HCFR - Open source projector and display calibration software - Page 367 - AVS Forum | Home Theater Discussions And Reviews
Forum Jump: 
 1440Likes
Reply
 
Thread Tools
post #10981 of 12071 Old 09-17-2018, 10:31 AM
AVS Forum Addicted Member
 
lovingdvd's Avatar
 
Join Date: Jul 2002
Posts: 12,930
Mentioned: 79 Post(s)
Tagged: 0 Thread(s)
Quoted: 3659 Post(s)
Liked: 844
Quote:
Originally Posted by Dominic Chan View Post
I wouldn't know what to do in that case. That would defy logic.
I'm going to get my Spyder5 set up to work with HCFR so I can do the same measurement as I'm doing with the i1D3 except with the Spyder5 - to see if it reads the same. If the Spyder shows a straight (normal) gamut maybe one of the meters has gone bad.

Maybe there was an error in the auto-cal with the Spyder and it didn't get good readings. Although the graph at the end showed a perfectly white 45 degree diag line.

Quote:
I have not seen in your previous posts that you measured the gamma droop and subsequent fix in Autocal using the BT2020 profile.
Please see this post and the pic attached there: https://www.avsforum.com/forum/24-di...l#post56811818 . That's how it looked in the pre-calibrated stage. Afterward was a straight white line. So it certainly thinks it got it right. But maybe the Spyder5 didn't read correctly. Have you ever seen that happen? Note that I have never checked the autocal results with a separate meter and process and always since day 1 just assumed it got it right. So it may have always been messed up like this and each calibration may come out identical, or maybe a fluke auto-cal that will be fixed with a redo.

When I rerun the autocal, will I have to redo the CMS and white point?

To avoid the ambiguity, I made sure I said 50% stimulus level, so that it won't be confused with 50% saturation.[/QUOTE]
lovingdvd is online now  
Sponsored Links
Advertisement
 
post #10982 of 12071 Old 09-17-2018, 10:50 AM
AVS Forum Special Member
 
Join Date: Dec 2014
Location: Mississauga, ON, Canada
Posts: 6,946
Mentioned: 130 Post(s)
Tagged: 0 Thread(s)
Quoted: 4883 Post(s)
Liked: 1847
Quote:
Originally Posted by lovingdvd View Post
Please see this post and the pic attached there: https://www.avsforum.com/forum/24-di...l#post56811818 . That's how it looked in the pre-calibrated stage. Afterward was a straight white line.
I saw that. Nowhere does it say that it’s based on the BT.2020 profile.

Quote:
When I rerun the autocal, will I have to redo the CMS and white point?
Yes if you rerun colour autocal, no if you run gamma only autocal.

Note: I’ve just noticed that CMS has some effect on gamma (in theory it shouldn’t), so before you do anything major, check the gamma with CMS turned off, to see if that makes any difference.
Dominic Chan is online now  
post #10983 of 12071 Old 09-17-2018, 11:01 AM
AVS Forum Addicted Member
 
lovingdvd's Avatar
 
Join Date: Jul 2002
Posts: 12,930
Mentioned: 79 Post(s)
Tagged: 0 Thread(s)
Quoted: 3659 Post(s)
Liked: 844
Quote:
Originally Posted by Dominic Chan View Post
I saw that. Nowhere does it say that it’s based on the BT.2020 profile.
Prior to autocal, all my modes looked like this, including Standard and Reference in both high and low lamp.

Quote:
Note: I’ve just noticed that CMS has some effect on gamma (in theory it shouldn’t), so before you do anything major, check the gamma with CMS turned off, to see if that makes any difference.
Oh, great idea. When you turned your CMS on did you see a similar BT1886-like effect applied to the low end of your gamma? I'll check mine tonight and report back.

My CMS was turned off at the time of the autocal. Not sure that matters, or how to deal with that if indeed it is the CMS doing it. I suppose in that case it'll be best to do the RGBW spreadsheet Chad method instead of CMS, which should accomplish the same thing, yet avoid the gamma change (assuming CMS turns out to be to blame)?
lovingdvd is online now  
Sponsored Links
Advertisement
 
post #10984 of 12071 Old 09-17-2018, 12:54 PM
AVS Forum Addicted Member
 
lovingdvd's Avatar
 
Join Date: Jul 2002
Posts: 12,930
Mentioned: 79 Post(s)
Tagged: 0 Thread(s)
Quoted: 3659 Post(s)
Liked: 844
@Dominic Chan I'm going to try with a custom profile instead of CMS. To get the numbers needed for the spreadsheet, I need to measure RGBW at 50-100% *saturation*? So please let me know if I have these steps correct:

1. Go to Measures->Parameters and change "Number of saturation color levels" from 4 to 10? This will give me 10-100% for RGB, because I need the 50-100% values for the spreadsheet. And without this change I'd only have 0, 25, 50, and 100% instead of 50-100%.

2. Run a Measures->Saturation->Primary Colors.

3. Get the 50-100% values for the spreadsheet from the r/g/b "saturations scale" grid in HCFR, and get these same values from the grayscale for W.

Is that correct?

Also, regarding creating the offsets for making the custom color profile (if abandoning I believe you had previously mentioned something about using 50% saturation for HDR instead of 100%? So for gathering the spreadsheet numbers for Chad's sheet, for my HDR calibration would I want to change the Pattern Intensity from 100% to 50%? And for SDR use 100% and put that into a separate custom color profile?

Thanks!!
lovingdvd is online now  
post #10985 of 12071 Old 09-17-2018, 12:56 PM
AVS Forum Addicted Member
 
lovingdvd's Avatar
 
Join Date: Jul 2002
Posts: 12,930
Mentioned: 79 Post(s)
Tagged: 0 Thread(s)
Quoted: 3659 Post(s)
Liked: 844
@Dominic Chan I'm going to try with a custom profile instead of CMS. To get the numbers needed for the spreadsheet, I need to measure RGBW at 50-100% *saturation*? So please let me know if I have these steps correct:

1. Go to Measures->Parameters and change "Number of saturation color levels" from 4 to 10? This will give me 10-100% for RGB, because I need the 50-100% values for the spreadsheet. And without this change I'd only have 0, 25, 50, and 100% instead of 50-100%.

2. Run a Measures->Saturation->Primary Colors.

3. Get the 50-100% values for the spreadsheet from the r/g/b "saturations scale" grid in HCFR, and get these same values from the grayscale for W.

Is that correct?

Also, regarding creating the offsets for making the custom color profile (if abandoning I believe you had previously mentioned something about using 50% saturation for HDR instead of 100%? So for gathering the spreadsheet numbers for Chad's sheet, for my HDR calibration would I want to change the Pattern Intensity from 100% to 50%? And for SDR use 100% and put that into a separate custom color profile?

