandwidth 0-265 settings, Color output settings RGB 8bit, Ycbr 12 bit, etc and the HDR optimizer on or off?

I am new to display calibration but want to get started and try calibrating my new JVC RS3000 Projector. I have about 400 hours on the bulb and would like to calibrate for both SDR and HDR.

I think I can use Mobile Forge for the pattern generator for SDR but want to use the Masciola disc for HDR with my Panasonic. However I am unsure what settings on the Panasonic are best for calibration procedures as I only set them up for best results playing back HDR10 discs on my projector.
GregCh is online now  
post #1811 of 1833 Old 08-16-2019, 03:26 PM
AVS Forum Special Member
 
asharma's Avatar
 
Join Date: Apr 2007
Location: Canada
Posts: 2,224
Mentioned: 14 Post(s)
Tagged: 0 Thread(s)
Quoted: 1803 Post(s)
Liked: 562
Quote:
Originally Posted by Dominic Chan View Post
John Wick 2 has MaxCLL of 1000 nits. How well does this setting work with discs mastered at higher nits?

Ultimately, if you’re doing adjustments primarily based on visual feedback then there’s really no “harm” in that.
Hey Dom, just testing MadMax...looks “fine” to me...Are u able to tell me what “technically” is not right by setting contrast to 9 at the point where it clips at a 1000 nits as per the 4K patterns? Thanks man

Video:JVC N7 Sony 85” 900F Sony 55” 900F
Audio: Paradigm Prestige 95, 55C, 15b Dual JL Audio D110
Receiver/Amp: Anthem 1120, Emotiva XPA-5 Gen 2
Source: Panasonic UB820, triple black velvet batcave
Screen:120” 2.35 Elunevision Tab tensioned motorized drop down screen;
asharma is offline  
post #1812 of 1833 Old 08-16-2019, 03:37 PM
AVS Forum Special Member
 
Join Date: Dec 2014
Location: Mississauga, ON, Canada
Posts: 6,781
Mentioned: 126 Post(s)
Tagged: 0 Thread(s)
Quoted: 4744 Post(s)
Liked: 1779
Quote:
Originally Posted by asharma View Post
Hey Dom, just testing MadMax...looks “fine” to me...Are u able to tell me what “technically” is not right by setting contrast to 9 at the point where it clips at a 1000 nits as per the 4K patterns? Thanks man
If you adjust Contrast to +9, any highlights beyond 1000 nits May be clipped.

Last edited by Dominic Chan; 08-16-2019 at 03:43 PM.
Dominic Chan is online now  
post #1813 of 1833 Old 08-16-2019, 03:45 PM
AVS Forum Special Member
 
asharma's Avatar
 
Join Date: Apr 2007
Location: Canada
Posts: 2,224
Mentioned: 14 Post(s)
Tagged: 0 Thread(s)
Quoted: 1803 Post(s)
Liked: 562
Quote:
Originally Posted by Dominic Chan View Post
If you adjust Contrast to +9, any highlights beyond 1000 nits will be clipped.
Ahhhhh yes, that makes sense...perhaps I need 2 profiles...one for 1000 nit titles and one for 4000 nit titles OR back contrast down to a happy medium like 4 or something that will serve both 1000 and 4000 nit titles...

Video:JVC N7 Sony 85” 900F Sony 55” 900F
Audio: Paradigm Prestige 95, 55C, 15b Dual JL Audio D110
Receiver/Amp: Anthem 1120, Emotiva XPA-5 Gen 2
Source: Panasonic UB820, triple black velvet batcave
Screen:120” 2.35 Elunevision Tab tensioned motorized drop down screen;
asharma is offline  
post #1814 of 1833 Old 08-16-2019, 03:49 PM
AVS Forum Special Member
 
Join Date: Dec 2014
Location: Mississauga, ON, Canada
Posts: 6,781
Mentioned: 126 Post(s)
Tagged: 0 Thread(s)
Quoted: 4744 Post(s)
Liked: 1779
Quote:
Originally Posted by asharma View Post
Ahhhhh yes, that makes sense...perhaps I need 2 profiles...one for 1000 nit titles and one for 4000 nit titles OR back contrast down to a happy medium like 4 or something that will serve both 1000 and 4000 nit titles...
The NX-7 adjusts the automatic tone mapping based on the metadata, but I’m not sure how the manual Contrast setting affects that.
Dominic Chan is online now  
post #1815 of 1833 Old 08-16-2019, 03:55 PM
AVS Forum Special Member
 
asharma's Avatar
 
Join Date: Apr 2007
Location: Canada
Posts: 2,224
Mentioned: 14 Post(s)
Tagged: 0 Thread(s)
Quoted: 1803 Post(s)
Liked: 562
Quote:
Originally Posted by Dominic Chan View Post
The NX-7 adjusts the automatic tone mapping based on the metadata, but I’m not sure how the manual Contrast setting affects that.
Ok, good point...if u stumble across the answer, pls let me know...for now I decided to try contrast at 0 and JVC auto TMapping at 1 which gives a wee bit more brightness...also DRA on Panny set to 3...plenty of pop on Madmax

