MadVR - ArgyllCMS - Page 173 - AVS Forum | Home Theater Discussions And Reviews
Forum Jump: 
 593Likes
Reply
 
Thread Tools
post #5161 of 5181 Old 06-16-2019, 03:10 PM
AVS Forum Special Member
 
3ll3d00d's Avatar
 
Join Date: Sep 2007
Location: London, UK
Posts: 3,813
Mentioned: 239 Post(s)
Tagged: 0 Thread(s)
Quoted: 2286 Post(s)
Liked: 2372
Quote:
Originally Posted by Manni01 View Post
You don't want the JVC to be in a BT2020 color profile. You want to select a DCI-P3 color profile.



A BT2020 color profile will give you huge errors outside of DCI-P3. There is no advantage selecting such a wider gamut with madVR. It only increases errors during verification (unless you measure DCI-P3 within BT2020) and often causes posterization issues.
I am referring to the colour profile in the JVC picture mode menu. I was under the impression BT.2020 was the one that puts the filter in place. Which option is a dci-p3 profile in the JVC?
3ll3d00d is online now  
Sponsored Links
Advertisement
 
post #5162 of 5181 Old 06-16-2019, 03:31 PM
AVS Forum Special Member
 
Manni01's Avatar
 
Join Date: Sep 2008
Posts: 9,011
Mentioned: 307 Post(s)
Tagged: 0 Thread(s)
Quoted: 5360 Post(s)
Liked: 5412
Quote:
Originally Posted by 3ll3d00d View Post
I am referring to the colour profile in the JVC picture mode menu. I was under the impression BT.2020 was the one that puts the filter in place. Which option is a dci-p3 profile in the JVC?
Reference on older models (from the rs500) and DCI-P3 on the newest. They do use the filter, but they target DCI saturations, unlike BT2020.

The HDR color profile also uses BT2020 saturations (without the filter), so not useful to achieve this.

You can also sim0ly create a custom color profile with DCI-P3 coordinates, using the filter, and upload it with the JVC Autocal software (there is no need to have a Spyder to do this).

Please ask further questions on the JVCs in the calibration thread related to your JVC model as it’s getting a bit OT here.
3ll3d00d likes this.

JVC Autocal Software V11 Calibration for 2019 Models (Google)
Batch Utility V4.02 May 16 2019 to automate measurements files for madVR with support for BD Folders
Manni01 is online now  
post #5163 of 5181 Old 06-17-2019, 12:27 AM
AVS Forum Special Member
 
3ll3d00d's Avatar
 
Join Date: Sep 2007
Location: London, UK
Posts: 3,813
Mentioned: 239 Post(s)
Tagged: 0 Thread(s)
Quoted: 2286 Post(s)
Liked: 2372
Thanks for clarifying. I can see now that I misread your comment about madvr setup in your autocal thread.

One q re how displaycal interacts with madtpg and the madvr config for this use case.

As I understand it that conversion from bt2020 to dci-p3 occurs when madvr finds a lut in the dci-p3 slot or when you use the "this display is already calibrated to" option.

When measuring, displaycal disables the 3dlut via madtpg so I think that means I have to manually set the "this display is already calibrated to" option to dci-p3 and whatever gamma target I have set.

Is that right?
3ll3d00d is online now  
Sponsored Links
Advertisement
 
post #5164 of 5181 Old 06-17-2019, 03:20 AM
AVS Forum Special Member
 
Manni01's Avatar
 
Join Date: Sep 2008
Posts: 9,011
Mentioned: 307 Post(s)
Tagged: 0 Thread(s)
Quoted: 5360 Post(s)
Liked: 5412
Quote:
Originally Posted by 3ll3d00d View Post
Thanks for clarifying. I can see now that I misread your comment about madvr setup in your autocal thread.

One q re how displaycal interacts with madtpg and the madvr config for this use case.

As I understand it that conversion from bt2020 to dci-p3 occurs when madvr finds a lut in the dci-p3 slot or when you use the "this display is already calibrated to" option.

When measuring, displaycal disables the 3dlut via madtpg so I think that means I have to manually set the "this display is already calibrated to" option to dci-p3 and whatever gamma target I have set.