Thanks!!
lovingdvd is online now  
post #10986 of 12071 Old 09-17-2018, 01:16 PM
AVS Forum Special Member
 
Join Date: Dec 2014
Location: Mississauga, ON, Canada
Posts: 6,946
Mentioned: 130 Post(s)
Tagged: 0 Thread(s)
Quoted: 4883 Post(s)
Liked: 1847
Quote:
Originally Posted by lovingdvd View Post
@Dominic Chan I'm going to try with a custom profile instead of CMS. To get the numbers needed for the spreadsheet, I need to measure RGBW at 50-100% *saturation*? So please let me know if I have these steps correct:

1. Go to Measures->Parameters and change "Number of saturation color levels" from 4 to 10? This will give me 10-100% for RGB, because I need the 50-100% values for the spreadsheet. And without this change I'd only have 0, 25, 50, and 100% instead of 50-100%.

2. Run a Measures->Saturation->Primary Colors.

3. Get the 50-100% values for the spreadsheet from the r/g/b "saturations scale" grid in HCFR, and get these same values from the grayscale for W.

Is that correct?

Also, regarding creating the offsets for making the custom color profile (if abandoning I believe you had previously mentioned something about using 50% saturation for HDR instead of 100%? So for gathering the spreadsheet numbers for Chad's sheet, for my HDR calibration would I want to change the Pattern Intensity from 100% to 50%? And for SDR use 100% and put that into a separate custom color profile?

Thanks!!
I have answered these questions previously but you seem to keep repeating the same questions. Chad’s spreadsheet uses the average of stimulus levels, not saturation levels. In HCFR it’s easy to do for white, but somewhat tedius for colours as you have to take the measurements one at a time using different stimulus levels (pattern intensities).

Last edited by Dominic Chan; 09-17-2018 at 01:40 PM.
Dominic Chan is online now  
post #10987 of 12071 Old 09-17-2018, 05:35 PM
AVS Forum Addicted Member
 
lovingdvd's Avatar
 
Join Date: Jul 2002
Posts: 12,930
Mentioned: 79 Post(s)
Tagged: 0 Thread(s)
Quoted: 3659 Post(s)
Liked: 844
Quote:
Originally Posted by Dominic Chan View Post
I have answered these questions previously but you seem to keep repeating the same questions. Chad’s spreadsheet uses the average of stimulus levels, not saturation levels. In HCFR it’s easy to do for white, but somewhat tedius for colours as you have to take the measurements one at a time using different stimulus levels (pattern intensities).
I replied but it doesn't look like it posted, so I am trying again. I don't exactly understand the process for using different stimulus levels.

Is this the correct (updated) procedure to follow to get RGB using CalMAN for the spreadsheet?

1. Go to green saturation scale grid view in HCFR.
2. Click on the 100 column.
3. On the GDI, choose 50% pattern intensity.
4. Take a free measure and note xy for spreadsheet, use it for the 50% column in the spreadsheet.
5. Repeat step 3 and 4, except each time change the Pattern Intensity to 60, 70, 80, 90 and 100% (and putting the value in the corresponding % in the spreadsheet).
6. Repeat 1-5 for red and blue.

Is that correct?

Thanks.
lovingdvd is online now  
post #10988 of 12071 Old 09-17-2018, 06:51 PM
AVS Forum Special Member
 
Join Date: Dec 2014
Location: Mississauga, ON, Canada
Posts: 6,946
Mentioned: 130 Post(s)
Tagged: 0 Thread(s)
Quoted: 4883 Post(s)
Liked: 1847
Quote:
Originally Posted by lovingdvd View Post
I replied but it doesn't look like it posted, so I am trying again. I don't exactly understand the process for using different stimulus levels.

Is this the correct (updated) procedure to follow to get RGB using CalMAN for the spreadsheet?

1. Go to green saturation scale grid view in HCFR.
2. Click on the 100 column.
3. On the GDI, choose 50% pattern intensity.
4. Take a free measure and note xy for spreadsheet, use it for the 50% column in the spreadsheet.
5. Repeat step 3 and 4, except each time change the Pattern Intensity to 60, 70, 80, 90 and 100% (and putting the value in the corresponding % in the spreadsheet).
6. Repeat 1-5 for red and blue.

Is that correct?

Thanks.
You can do that, but it’s easier to use the Primary Colours grid where you can see and measure all the colours.
Dominic Chan is online now  
post #10989 of 12071 Old 09-17-2018, 06:59 PM
AVS Forum Addicted Member
 
lovingdvd's Avatar
 
Join Date: Jul 2002
Posts: 12,930
Mentioned: 79 Post(s)
Tagged: 0 Thread(s)
Quoted: 3659 Post(s)
Liked: 844
Quote:
Originally Posted by Dominic Chan View Post
You can do that, but it’s easier to use the Primary Colours grid where you can see and measure all the colours.
Let me see if I understand:

1. Select Primaries and Secondaries (ignore secondaries).
2. Set the GDI intensity to X% (50-100%).
3. Measure->Primary Colors.
4. Note RGB values for X% and note the RGB xy for the spreadsheet for X%.
5. Go to step 2 and choose the next increment (60, 70, 80, 90, 100).
6. You now have RGB offsets from spreadsheet.
7. For W, just use the xy for 50-100% from a grayscale.

Do I have this right now?
lovingdvd is online now  
post #10990 of 12071 Old 09-17-2018, 07:22 PM
AVS Forum Special Member
 
Join Date: Dec 2014
Location: Mississauga, ON, Canada
Posts: 6,946
Mentioned: 130 Post(s)
Tagged: 0 Thread(s)
Quoted: 4883 Post(s)
Liked: 1847
Quote:
Originally Posted by lovingdvd View Post
Let me see if I understand:

1. Select Primaries and Secondaries (ignore secondaries).
2. Set the GDI intensity to X% (50-100%).
3. Measure->Primary Colors.
4. Note RGB values for X% and note the RGB xy for the spreadsheet for X%.
5. Go to step 2 and choose the next increment (60, 70, 80, 90, 100).
6. You now have RGB offsets from spreadsheet.
7. For W, just use the xy for 50-100% from a grayscale.

Do I have this right now?
Yes. Alternatively, you can read W at the same time as R/G/B.

EDIT:
In Step 4: Not sure what you mean by Note RGB values
In Step 6: You have more than the offsets; the spreadsheet gives you the new coordinates.