Video:JVC N7 Sony 85” 900F Sony 55” 900F
Audio: Paradigm Prestige 95, 55C, 15b Dual JL Audio D110
Receiver/Amp: Anthem 1120, Emotiva XPA-5 Gen 2
Source: Panasonic UB820, triple black velvet batcave
Screen:120” 2.35 Elunevision Tab tensioned motorized drop down screen;
asharma is offline  
post #1816 of 1833 Old 08-17-2019, 10:36 PM
Member
 
Join Date: Jul 2018
Location: København
Posts: 114
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 104 Post(s)
Liked: 58
I tried to calibrate my LG C8's DV mode using the DVS DV suite, but I found when attempting to use the 20pt white balance controls that they didn't correspond to the provided 21pt greyscale patches?
Ken Masters is offline  
post #1817 of 1833 Old 08-17-2019, 10:55 PM
AVS Forum Special Member
 
skschatzman's Avatar
 
Join Date: Oct 2016
Posts: 3,603
Mentioned: 78 Post(s)
Tagged: 1 Thread(s)
Quoted: 2720 Post(s)
Liked: 1611
Quote:
Originally Posted by Ken Masters View Post
I tried to calibrate my LG C8's DV mode using the DVS DV suite, but I found when attempting to use the 20pt white balance controls that they didn't correspond to the provided 21pt greyscale patches?
Yes. This is quite normal and discussed often. You have to find the targets through trial and error. A lot of TV manufacturers do this.
Ken Masters likes this.
skschatzman is offline  
post #1818 of 1833 Old 08-30-2019, 04:34 AM
Senior Member
 
mombasa123's Avatar
 
Join Date: Dec 2012
Posts: 200
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 133 Post(s)
Liked: 40
So understood that for HDR calibration should never touch Brightness & Contrast.

However for Samsung QLED (Q9FN) the panel is way too bright (EOTF & Luminance curves above reference so have to reduce Contrast from default 50 to 45 to better match. That's all good but, I am finding that white is clipping at 1350 nits and 80,90 & 100% blending into each other on the various patterns of Ryan's disc. Reducing Contrast to 41/42 takes white up to 1600 nits (albeit pink tinge) etc. The EOTF & Luminance curve is then a bit below reference at the top end.

Any thoughts please?
mombasa123 is offline  
post #1819 of 1833 Old 08-30-2019, 05:57 AM
AVS Forum Special Member
 
Rolls-Royce's Avatar
 
Join Date: Apr 2003
Location: Victorville, CA
Posts: 4,110
Mentioned: 38 Post(s)
Tagged: 0 Thread(s)
Quoted: 1121 Post(s)
Liked: 1225
Quote:
Originally Posted by mombasa123 View Post
So understood that for HDR calibration should never touch Brightness & Contrast.

However for Samsung QLED (Q9FN) the panel is way too bright (EOTF & Luminance curves above reference so have to reduce Contrast from default 50 to 45 to better match. That's all good but, I am finding that white is clipping at 1350 nits and 80,90 & 100% blending into each other on the various patterns of Ryan's disc. Reducing Contrast to 41/42 takes white up to 1600 nits (albeit pink tinge) etc. The EOTF & Luminance curve is then a bit below reference at the top end.

Any thoughts please?
You should never accept color clipping at maximum white - which is what is causing the pink tinge - in order to get higher luminance when setting contrast, either in SDR or in HDR. Do keep in mind that HDR as it exists today cannot be fully calibrated as SDR can. The maximum targets for HDR luminance and color simply cannot be met by any display on the market, so they have to do tone mapping. Your results therefore will always be impacted by the display's tone mapping as well as its actual performance. Some things you can't fix without disrupting other aspects.
galonzo, mombasa123 and mrtickleuk like this.

...Royce...