Is that right?
I don't use DC to calibrate/verify, so I'll let @fhoech or others answer that one

With Calman (sorry for the OT, just in case it helps):

When I calibrate, I select DCI-P3 as a target, and madTPG automatically selects DCI-P3 patterns. 3D LUT is unchecked in Calman, so the 3D LUT isn't applied. The 3D LUT is uploaded into the DCI-P3 slot at the end of the calibration.

When I verify, I simply enable the 3D LUT option in Calman, the selected 3D LUT is enabled in madTPG, and I can verify the calibration that way. It's entirely automatic.

Whether I calibrate or verify, it's my SDR profile that is selected/enabled with madVR according to my profile rules, and the "calibrate using a 3D LUT" option is selected. I only select/load a rec-709 and a DCI-P3 LUT. If you select a BT2020 LUT, it will take precedence and be used when playing HDR content, even if a DCI-P3 LUT is available, due to the container being BT2020. If no BT2020 LUT is avaialble, madVR discards the BT2020 container and plays the P3 content (as 99% of the time it was mastered using P3 primaries, not BT2020). It's only when the content is actually BT2020 that it's converted to P3 (assuming that you have a P3 LUT loaded of course).

Apart from making sure that BT2020 is enabled in the calibration tab if you use 12bits with nVidia, there is nothing to do.
3ll3d00d likes this.

JVC Autocal Software V11 Calibration for 2019 Models (Google)
Batch Utility V4.02 May 16 2019 to automate measurements files for madVR with support for BD Folders
Manni01 is online now  
post #5165 of 5181 Old 06-17-2019, 03:54 AM
Senior Member
 
chros73's Avatar
 
Join Date: Jan 2015
Posts: 416
Mentioned: 10 Post(s)
Tagged: 0 Thread(s)
Quoted: 241 Post(s)
Liked: 105
Quote:
Originally Posted by cezarL View Post
I'm looking to generate 3D LUTs with DisplayCAL for both SDR and HDR (passthrough mode, not converting to SDR - but ideally it would be good if madVR tonemapped everything above 750nits).
Trouble is DisplayCAL has a lot of options available under "3D LUT" and I don't know whether I should go with its default madVR presets or if I should be adjusting them in any way?
I don't think it's possible and wise to create a 3dlut for HDR10 passthrough. There's no option for this in latest madvr stable (there was in 0.92.16).

As far as I know we can only create 3dlut for SDR content: for this just use the default madVR BT.709 SDR preset.

I have an LG B8 and I think the only way to calibrate HDR10 on it to use Calman that would set the certain settings (grayscale values and just 3 (!) RBG values) in the TV itself. I couldn't try it because you'd have to buy it.

If someone knows otherwise, let us know

Quote:
Originally Posted by cezarL View Post
There is one other thing to consider:
Panasonic have implemented Dynamic LUTs which makes standard clipping pluge patterns unusable - first 3 levels of black and last 3 levels of white are clipped on patterns.
Is it only in HDR mode or SDR as well? The latter would indicate that it's some sort of dynamic contrast.

Ryzen 5 2600,Asus Prime b450-Plus,16GB,MSI GTX 1060 Gaming X 6GB(v385.28),Win10 LTSB 1607,MPC-BEx64+LAV+MadVR,Yamaha RX-A870,LG OLED65B8(04.10.25+PC4:4:[email protected]/24/25/29/30/50/59/60Hz)
chros73 is online now  
post #5166 of 5181 Old 06-17-2019, 04:39 AM
AVS Forum Special Member
 
mightyhuhn's Avatar
 
Join Date: Nov 2013
Posts: 2,141
Mentioned: 11 Post(s)
Tagged: 0 Thread(s)
Quoted: 1157 Post(s)
Liked: 423
the current madVR version clearly wants that to work. why else would it like to know the HDR meta data you want to send to the screen... if 0.92.17 doesn't work use 0.92.16

the DV point is meaningless for madVR madVR doesn't support it and the only meaningful way to handle DV is to decode it and send an HDR 10 signal after that if you want to calibrate the image. it's like you want to process bitstream audio...

the WOLED argument has it points no question they are well known i still don't see why this should be limited to HDR it clearly effects SDR in a similar way not every one calibrates for 100 nits where an OLED is stable not everyone watches in an dark room simple as that. if you want an accurate image ABL and APL are your enemy nothing new and both should be avoid or disabled for 3D LUT calibration.
beware that your post comes from an LG OLED calibration thread. using an very small test pattern size should still give as accurate as possible results in the end the TV does what ever it wants with it anyway...