Last edited by Dominic Chan; 09-17-2018 at 07:48 PM.
Dominic Chan is online now  
post #10991 of 12071 Old 09-17-2018, 10:51 PM
AVS Forum Addicted Member
 
lovingdvd's Avatar
 
Join Date: Jul 2002
Posts: 12,930
Mentioned: 79 Post(s)
Tagged: 0 Thread(s)
Quoted: 3659 Post(s)
Liked: 844
Gamma mystery - progress, almost there

Quote:
Originally Posted by Dominic Chan View Post
Yes. Alternatively, you can read W at the same time as R/G/B.

EDIT:
In Step 4: Not sure what you mean by Note RGB values
In Step 6: You have more than the offsets; the spreadsheet gives you the new coordinates.
OK so remember how my gamma was reading more like BT1886 instead of 2.2 despite default Gamma settings (Normal) after an autocal? Well, I figured out the problem (tho not sure how to solve it). The issue is with the laptop generated patterns. Apparently the laptop (Windows 10 Acer with Intel HD 4600 and Geforce 8xx) is doing something to the gamma.

To figure this out, first I ran the HFCR gamma measurements against my BT2020F and Reference, and confirmed that gamma was not measuring right (BT1886-like curve). I then ran a JVC autocal Log-only with the same exact params set on the pj, and the white line was perfectly diagonal at 45 degrees. Then I wondered if the problem may be somehow with the laptop. I used the option to Chromecast patterns from HCFR to my Shield, and guess what... The gamma for Normal measured almost a perfectly straight line at 2.2 (just a little bit of rise creeping in at 80% and 90%). Unfortunately I can't use Chromecast for anything color related, because for some reason it shrinks my gamut to like half its normal size! For instance when it displays the Red primary through the Shield, it looks like light orange on the screen (yet the JVC Red pattern in the service menu looks Red as it should). I'm not sure why the colors do not cast properly, as the Shield is outputting BT2020 (confirmed with Vertex). BTW this is all with BT2020F (I have abandoned NF entirely).

See the attached screenshot of my GDI. It has disable video LUT checked, and if that does what I think it is supposed to do, I don't think it's doing the trick. I spent a couple hours digging through all the various Intel HD 4600 display properties. There are some things about color correct and so forth and I disabled everything. There was something about gamma but it was a straight diagonal line. I don't think the Geforce card is being used by HCFR.

Any ideas how to ensure that Windows and or any display adapters or properties etc is not messing with the gamma? Maybe it's not, and its some other setting in HCFR or Windows? I'm connecting the laptop to my Marantz and sending it as 1080p and my JVC is set for 16-235 video levels. I didn't try the 0-255 setting now that I think about it, but I don't think that would be a fit for a 16-235 pj setup?

Thank you for all your help with this Dominic!
Attached Thumbnails
Click image for larger version

Name:	IMG_5429.jpg
Views:	41
Size:	1.51 MB
ID:	2456628  
lovingdvd is online now  
post #10992 of 12071 Old 09-18-2018, 05:05 AM
AVS Forum Special Member
 
Join Date: Dec 2014
Location: Mississauga, ON, Canada
Posts: 6,946
Mentioned: 130 Post(s)
Tagged: 0 Thread(s)
Quoted: 4883 Post(s)
Liked: 1847
Quote:
Originally Posted by lovingdvd View Post
I'm connecting the laptop to my Marantz and sending it as 1080p and my JVC is set for 16-235 video levels. I didn't try the 0-255 setting now that I think about it, but I don't think that would be a fit for a 16-235 pj setup?
Actually you should try using 0-255 in HCFR GDI. Even though the JVC is set for 16-235,most likely your graphics card is converting everything from full range to limited range so you don’t want double conversion.

To make sure you have no level mismatch issues, measure 100% white using Chromecast, then using GDI (first with the current settings, then with the change I mentioned above). A correctly set GDI should give you the same white (and black) level as Chromecast.

EDIT: A further refinement would be to set your graphics card to full range and HCFR to limited, but that’s for another day.

EDIT2: I just looked at your measurements and the black level was 0.053 nits (Contrast Ratio 746:1)! For sure that indicates a level mismatch issue. The good news is that your white level should also be higher than the 39.7 nits measurement.

Last edited by Dominic Chan; 09-18-2018 at 07:18 AM.
Dominic Chan is online now  
post #10993 of 12071 Old 09-19-2018, 01:19 AM
Senior Member
 
ebr9999's Avatar
 
Join Date: Jun 2013
Posts: 394
Mentioned: 9 Post(s)
Tagged: 0 Thread(s)
Quoted: 228 Post(s)
Liked: 92
Quote:
Originally Posted by ebr9999 View Post
I have saved in usercolors.csv targets for B6 HDR adjustment points (annexed, changed to .txt for quick upoad). Then I han run it as color checker.
Terrible dE as the Y target is taken for all the code value from Gray White at 100%. Any way I can fix that? Any way I can have something to whatch like a an RGB level graph?
Starting from defualt I have set SMTP 2084 HDR and UHDTV - Rec 2020. That's all I have done
Getting back to this question (issue?), it appears from measure that Y target has been taken from 100% Grey Scale, but not scaled:
Attached Thumbnails
Click image for larger version

Name:	Capture.PNG
Views:	373
Size:	108.5 KB
ID:	2456974  

TV: LG 55E8PLA, Pana TX-L37ET5E BD/DVR: Panasonic BMR-BWT735 Media player: Teufel Connector 2, HT Rec: ANTHEM MRX 510,
Front: Woofer&Tweeter, Center: 2Woofer&1Tweeter, Surround: Bose acoustimass 5 series II, Sub: SVS SB2000
ebr9999 is offline  
post #10994 of 12071 Old 09-19-2018, 10:54 AM
AVS Forum Addicted Member
 
lovingdvd's Avatar
 
Join Date: Jul 2002
Posts: 12,930
Mentioned: 79 Post(s)
Tagged: 0 Thread(s)
Quoted: 3659 Post(s)
Liked: 844
Quote:
Originally Posted by Dominic Chan View Post
Actually you should try using 0-255 in HCFR GDI. Even though the JVC is set for 16-235,most likely your graphics card is converting everything from full range to limited range so you don’t want double conversion.
Yes, thanks, changing to 00255 in HCFR GDI did the trick!

Quote:
EDIT: A further refinement would be to set your graphics card to full range and HCFR to limited, but that’s for another day.
Can you please explain more about why that is a further refinement, and how much of a difference it is likely to make, and in what way?

I changed my Intel HD 6400 display properties so that the Marantz 8805 I am connecting the laptop to through HDMI is set to full range. And confirmed with a test pattern in my browser (outside of HCFR) that it was full range. However I had the same gamma issue when set to full range if HCFR was set to 16-235. Unless the issue is that I need to set the laptop card itself to full range, as opposed to the "display". I googled for a good hour but there was nothing about it, and the limited vs full range option is not provided in the graphic card adapter for the "internal display" only when I select the secondary display (the Marantz its connected to). So the only way to get the correct gamma was with HCFR set to 0-255 and regardless of how I had the display properties set.