"I never drink...wine."
Bela Lugosi, DRACULA, 1931

Last edited by Rolls-Royce; 08-30-2019 at 06:46 AM.
Rolls-Royce is offline  
post #1820 of 1833 Old 08-30-2019, 09:37 AM
AVS Forum Special Member
 
Join Date: Dec 2014
Location: Mississauga, ON, Canada
Posts: 6,781
Mentioned: 126 Post(s)
Tagged: 0 Thread(s)
Quoted: 4744 Post(s)
Liked: 1779
Quote:
Originally Posted by mombasa123 View Post
However for Samsung QLED (Q9FN) the panel is way too bright (EOTF & Luminance curves above reference so have to reduce Contrast from default 50 to 45 to better match.
If reducing Contrast (slightly) gives you better match to the PQ EOTF, then by all means.

Quote:
That's all good but, I am finding that white is clipping at 1350 nits and 80,90 & 100% blending into each other on the various patterns of Ryan's disc. Reducing Contrast to 41/42 takes white up to 1600 nits (albeit pink tinge) etc. The EOTF & Luminance curve is then a bit below reference at the top end.
I believe the test patterns have metedata that indicates a MaxCLL (Maximum Content Light Level) of 1000 nits, in which case it's quite ligit for the display to clip above 1000 nits (~75% input), as there is not supposed to be any contents above that. Ideally you should also use test patterns with MaxCLL of 4000 nits, to confirm that the display adjusts its tone mapping accordingly (clipping above ~90% input).
galonzo, mombasa123 and jrref like this.
Dominic Chan is online now  
post #1821 of 1833 Old 08-30-2019, 08:35 PM
Senior Member
 
mombasa123's Avatar
 
Join Date: Dec 2012
Posts: 200
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 133 Post(s)
Liked: 40
Quote:
Originally Posted by Dominic Chan View Post
If reducing Contrast (slightly) gives you better match to the PQ EOTF, then by all means.


I believe the test patterns have metedata that indicates a MaxCLL (Maximum Content Light Level) of 1000 nits, in which case it's quite ligit for the display to clip above 1000 nits (~75% input), as there is not supposed to be any contents above that. Ideally you should also use test patterns with MaxCLL of 4000 nits, to confirm that the display adjusts its tone mapping accordingly (clipping above ~90% input).
Thanks. Yes, I read the documentation and the disk is mastered at 1000 nits but was not quite sure what that actually meant. So, I will leave the contrast at 45 to match EOTF curve and not reduce it further as the clipping above 1000 nits is expected.

White Clipping Demonstrates how well your display tone maps peak luminance information based on the metadata in which it’s being feed. The metadata within this particular pattern is designed to clip at 1,000 nit. However, no matter how much you raise or lower the contrast controls, the peak luminance information (flashing bars) will always remain the same, even though the light output may increase or decrease onscreen. This is due to the internal tone mapping of your display.

Will try and find some 4K nit patterns - Spears & Munsil on order so maybe that has some.
mombasa123 is offline  
post #1822 of 1833 Old 09-01-2019, 06:24 AM - Thread Starter
A/V Addicted Member
 
mascior's Avatar
 
Join Date: Apr 2008
Location: Currituck, NC
Posts: 837
Mentioned: 49 Post(s)
Tagged: 0 Thread(s)
Quoted: 413 Post(s)
Liked: 780
Quote:
Originally Posted by mombasa123 View Post
Thanks. Yes, I read the documentation and the disk is mastered at 1000 nits but was not quite sure what that actually meant. So, I will leave the contrast at 45 to match EOTF curve and not reduce it further as the clipping above 1000 nits is expected.

White Clipping Demonstrates how well your display tone maps peak luminance information based on the metadata in which it’s being feed. The metadata within this particular pattern is designed to clip at 1,000 nit. However, no matter how much you raise or lower the contrast controls, the peak luminance information (flashing bars) will always remain the same, even though the light output may increase or decrease onscreen. This is due to the internal tone mapping of your display.

Will try and find some 4K nit patterns - Spears & Munsil on order so maybe that has some.
Hi mombasa123,

These patterns can be found within the miscellaneous section of our HDR-10 test pattern suite (digital files). Under "HDR Clipping Test Patterns" there are four different variations of the HDR Clipping Grayscale patterns - 1,000 nit, 2,000 nit, 4,000 nit and 10,000 nit mastering peak luminance, along with the corresponding MaxCLL.

Also, under the "White Clipping 80-100 Different nit Levels" I have nine variations of white clipping test patterns from 2,000 nit to 10,000 nit in 1,000 nit increments, all with the proper corresponding metadata.