beware that HDR process 3D LUT ware meant to overwrite the meta data that clearly helps at least that's how i understand them.
mightyhuhn is online now  
post #5167 of 5181 Old 06-17-2019, 04:42 AM
AVS Forum Special Member
 
3ll3d00d's Avatar
 
Join Date: Sep 2007
Location: London, UK
Posts: 3,813
Mentioned: 239 Post(s)
Tagged: 0 Thread(s)
Quoted: 2286 Post(s)
Liked: 2372
Quote:
Originally Posted by Manni01 View Post

With Calman (sorry for the OT, just in case it helps):



When I calibrate, I select DCI-P3 as a target, and madTPG automatically selects DCI-P3 patterns. 3D LUT is unchecked in Calman, so the 3D LUT isn't applied. The 3D LUT is uploaded into the DCI-P3 slot at the end of the calibration.
When you say, you select a target do you mean in calman or somewhere else?

If in calman then I think this bit is different in displaycal in that I don't have a target to select during profiling (or I don't know how to do it). The docs make no mention of this either as far as I can see.
3ll3d00d is online now  
post #5168 of 5181 Old 06-17-2019, 04:53 AM
AVS Forum Special Member
 
Manni01's Avatar
 
Join Date: Sep 2008
Posts: 9,011
Mentioned: 307 Post(s)
Tagged: 0 Thread(s)
Quoted: 5360 Post(s)
Liked: 5412
Quote:
Originally Posted by 3ll3d00d View Post
When you say, you select a target do you mean in calman or somewhere else?

If in calman then I think this bit is different in displaycal in that I don't have a target to select during profiling (or I don't know how to do it). The docs make no mention of this either as far as I can see.
Yes, in Calman. Calman fully controls madTPG/madVR as a source/display.

JVC Autocal Software V11 Calibration for 2019 Models (Google)
Batch Utility V4.02 May 16 2019 to automate measurements files for madVR with support for BD Folders
Manni01 is online now  
post #5169 of 5181 Old 06-17-2019, 05:14 AM
AVS Forum Special Member
 
fhoech's Avatar
 
Join Date: Apr 2013
Location: Stuttgart, Germany
Posts: 1,047
Mentioned: 72 Post(s)
Tagged: 0 Thread(s)
Quoted: 572 Post(s)
Liked: 378
During profiling, there are no "targets" in madTPG - everything needs to be set to passthrough (no 3D LUT, no videoLUT), and this is done automatically.

During verification, you need to enable the 3D LUT (obviously) - in DisplayCAL this is done on the verification tab.
Manni01 likes this.

DisplayCAL - Display Calibration and Characterization powered by ArgyllCMS
Current stable version 3.8.3 released 2019-07-02 | Changelog
DisplayCAL on Facebook
fhoech is offline  
post #5170 of 5181 Old 06-17-2019, 05:35 AM
AVS Forum Special Member
 
3ll3d00d's Avatar
 
Join Date: Sep 2007
Location: London, UK
Posts: 3,813
Mentioned: 239 Post(s)
Tagged: 0 Thread(s)
Quoted: 2286 Post(s)
Liked: 2372
Quote:
Originally Posted by fhoech View Post
During profiling, there are no "targets" in madTPG - everything needs to be set to passthrough (no 3D LUT, no videoLUT), and this is done automatically.



During verification, you need to enable the 3D LUT (obviously) - in DisplayCAL this is done on the verification tab.
Ok but what does this mean for madvr and the display device? Is my earlier post correct that I have to tell madvr "display is already calibrated to dci-p3" and put the display into a dci-p3 mode?
3ll3d00d is online now  
post #5171 of 5181 Old 06-17-2019, 05:52 AM
Newbie
 
cezarL's Avatar
 
Join Date: Apr 2011
Posts: 5
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 0
@chros73 , @mightyhuhn


Thank you both for taking the time to respond.
For the time being, even a 3D LUT for SDR is giving me headaches, I don't even dare think about HDR