Quote:
EDIT2: I just looked at your measurements and the black level was 0.053 nits (Contrast Ratio 746:1)! For sure that indicates a level mismatch issue. The good news is that your white level should also be higher than the 39.7 nits measurement.
Yes, with the gamma fixed I get 48 nits.

OK so with all that out of the way and the gamma fixed, I went back to square one and redid all my testing. As it turns out, using the compressed gamma was also messing up some of my colors at the primary/secondary targets so I had to take a fresh look at the CMS and how/if that would help.

I did a autocal gamma+color and then tried several different methods to fine tune the calibration.

1) standard BT2020 profile with RGB gains adjusted at 100% with CMS.
2) standard BT2020 profile with RGB gains adjusted at 100% without CMS.
3) custom BT2020 color profile with RGB gains at 0 with CMS using Chad's BT2020 targets.
4) custom BT2020 color profile with RGB gains at 0 with CMS using Manni's refined BT2020 targets (designed I think not to try and hit the edges of BT2020 since we can't reach that anyway).

It was around 4am when I finished taking all the measurements so I'm going to look through them shortly and post a quick summary then. I'll also upload all the chc files so you can do your own assessment of which is the best approach. So far the results surprise me. More soon.

In the meantime - can you please clarify whether there is benefit to using the laptop at full range and HCFR set to 16-235 (which I couldn't get to work) because unless there is a significant difference it's probably not worth more troubleshooting to figure out why that does not work. Thanks!
lovingdvd is online now  
post #10995 of 12071 Old 09-19-2018, 10:54 AM
AVS Forum Addicted Member
 
lovingdvd's Avatar
 
Join Date: Jul 2002
Posts: 12,930
Mentioned: 79 Post(s)
Tagged: 0 Thread(s)
Quoted: 3659 Post(s)
Liked: 844
Quote:
Originally Posted by Dominic Chan View Post
Any luck?
Yes - was just typing up a reply when your post came in. See above.
lovingdvd is online now  
post #10996 of 12071 Old 09-19-2018, 11:58 AM
AVS Forum Special Member
 
Join Date: Dec 2014
Location: Mississauga, ON, Canada
Posts: 6,946
Mentioned: 130 Post(s)
Tagged: 0 Thread(s)
Quoted: 4883 Post(s)
Liked: 1847
Quote:
Originally Posted by lovingdvd View Post
In the meantime - can you please clarify whether there is benefit to using the laptop at full range and HCFR set to 16-235 (which I couldn't get to work) because unless there is a significant difference it's probably not worth more troubleshooting to figure out why that does not work. Thanks!
See this post and the discussion that led to it.
https://www.avsforum.com/forum/139-d...l#post55943440
There’s not much difference at high levels so for your purpose there’s no need to switch.

Last edited by Dominic Chan; 09-19-2018 at 12:14 PM.
Dominic Chan is online now  
post #10997 of 12071 Old 09-19-2018, 10:53 PM
Senior Member
 
howzz1854's Avatar
 
Join Date: Feb 2008
Posts: 409
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 15 Post(s)
Liked: 26
hey guys, i am at a loss here. i tried calibrating my HDR TV Samsung KU6300 that's connected to my Gaming machine running Windows 10 using the latest HCFR but discovered that there's a huge gamma discrepancy with Windows 10's native HDR WCG mode, vs the MadVR HDR passthrough method. the HCFR HDR10 output seem to be similar to the MadVR, in which when Meta is sent to the tv, all the desktop icons and wallpaper increases in contrast dramatically. but when i use the Windows 10's native HDR WCG toggle, the desktop elements becomes very dull. i know the TV is receiving HDR signal in both cases, but there's just a huge difference in terms of gamma curve and color mapping.

long story short, i tried to calibrate for both output method to see which one will yield the best result and found that when sending the HDR meta through MadVR/HCFR HDR10 it yields the best final calibrated result, while the Windows 10 native HDR toggle after calibrating it with HCFR (without the HCFR's internal HDR10 signal output), I can't get close to the HDR gamma/luminance reference curve and on top of that the color saturation mapping doesn't expand as far out as the MadVR/HCFR HDR10 output method. with the MadVR/HCFR, i can get very very close to the SMPTE 2084 curve, which resembles a snake shape, and the saturation coverage goes out a lot farther on the Rec2020 spectrum. but with the Windows 10 HDR output, the closest i can get the gamma curve is more inline with the regular BT 1886 reference curve, almost looks like a 2.1/2.2 gamma. needless to say, this provides less punch when watching a HDR movie or an HDR game.

when investigating more by looking at the MPC-HE/MadVR info toggle switch (Ctrl + J), i noticed that when using MadVR method, the info shows that the HDR method is "NV HDR". but when doing the Windows 10 method, it shows "OS HDR". all the other info are identical.

i would have just stick with the MadVR method except there're a lot of HDR games out there on PC, some uses the native OS HDR toggle method, while others send meta data when the game boots and triggers the HDR mode method. as you can imagine, this makes the former game look more dull.

can someone explain to me what's going on here.

spec:
Windows 10 Home 64bit version 1803
Video card: Geforce 980Ti
CPU Ryzen 2 2700X
TV Samsung KU6300
connection HDMI

NVidia control panel setting output is set to Default. but tried YcBCR 4:2:2 10 bit limited but the same result.
Attached Thumbnails
Click image for larger version

Name:	Preference.PNG
Views:	35
Size:	27.8 KB
ID:	2457398   Click image for larger version

Name:	advanced.PNG
Views:	33
Size:	18.2 KB
ID:	2457400  

TV: Samsung 60inch 60ES8000 LED TV,
Sound System: Pioneer VSX 1120-K, 2x JBL L890 fronts, 1x JBL LC2 center, 4x JBL L820 surrounds, 1x SVS PB12 NSD for a complete 7.1
Headphones: Audio Technica ATH-AD700, AD900X, Sennheiser HD598, HD650, HD800S
Headphone DAC/AMP: Schiit Bifrost / Bottlehead Crack Tube Amp

Last edited by howzz1854; 09-20-2018 at 01:15 PM.
howzz1854 is offline  
post #10998 of 12071 Old 09-21-2018, 06:54 AM
Member
 