The MediaInfo of the 4,000 nit pattern regarding nit level is as follows:

Mastering display luminance: min: 0.0005 cd/m2, max: 4000 cd/m2
Maximum Content Light Level: 4000 cd/m2
Maximum Frame-Average Light Level: 400 cd/m2

I hope this helps. Please let me know if you have any further questions or concerns. Thank you!

- Ryan M.
galonzo and mombasa123 like this.

Last edited by mascior; 09-01-2019 at 06:30 AM.
mascior is offline  
post #1823 of 1833 Old 09-02-2019, 12:48 AM
Senior Member
 
mombasa123's Avatar
 
Join Date: Dec 2012
Posts: 200
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 133 Post(s)
Liked: 40
Quote:
Originally Posted by mascior View Post
Hi mombasa123,

These patterns can be found within the miscellaneous section of our HDR-10 test pattern suite (digital files). Under "HDR Clipping Test Patterns" there are four different variations of the HDR Clipping Grayscale patterns - 1,000 nit, 2,000 nit, 4,000 nit and 10,000 nit mastering peak luminance, along with the corresponding MaxCLL.

Also, under the "White Clipping 80-100 Different nit Levels" I have nine variations of white clipping test patterns from 2,000 nit to 10,000 nit in 1,000 nit increments, all with the proper corresponding metadata.

The MediaInfo of the 4,000 nit pattern regarding nit level is as follows:

Mastering display luminance: min: 0.0005 cd/m2, max: 4000 cd/m2
Maximum Content Light Level: 4000 cd/m2
Maximum Frame-Average Light Level: 400 cd/m2

I hope this helps. Please let me know if you have any further questions or concerns. Thank you!

- Ryan M.
Thanks Ryan. I played with those patterns today. However no matter which nits pattern I use my meter readings stay pretty much the same, eg 1000 nits and 2000 nits patterns both read ~1400 nits. I would have thought the 1000 nits patch would give me a 1000 meter reading?

Also, similarly with the HDR clipping patterns. I get the bars going up to 79% amplitude regardless of the nits level. Reducing contrast increases the number of bars across all nit pattern levels.

Worse, the nits level drops gradually from 1500 to around 1200 nits as the panel warms up.

Really struggling to understand what's going on

Edit : Wondering if my media player (Vero 4K+) is sending MaxCLL metadata ……

Edit 2 : Just popped over to the OSMC forums and confirmed that the Vero does not pass MaxCLl metadata. So what does that mean wrt display tone mapping in the real world?

Last edited by mombasa123; 09-02-2019 at 01:23 AM.
mombasa123 is offline  
post #1824 of 1833 Old 09-02-2019, 08:29 AM
AVS Forum Special Member
 
Rolls-Royce's Avatar
 
Join Date: Apr 2003
Location: Victorville, CA
Posts: 4,110
Mentioned: 38 Post(s)
Tagged: 0 Thread(s)
Quoted: 1121 Post(s)
Liked: 1225
Quote:
Originally Posted by mombasa123 View Post
Thanks Ryan. I played with those patterns today. However no matter which nits pattern I use my meter readings stay pretty much the same, eg 1000 nits and 2000 nits patterns both read ~1400 nits. I would have thought the 1000 nits patch would give me a 1000 meter reading?

Also, similarly with the HDR clipping patterns. I get the bars going up to 79% amplitude regardless of the nits level. Reducing contrast increases the number of bars across all nit pattern levels.

Worse, the nits level drops gradually from 1500 to around 1200 nits as the panel warms up.

Really struggling to understand what's going on

Edit : Wondering if my media player (Vero 4K+) is sending MaxCLL metadata ……

Edit 2 : Just popped over to the OSMC forums and confirmed that the Vero does not pass MaxCLl metadata. So what does that mean wrt display tone mapping in the real world?
Welcome to what pro calibrators have called the "Wild West" of HDR. Different displays respond to metadata differently. Some, like the Samsungs, don't use all of it. At least, they haven't in some past models. Metadata does not cap your display's output, but instead informs it of the particulars of the signal and the mastering display used. It is up to the tone mapping to decide what to do with that info. There is still likely information in the signal coded above the max given by the metadata, and the display will try to reproduce it to the best of its ability.
galonzo and mombasa123 like this.

...Royce...