Join Date: Dec 2015
Posts: 158
Mentioned: 8 Post(s)
Tagged: 0 Thread(s)
Quoted: 103 Post(s)
Liked: 47
Quote:
Originally Posted by howzz1854 View Post
hey guys, i am at a loss here. i tried calibrating my HDR TV Samsung KU6300 that's connected to my Gaming machine running Windows 10 using the latest HCFR but discovered that there's a huge gamma discrepancy with Windows 10's native HDR WCG mode, vs the MadVR HDR passthrough method. the HCFR HDR10 output seem to be similar to the MadVR, in which when Meta is sent to the tv, all the desktop icons and wallpaper increases in contrast dramatically. but when i use the Windows 10's native HDR WCG toggle, the desktop elements becomes very dull. i know the TV is receiving HDR signal in both cases, but there's just a huge difference in terms of gamma curve and color mapping.

long story short, i tried to calibrate for both output method to see which one will yield the best result and found that when sending the HDR meta through MadVR/HCFR HDR10 it yields the best final calibrated result, while the Windows 10 native HDR toggle after calibrating it with HCFR (without the HCFR's internal HDR10 signal output), I can't get close to the HDR gamma/luminance reference curve and on top of that the color saturation mapping doesn't expand as far out as the MadVR/HCFR HDR10 output method. with the MadVR/HCFR, i can get very very close to the SMPTE 2084 curve, which resembles a snake shape, and the saturation coverage goes out a lot farther on the Rec2020 spectrum. but with the Windows 10 HDR output, the closest i can get the gamma curve is more inline with the regular BT 1886 reference curve, almost looks like a 2.1/2.2 gamma. needless to say, this provides less punch when watching a HDR movie or an HDR game.

when investigating more by looking at the MPC-HE/MadVR info toggle switch (Ctrl + J), i noticed that when using MadVR method, the info shows that the HDR method is "NV HDR". but when doing the Windows 10 method, it shows "OS HDR". all the other info are identical.

i would have just stick with the MadVR method except there're a lot of HDR games out there on PC, some uses the native OS HDR toggle method, while others send meta data when the game boots and triggers the HDR mode method. as you can imagine, this makes the former game look more dull.

can someone explain to me what's going on here.

spec:
Windows 10 Home 64bit version 1803
Video card: Geforce 980Ti
CPU Ryzen 2 2700X
TV Samsung KU6300
connection HDMI

NVidia control panel setting output is set to Default. but tried YcBCR 4:2:2 10 bit limited but the same result.
Not all things in Windows 10 are made for HDR of any sort. When you turn it on, if there is no HDR10 signal, it pushes the SDR content within a lower set. On Windows 10 1803, they added a slider in the HDR and WCG menu for Brightness for SDR content. This is to try to adjust the nit level for the SDR content, that way to provide better mapping and try to avoid issues related to washed out or dull looking images and icons, which happens when you turn it on with windows controller. HCFR applies the HDR signal in 8-bit, IIRC, but the signal is applied to everything, so those SDR images get put within the stretched HDR/WCG color space. So, there is an extra step when playing with the OS toggle where you should have HDR on in OS, then put up a 100% IRE white/gray pattern being sent in SDR, then adjust the brightness slider in the HDR menu so that you reach your target luminance for SDR content (usually 80-120 nit, I usually go for 120, personally), which then should give slightly better mapping so that you get closer to proper SDR content in HDR on in the OS, while the HDR content, when it is properly sent, should provide correct values. You may need to balance the settings in Nvidia control panel to make sure things are being sent properly to the display. MadVR has the settings for 0-255 and 16-234 as well, so having another pattern to make sure you are sending it properly, whatever means you choose to present that pattern with, would be good to try.

Also, if doing madVR, you may look into reshade, which can allow the use of a 3D LUT with games, although with some tradeoffs. Check it out to see if it is right for you (sorry that sounded like a pharmaceutical commercial, lol).

So, just to be clear, Windows 10's HDR mode allows SDR content to not be stretched into the HDR color space, but without the luminance correction, washes out all SDR content. It is making the effort to try. Most apps are SDR, not HDR. The other two kick it so that it presents the material as HDR, even if not HDR content. I would calibrate with the HCFR or MadVR image (newest version out 9/19), then try to do what I said above for changing the brightness level for SDR. With that, you should have better results and less washout. I have not fully worked out the kinks with OS HDR because it isn't as good as just allowing the graphics card to pass it along. But try that and you may get better results.
ajc9988 is offline  
post #10999 of 12071 Old 09-21-2018, 11:22 AM
Senior Member
 
howzz1854's Avatar
 
Join Date: Feb 2008
Posts: 409
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 15 Post(s)
Liked: 26
Quote:
Originally Posted by ajc9988 View Post
Not all things in Windows 10 are made for HDR of any sort. When you turn it on, if there is no HDR10 signal, it pushes the SDR content within a lower set. On Windows 10 1803, they added a slider in the HDR and WCG menu for Brightness for SDR content. This is to try to adjust the nit level for the SDR content, that way to provide better mapping and try to avoid issues related to washed out or dull looking images and icons, which happens when you turn it on with windows controller. HCFR applies the HDR signal in 8-bit, IIRC, but the signal is applied to everything, so those SDR images get put within the stretched HDR/WCG color space. So, there is an extra step when playing with the OS toggle where you should have HDR on in OS, then put up a 100% IRE white/gray pattern being sent in SDR, then adjust the brightness slider in the HDR menu so that you reach your target luminance for SDR content (usually 80-120 nit, I usually go for 120, personally), which then should give slightly better mapping so that you get closer to proper SDR content in HDR on in the OS, while the HDR content, when it is properly sent, should provide correct values. You may need to balance the settings in Nvidia control panel to make sure things are being sent properly to the display. MadVR has the settings for 0-255 and 16-234 as well, so having another pattern to make sure you are sending it properly, whatever means you choose to present that pattern with, would be good to try.

Also, if doing madVR, you may look into reshade, which can allow the use of a 3D LUT with games, although with some tradeoffs. Check it out to see if it is right for you (sorry that sounded like a pharmaceutical commercial, lol).

So, just to be clear, Windows 10's HDR mode allows SDR content to not be stretched into the HDR color space, but without the luminance correction, washes out all SDR content. It is making the effort to try. Most apps are SDR, not HDR. The other two kick it so that it presents the material as HDR, even if not HDR content. I would calibrate with the HCFR or MadVR image (newest version out 9/19), then try to do what I said above for changing the brightness level for SDR. With that, you should have better results and less washout. I have not fully worked out the kinks with OS HDR because it isn't as good as just allowing the graphics card to pass it along. But try that and you may get better results.

thanks for the quick reply. I really appreciate it.

I already calibrated using MadVR in OS HDR mode, and then again in MadVR with Driver (Nvidia) mode, and the best result comparing the two was exactly what I mentioned above. the OS HDR calibrated result still doesn't come close to the reference HDR Gamma Luminance curve. at best it looks like a 2.1 gamma. I was however able to the Driver HDR mode using MadVR to very very close to the reference HDR luminance curve. so this tells me that the OS HDR mode really is just f'ed up. I verified this with HCFR's internal pattern generator. the HCFR's internal HDR signal output is identical to the MadVR HDR output, in which both utilize the driver. And if I disable HCFR's internal HDR signal output, and use the OS HDR I also get the same broken result mentioned earlier.

I'd be happy to just stick with the Driver HDR mode but the issue going back to the source of why I posted this is, there're some games, like the recent release of Shadow of Tomb Raider, uses the OS HDR. So either they designed the PC release with the broken OS HDR luminance curve in place or they simply are just not aware that the OS HDR is no where near the standard. but I think it's probably both. consider these game developers 90% of the time get the color space wrong anyway. most game devs don't even include a grey scale chart when they provide you with a brightness control slider in game, what are you supposed to compare that to??

HDR is still such a mess still. seems like developers have no understanding of color space, and Microsoft doesn't either. so you have tv manufactures releasing their tvs trying to guess what reference curve their target source is using. it's just a mess.

TV: Samsung 60inch 60ES8000 LED TV,
Sound System: Pioneer VSX 1120-K, 2x JBL L890 fronts, 1x JBL LC2 center, 4x JBL L820 surrounds, 1x SVS PB12 NSD for a complete 7.1
Headphones: Audio Technica ATH-AD700, AD900X, Sennheiser HD598, HD650, HD800S
Headphone DAC/AMP: Schiit Bifrost / Bottlehead Crack Tube Amp
howzz1854 is offline  
post #11000 of 12071 Old 09-21-2018, 11:48 AM
AVS Forum Special Member
 
Join Date: Dec 2014
Location: Mississauga, ON, Canada
Posts: 6,946
Mentioned: 130 Post(s)
Tagged: 0 Thread(s)
Quoted: 4883 Post(s)
Liked: 1847
I’m not sure how the discussion migrated back to the JVC Autocal thread.

Quote:
Originally Posted by lovingdvd View Post
1. In your Preferences->References tab you have BT.2390 checked. What impact does this have and when it is appropriate to use that option vs not? How are things impacted if you don't check this box when it should be checked off?
BT.2390 implements soft clipping (highlight rolloff). If you uncheck it the curve will hard clip at the maximum luminance. You can view the behaviour easily if you display the EOTF (Luminance graph) while toggling it on and off. Unchecking it equivalent to setting sc to same value as bh in Arve's Tool. However, Arve's Tool does not use the same algorithm as BT.2390.

Quote:
2. In the References tab you selected Override Targets. For Diffuse white you have 25.6 which is the reading at 50%. How did you come up with that number? Is it the nit measurement of 50% white from the grayscale based on the selected HDR curve?
This is the same value as bw in Arves Tool. I adjust it slightly to match the measured value at 50%.

Quote:
3. How does the diffuse white number tie back into the parameters in Arve's Tool? Is that 25.6 what should be set for bm or bw? Or do you use Arve's tool to figure out what the diffuse white should be?
See 2. above for bw. TargetMaxL is the measured peak luminance, same as bm in Arve's Tool.

Quote:
4. What does the Target MaxL override value represent? How did you come up with or decide on using 120 as this value, so I know what value I should use?
See 3. above.

Quote:
5. I recall seeing a HDR10 check box somewhere in HCFR, but can't find it now - thought it was on the Preferences->References tab. What is this check box used for?
That is only used to force the graphics card into HDR mode, if you have compatible hardware. Not relevant to the discussion.

Quote:
6. The J9_HDR and J9_HDR_CMS seems to have two different HDR curves loaded? What's the difference between these two curves?
One is for 1200 nits, the other 4000 nits. The difference was not really intended for the comparison, but does not affect it since they only differ above 50% (and all the patterns are at, or below, 50%).
Dominic Chan is online now  
post #11001 of 12071 Old 09-21-2018, 01:03 PM
AVS Forum Special Member
 
Join Date: Dec 2014
Location: Mississauga, ON, Canada
Posts: 6,946
Mentioned: 130 Post(s)
Tagged: 0 Thread(s)
Quoted: 4883 Post(s)
Liked: 1847
Quote:
Originally Posted by lovingdvd View Post
Strange Arve does not implement the same BT.2390. Yet your measured gamma curve seems to follow it exactly?
Both HCFR and Arve's tool follow the ST2084 PQ curve, so they match each other until the highlight rolloff (the "knee").

zoyd added a feature to customize the rolloff. You can match Arve's curve quite closely if you like; I used NW=3, 20%. (These are related to se and sc in Arve's Tool). HCFR also implements ST2390's black compensation (the "toe") which Arve's Tool lacks.

Recently I've been generating my curves using HCFR using a Excel macro to create a jgd file from HCFR's tone curve, although you can also use the 12-point gamma control to do that manually.

Quote:
I am wondering if I can apply what I am learning/measuring through HCFR to maximize my curves creation with Arve's Tool. I know know my peak nits for 2.40 zoom are 47. This is what I'm thinking I could try but wondering if this makes sense:

1. Run Arve's tool and lp 3 to load the HDR 1200 nit curve.
2. Set the hard clip point to 1100 since I don't mind cutting it off these and gaining some brightness.
3. Set bm to 47
I would use a lower diffuse white level (e.g., 15) when bm is very low.

Quote:
4. Pw to write to pj.
5. Set Preferences->Reference in HCFR and override Target MaxL to 47.
6. Measure grayscale with HCFR and with EOTF HDR10.
7. Note my nits at 50% from grayscale and update the Preferences->References to set Diffuse White to this value.
8. Check the gamma and luminance graphs and see how well it tracks.

Anything you'd change about that approach? Does that sound like a worthwhile exercise and might I learn something from it?

Can reviewing this give me any insight into the way in which I can customize the curve? Or do I have it backward - I should customize the curve and then measure to assess what impact my customizations have?
I'm not quite sure what your objectives are. This will help you get the grey scale to match the reference curve, in case you want to finetune CMS. However, I don't think it will help you decide how to shape the highlight rolloff.

Quote:
The other big thing I'm wondering is how to intrepret these HDR curves. For instance on your J9_HDR curve, at 75% it clips and reads the same brightness up to 100%. So what does this mean exactly when watching an HDR movie? Say content is encoded at 85% brightness, is that then crushed along with everything that's about 75%, and if so, isn't that not a good thing to have? Yet it seems to be by design, so I apparently do not understand the theory of what this curve is supposed to be doing. That is what I'm trying to better understand.
I have explained this in a previous post. You need to look at the ST2084 PQ curve to understand what the % levels mean. For example, 1200 nits corresponds to about 75%; 4000 nits corresponds to about 90%.