"I never drink...wine."
Bela Lugosi, DRACULA, 1931

Last edited by Rolls-Royce; 09-02-2019 at 08:42 AM.
Rolls-Royce is offline  
post #1825 of 1833 Old 09-03-2019, 03:52 AM
Senior Member
 
mombasa123's Avatar
 
Join Date: Dec 2012
Posts: 200
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 133 Post(s)
Liked: 40
I tried some of the patterns on the Samsung internal player on the assumption that it respects the HDR metadata. The results are interesting to say the least.

With the Q9FN default Contrast at 50 :
1000 nits pattern clips at 1000 nits (75%)
2000 nits pattern clips at just under 2000 (80%)

With the Q9FN default Contrast at 45 (to match the EOTF curve) :
1000 nits pattern shows bars above 1000 nits (78%)
2000 nits pattern shows bars well above 2000 nits (85)

So, question is, do I follow the EOTF curve (contrast at 45) or the above patterns so that that they clip correctly (contrast at 50)?
Attached Thumbnails
Click image for larger version

Name:	2k nits Contrast 45.jpg
Views:	30
Size:	297.3 KB
ID:	2610556   Click image for larger version

Name:	2k nits Contrast 50.jpg
Views:	30
Size:	268.3 KB
ID:	2610558   Click image for larger version

Name:	1k nits Contrast 50.jpg
Views:	29
Size:	267.7 KB
ID:	2610560   Click image for larger version

Name:	1k nits Contrast 45.jpg
Views:	33
Size:	278.2 KB
ID:	2610562  
mombasa123 is offline  
post #1826 of 1833 Old 09-03-2019, 04:52 AM
AVS Forum Special Member
 
Join Date: Dec 2014
Location: Mississauga, ON, Canada
Posts: 6,781
Mentioned: 126 Post(s)
Tagged: 0 Thread(s)
Quoted: 4744 Post(s)
Liked: 1779
Quote:
Originally Posted by mombasa123 View Post
So, question is, do I follow the EOTF curve (contrast at 45) or the above patterns so that that they clip correctly (contrast at 50)?
Clipping is a “necessary evil”, not an objective. With Contrast at 45, in addition to better following the PQ curve, it gives you much better gradation of the highlights above 70% (~600 nits), but of course the brightest highlights will not be quite as bright.
galonzo and jrref like this.

Last edited by Dominic Chan; 09-04-2019 at 06:00 AM.
Dominic Chan is online now  
post #1827 of 1833 Old 09-03-2019, 06:55 AM
Senior Member
 
mombasa123's Avatar
 
Join Date: Dec 2012
Posts: 200
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 133 Post(s)
Liked: 40
Quote:
Originally Posted by Dominic Chan View Post
Clipping is a “necessary evil”, not an objective. You can see that Contrast at 45 gives you much better gradation of the highlights above 70% (~600 nits), in addition to following the PQ curve.
Thanks & noted. So best to set Contrast to follow EOTF for Samsung QLED.

Still a little confused why lowering Contrast displays bars above 1000 (75%) nits though for a 1000 nits master. It shouldn't do according to Ryan :

White Clipping Demonstrates how well your display tone maps peak luminance information based on the metadata in which it’s being feed. The metadata within this particular pattern is designed to clip at 1,000 nit. However, no matter how much you raise or lower the contrast controls, the peak luminance information (flashing bars) will always remain the same, even though the light output may increase or decrease onscreen. This is due to the internal tone mapping of your display.
mombasa123 is offline  
post #1828 of 1833 Old 09-03-2019, 09:39 PM
Senior Member
 
mombasa123's Avatar
 
Join Date: Dec 2012
Posts: 200
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 133 Post(s)
Liked: 40
Just an fyi but here are the 4K nits HDR clipping patterns with Contrast at 50 & 45.

Contrast at 50 clips above 4k nits while Contrast at 45 shows a bars upto 5k nits.

At least it proves that the displays tone mapping is good.

@mascior , would love to have your thoughts on what's happening here.
Attached Thumbnails
Click image for larger version

Name:	4k nits Contrast 50.jpg
Views:	30
Size:	290.8 KB
ID:	2610886   Click image for larger version

Name:	4k nits Contrast 45.jpg
Views:	29
Size:	324.4 KB
ID:	2610888  
mombasa123 is offline  
post #1829 of 1833 Old 09-04-2019, 05:48 PM - Thread Starter
A/V Addicted Member
 
mascior's Avatar
 
Join Date: Apr 2008
Location: Currituck, NC
Posts: 837
Mentioned: 49 Post(s)
Tagged: 0 Thread(s)
Quoted: 413 Post(s)
Liked: 780
Quote:
Originally Posted by mombasa123 View Post
Thanks & noted. So best to set Contrast to follow EOTF for Samsung QLED.