Last edited by Dominic Chan; 09-21-2018 at 01:18 PM.
Dominic Chan is online now  
post #11002 of 12071 Old 09-21-2018, 01:36 PM
AVS Forum Addicted Member
 
lovingdvd's Avatar
 
Join Date: Jul 2002
Posts: 12,930
Mentioned: 79 Post(s)
Tagged: 0 Thread(s)
Quoted: 3659 Post(s)
Liked: 844
Quote:
Originally Posted by Dominic Chan View Post
I would use a lower diffuse white level (e.g., 15) when bm is very low.
What is a good way to determine what an appropriate diffuse white level is? I mean, I could just pick 15 like you mentioned in your example. But what's the significance of this? I still don't really understand what this has to do with anything, but know it is important. I know that the lower this number, the darker the picture and the more room for highlights, and vice versa. But I'm unsure of what guidelines to use in determining the ideal balance between the two.
[/quote]

Quote:
I'm not quite sure what your objectives are. This will help you get the grey scale to match the reference curve, in case you want to finetune CMS. However, I don't think it will help you decide how to shape the highlight rolloff.
Yes the objective was to try and see if plotting that can help me somehow understand what the curves are doing. People say they can just look at a curve and know how the picture will look. I'd like to understand the curves to the point of being able to do that.

Quote:
I have explained this in a previous post. You need to look at the ST2084 PQ curve to understand what the % levels mean. For example, 1200 nits corresponds to about 75%; 4000 nits corresponds to about 90%.
Yes I recall you talking about it, but I did not really understand it. In your example 1200 nits correspondes to about 75%. So if a title is properly mastered and MaxCLL is 1200 nits, then it doesn't matter that you crush 75% and up, because there will be (should be) no content mastered in there about 1200 nits. Looking at the curve I understand that part.

But how do you know that 1200 nits corresponds to 75%? What about for 4000 nit titles?

Let's see if I have this part now - for low nit setups like mine, if you want HDR brighter overall then you need to use a higher bw. But this means the range between bw and the clipping point has more compression. Which then means less details in the highlights and more crush?

How does the soft clip se and sc then tie into this?

Thanks!!
lovingdvd is online now  
post #11003 of 12071 Old 09-21-2018, 08:25 PM
AVS Forum Special Member
 
Join Date: Dec 2014
Location: Mississauga, ON, Canada
Posts: 6,946
Mentioned: 130 Post(s)
Tagged: 0 Thread(s)
Quoted: 4883 Post(s)
Liked: 1847
Quote:
Originally Posted by lovingdvd View Post
What is a good way to determine what an appropriate diffuse white level is? I mean, I could just pick 15 like you mentioned in your example. But what's the significance of this? I still don't really understand what this has to do with anything, but know it is important. I know that the lower this number, the darker the picture and the more room for highlights, and vice versa. But I'm unsure of what guidelines to use in determining the ideal balance between the two.
I don't know if it's possible to come up with any guidelines. I've always preferred a bw of 25, but manni and javs prefer a lower value.

Quote:
Yes the objective was to try and see if plotting that can help me somehow understand what the curves are doing. People say they can just look at a curve and know how the picture will look. I'd like to understand the curves to the point of being able to do that.
Yes, that's very useful but you can also look at the plot in Arve's Tool without introducing another player.

Quote:
Yes I recall you talking about it, but I did not really understand it. In your example 1200 nits correspondes to about 75%. So if a title is properly mastered and MaxCLL is 1200 nits, then it doesn't matter that you crush 75% and up, because there will be (should be) no content mastered in there about 1200 nits. Looking at the curve I understand that part.

But how do you know that 1200 nits corresponds to 75%? What about for 4000 nit titles?
I have explained this before. If you look at the PQ curve I posted, 1200 nits corresponds to ~75% input and 4000 nits to ~90% input. With tone mapping, the vertical scale is compressed, but the horizontal scale remains unaffected.

Quote:
Let's see if I have this part now - for low nit setups like mine, if you want HDR brighter overall then you need to use a higher bw. But this means the range between bw and the clipping point has more compression. Which then means less details in the highlights and more crush?

How does the soft clip se and sc then tie into this?
sc determine where the compression starts, i.e., where the tone curve starts departing from the PQ curve. It should never be lower than the reference white (100 nits) as otherwise you get tonal distortion in the regular (non-highlight) region. With limited headroom I would set sc very close to 100. (Arve's Tool would use 100 for sc if bm is 50 or less).
se determines the end slope of the curve, which in turn determines allocation of nits between "low highlight" and "high highlight". In one extreme case, se=1 means a horizontal end-slope, i.e., most of the nits are allocated to the low highlighs (100-400 nits), and there's little gradation between 400 nits and 1200 nits. In the other extreme, se=1 gives most gradation to the high highlights, at the expense of the low highlights.
Dominic Chan is online now  
post #11004 of 12071 Old 09-21-2018, 08:50 PM
AVS Forum Addicted Member
 
lovingdvd's Avatar
 
Join Date: Jul 2002
Posts: 12,930
Mentioned: 79 Post(s)
Tagged: 0 Thread(s)
Quoted: 3659 Post(s)
Liked: 844
Quote:
Originally Posted by Dominic Chan View Post
I don't know if it's possible to come up with any guidelines. I've always preferred a bw of 25, but manni and javs prefer a lower value.


Yes, that's very useful but you can also look at the plot in Arve's Tool without introducing another player.


I have explained this before. If you look at the PQ curve I posted, 1200 nits corresponds to ~75% input and 4000 nits to ~90% input. With tone mapping, the vertical scale is compressed, but the horizontal scale remains unaffected.



sc determine where the compression starts, i.e., where the tone curve starts departing from the PQ curve. It should never be lower than the reference white (100 nits) as otherwise you get tonal distortion in the regular (non-highlight) region. With limited headroom I would set sc very close to 100. (Arve's Tool would use 100 for sc if bm is 50 or less).
se determines the end slope of the curve, which in turn determines allocation of nits between "low highlight" and "high highlight". In one extreme case, se=1 means a horizontal end-slope, i.e., most of the nits are allocated to the low highlighs (100-400 nits), and there's little gradation between 400 nits and 1200 nits. In the other extreme, se=1 gives most gradation to the high highlights, at the expense of the low highlights.
Thanks Dominic. Wow I think this is actually starting to sink in for me! So sc sets the line between normal gamma and where it diverges for the HDR part of the image. Effectively then the difference you leave between sc and bm defines the headroom that one has for HDR-like effects, yes? And se defines the balance between low highlights vs high highlights.

Can you provide a few examples (descriptive or screenshots) that demonstrate the types of things that are considered low highlights vs high highlights? For instance a car's headlights or an explosion - those would be considered high highlights? And things such as a shimmer or reflection off of an object would be considered low highlights (just guessing here)? What other types of highlights would be examples of low ones?

Finally regarding bw - this is the one parameter I'm not quite sure I understand it's effect - other than knowing that a higher bm makes for a brighter picture and lower makes a dimmer picture. But ultimately what is the impact of a lower bm vs higher bm with regards to the rest of the picture and HDR qualities, besides just making for a brighter/dimmer picture overall? Why do you prefer 25, and any idea why Javs and Manni prefer lower?
lovingdvd is online now  
post #11005 of 12071 Old 09-22-2018, 04:17 AM
Advanced Member
 
jqmn's Avatar
 
Join Date: Apr 2005
Location: Houston, TX
Posts: 792
Mentioned: 6 Post(s)
Tagged: 0 Thread(s)
Quoted: 433 Post(s)
Liked: 155
Is it fair to say it comes down to how much of your light you want to allocate to highlights and how much to the regular picture? It also may depend on the type of content you watch. Some people are looking at 1K vs 4K curves but it might make more sense to think about it in terms of films with a lot of specular highlights (sci-fi) vs films that are more regular (Dunkirk) and have a setting based on that. If we believe the literature and the APL is still in the up to 100 range or so then reserving so much for highlights really brings the APL down. If you like around 14ftL that would make sense. When I was looking at levels down around 25 I found the picture pleasing but the APL just too dim. With BD's I moved to near 100 nits on the theory that is what the discs are mastered at and for HDR this translated (for me) to a general purpose curve with DW around 45. This also allowed the low-end to better track the curve up to about 30%, not crush anything and there is still a balance to highlights.



Quote:
Originally Posted by Dominic Chan View Post
I don't know if it's possible to come up with any guidelines. I've always preferred a bw of 25, but manni and javs prefer a lower value.
jqmn is offline  
post #11006 of 12071 Old 09-22-2018, 05:06 AM
Senior Member
 
ebr9999's Avatar
 
Join Date: Jun 2013
Posts: 394
Mentioned: 9 Post(s)
Tagged: 0 Thread(s)
Quoted: 228 Post(s)
Liked: 92
RPG (RPI Pattern Generator by Biasi)


For those interested to RPG (RPI Pattern Generator by Biasi): http://translate.google.com/translat...rator-by-biasi

TV: LG 55E8PLA, Pana TX-L37ET5E BD/DVR: Panasonic BMR-BWT735 Media player: Teufel Connector 2, HT Rec: ANTHEM MRX 510,
Front: Woofer&Tweeter, Center: 2Woofer&1Tweeter, Surround: Bose acoustimass 5 series II, Sub: SVS SB2000
ebr9999 is offline  
post #11007 of 12071 Old 09-22-2018, 12:43 PM
AVS Forum Special Member
 
Join Date: Dec 2014
Location: Mississauga, ON, Canada
Posts: 6,946
Mentioned: 130 Post(s)
Tagged: 0 Thread(s)
Quoted: 4883 Post(s)
Liked: 1847
Quote:
Originally Posted by lovingdvd View Post
Thanks Dominic. Wow I think this is actually starting to sink in for me! So sc sets the line between normal gamma and where it diverges for the HDR part of the image. Effectively then the difference you leave between sc and bm defines the headroom that one has for HDR-like effects, yes? And se defines the balance between low highlights vs high highlights.
Getting close. "Normal" video is the region at and below bw. You may want to continue on the PQ curve a little further for the highlights, until you hit sc, at which point you start diverging. So the HDR effect is between bw and bm, not just between sc and bm, although in your case sc is very close to bw.
Dominic Chan is online now  
post #11008 of 12071 Old 09-22-2018, 05:39 PM
AVS Forum Addicted Member
 
lovingdvd's Avatar
 
Join Date: Jul 2002
Posts: 12,930
Mentioned: 79 Post(s)
Tagged: 0 Thread(s)
Quoted: 3659 Post(s)
Liked: 844
Quote:
Originally Posted by Dominic Chan View Post
...You may want to continue on the PQ curve a little further for the highlights, until you hit sc, at which point you start diverging.
What parameter(s) would I change in order to continue on the PQ curve a little further?

Quote:
So the HDR effect is between bw and bm, not just between sc and bm, although in your case sc is very close to bw.
What does the range between bm and sc control?
lovingdvd is online now  
post #11009 of 12071 Old 09-22-2018, 05:51 PM
AVS Forum Special Member
 
Join Date: Dec 2014
Location: Mississauga, ON, Canada
Posts: 6,946
Mentioned: 130 Post(s)
Tagged: 0 Thread(s)
Quoted: 4883 Post(s)
Liked: 1847
Quote:
Originally Posted by lovingdvd View Post
What parameter(s) would I change in order to continue on the PQ curve a little further?
To continue on PQ further you would just raise sc. I thought my explanation was clear, but maybe not.

Quote:
Originally Posted by Dominic Chan View Post
sc determine where the compression starts, i.e., where the tone curve starts departing from the PQ curve.
Quote:
Originally Posted by lovingdvd View Post
What does the range between bm and sc control?
The “high highlights”.
Really seem to be going in a circle

Last edited by Dominic Chan; 09-22-2018 at 06:22 PM.
Dominic Chan is online now  
post #11010 of 12071 Old 09-22-2018, 10:16 PM
AVS Forum Addicted Member
 
lovingdvd's Avatar
 
Join Date: Jul 2002
Posts: 12,930
Mentioned: 79 Post(s)
Tagged: 0 Thread(s)
Quoted: 3659 Post(s)
Liked: 844
Quote:
Originally Posted by Dominic Chan View Post
To continue on PQ further you would just raise sc. I thought my explanation was clear, but maybe not.



The “high highlights”.
Really seem to be going in a circle
Sounds good. Thanks for all the help - much appreciated.

After much testing I have concluded that it is MUCH faster and easier to just do the autocal and then measure of the screen afterward with the i1D3 and set the RGB at 100% (I tried 80% but liked the results better at 100%). As opposed to use a custom color profile with just the W xy target changed. The results are excellent and indistinguishable from using the custom color profile. No CMS required in my case. I have four different RGB gains set (one in 6500K, and then 2 from Custom 1-2). These are for high lamp in 16:9 and 2.40, and low lamp in 16:9, all with BT2020F. I also did a new curve specifically for 16:9 HDR, where I have about 70 nits so significantly more headroom than my 46 nit 2.40 setup. The results are excellent.
lovingdvd is online now  
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