Still a little confused why lowering Contrast displays bars above 1000 (75%) nits though for a 1000 nits master. It shouldn't do according to Ryan :

White Clipping Demonstrates how well your display tone maps peak luminance information based on the metadata in which it’s being feed. The metadata within this particular pattern is designed to clip at 1,000 nit. However, no matter how much you raise or lower the contrast controls, the peak luminance information (flashing bars) will always remain the same, even though the light output may increase or decrease onscreen. This is due to the internal tone mapping of your display.
Quote:
Originally Posted by mombasa123 View Post
Just an fyi but here are the 4K nits HDR clipping patterns with Contrast at 50 & 45.

Contrast at 50 clips above 4k nits while Contrast at 45 shows a bars upto 5k nits.

At least it proves that the displays tone mapping is good.

@mascior , would love to have your thoughts on what's happening here.
Hi mombasa123,

I would have to echo Rolls-Royce on this...It has been my experience that Samsung completely ignores both MaxCLL/MaxFALL metadata, takes the peak mastering luminance and applies their own tone mapping methodology. This completely contradicts the statement within my instruction manual. However, may explain what you are seeing onscreen.

This test pattern in particular has onscreen digital information up to code 940 (100% / 10,000 nit). With a mastering display peak luminance of 4,000 nit...everything above 856 should be clipped. However, your display may be taking into account the entire available range, applying tone mapping based on the mastering display luminance of 4,000 nit and providing a variable roll-off based on your contrast adjustment. With a contrast setting of 50 seems to provide proper tone mapping of this particular pattern. Whereas, 45 may provide you more onscreen data at the expense of light output. This however should be a moot point, as the content was created to have a peak mastering luminance of 4,000 nit, MaxCLL of 4,000 nit and a MaxFALL of 400. Regardless, the correct contrast adjustment for your display seems to be 50. I hope this helps. Thanks!

- Ryan M.
mascior is offline  
post #1830 of 1833 Old 09-05-2019, 05:14 AM
Senior Member
 
mombasa123's Avatar
 
Join Date: Dec 2012
Posts: 200
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 133 Post(s)
Liked: 40
Quote:
Originally Posted by mascior View Post
Hi mombasa123,

I would have to echo Rolls-Royce on this...It has been my experience that Samsung completely ignores both MaxCLL/MaxFALL metadata, takes the peak mastering luminance and applies their own tone mapping methodology. This completely contradicts the statement within my instruction manual. However, may explain what you are seeing onscreen.

This test pattern in particular has onscreen digital information up to code 940 (100% / 10,000 nit). With a mastering display peak luminance of 4,000 nit...everything above 856 should be clipped. However, your display may be taking into account the entire available range, applying tone mapping based on the mastering display luminance of 4,000 nit and providing a variable roll-off based on your contrast adjustment. With a contrast setting of 50 seems to provide proper tone mapping of this particular pattern. Whereas, 45 may provide you more onscreen data at the expense of light output. This however should be a moot point, as the content was created to have a peak mastering luminance of 4,000 nit, MaxCLL of 4,000 nit and a MaxFALL of 400. Regardless, the correct contrast adjustment for your display seems to be 50. I hope this helps. Thanks!

- Ryan M.
Thank you Ryan. That is v informative and educational!

I think the correct Contrast for the Q9FN is 45 based on EOTF/Luminance curves (in preference to the clipping patterns). Vincent Teoh also pointed out the blownup highlights (ie loss of detail) in his review (default Contrast of 50). I have attached both 45 & 50 Contrast curves to illustrate. The 45 one is much closer to reference and inline with rtings.com post review analysis.

It has been all over the net that Samsung do artificially brighten the PQ so your comments do fit. I am not convinced that Samsung do not respect MaxCLL/FLL as my Vero 4K+ which (currently) does noy pass that metadata has very different clipping readings compared to the internal player. My Odroid N2 and Oppo based M9702 media streamer arriving soon so looking fwd to some comparisions.

Many thanks again for replying.
Attached Thumbnails
Click image for larger version

Name:	Contrast 45.JPG
Views:	27
Size:	235.6 KB
ID:	2611298   Click image for larger version

Name:	Contrast 50.JPG
Views:	22
Size:	235.9 KB
ID:	2611300  
mombasa123 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