My first attempt at generating something for SDR failed miserably... I went wild with the WB 2-point adjustments and ended up with a terrible picture. Oh well, at least this way I learned that one shouldn't increase the red gain, nor muck about with the cuttoffs for any channel.
Half a day later (after reading some 1000 posts from this thread) I gave it another try. Looks somewhat better but I'm still not happy with:
1) how it elevates near-blacks - Yes, I was hoping the 3D LUT would help with the minor black crush on the set, but dunno... maybe I'm not used to this but it looks off to me in mostly dark scenes.
2) a slight greenish tint to skin tones - At the initial whitepoint check, I had to set the main green gain to -1, main blue gain to -2, 100IRE green gain to -2, 100IRE blue gan to +2 so as to give me as little of a DeltaE value as possible to start with.
Now, my feeling is the 3D LUT somehow overcompensates for these negative gains to green and overshoots the results. Just a guess as I have no experience whatsoever.


Also, the profile self check DeltaE*76 average sounds wrong to me: average 0.52, max 5.09, RMS 0.65?
(I'm basing this on the general consensus from reviewers that the Panasonics are closest to reference out-of-the-box among the 2018 crop of OLEDs - guess I expected lower average and max values?)



I could upload the calibration files if anyone would be kind enough to take a look in case they can spot problems. Just need to know which files to provide


I started with the madVR preset for Rec.709 but with these deviations from defaults:
- enabled white level drift compensation
- added full field pattern insertion (every 60 seconds, for 10 seconds, 0%)
- selected the LG OLED 6-Series correction profile - wonder if I should've went with 'none' instead?
- auto-optimized test chart with 2060 patches, with the "maximize brightness differences" option (think that's how it's labeled)
If any of the above are wrong, hopefully someone will chime in with suggestions



I'm doing another run now, with the same settings but without any whitepoint adjustments - I'm curious to see how the new 3D LUT will compare to the previous one...
cezarL is offline  
post #5172 of 5181 Old 06-17-2019, 06:06 AM
Newbie
 
cezarL's Avatar
 
Join Date: Apr 2011
Posts: 5
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 0
I can't edit the above post for some reason. Just wanted to add:
In madTPG, I went with 10% image area, 20% linear light APL background, full screen.
cezarL is offline  
post #5173 of 5181 Old 06-17-2019, 08:07 AM
Newbie
 
cezarL's Avatar
 
Join Date: Apr 2011
Posts: 5
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 0
Hmmm... Pretty much the same behaviour - elevated shadows and "greenish" (or better said, less warm) skin tones - even without touching WB controls on the TV.

Don't know how to "interpret" this, though.
Could it be a screwed up LUT because of the settings I used?
Is the meter at fault, or the lack of a correction profile for my TV?
Or am I now really seeing a calibrated rendition, but being used to the, erm, uncalibrated image, I find it "incorrect"?


TBH the latter is a bit hard to believe, because pro calibrators have stated that hardware calibration brings only marginal improvements to the overall PQ of the FZ series. Unless, of course, my set has a problem that I wasn't aware of and is way off standard at default settings.
cezarL is offline  
post #5174 of 5181 Old 06-17-2019, 01:38 PM
AVS Forum Special Member
 
3ll3d00d's Avatar
 
Join Date: Sep 2007
Location: London, UK
Posts: 3,813
Mentioned: 239 Post(s)
Tagged: 0 Thread(s)
Quoted: 2286 Post(s)
Liked: 2372
Quote:
Originally Posted by fhoech View Post
During profiling, there are no "targets" in madTPG - everything needs to be set to passthrough (no 3D LUT, no videoLUT), and this is done automatically.
I ran a reasonably quick (~400 patches) profile and v quick 3dlut verification (~50 patches) with the projector in DCI mode & madvr set to "this display is already calibrated to DCI-P3" and then again leaving it with the 3dlut settings in place when profiling started. Resulting verification report is v similar, the latter is slightly better (maybe 0.3 dE better on average) though.

so ultimately I think I do not understand what is in the profile patch set & how that is interpreted by the projector. I can see, according to the editor, that a patchset is defined in RGB terms so I guess this means there is no conversion done at all in madvr at this stage, the display just interprets this according to whatever gamut is selected in the projector at the time.

fwiw verification after profiling with "display is already calibrated"



verification with no config change

Attached Thumbnails
Click image for larger version

Name:	dci.png
Views:	92
Size:	297.0 KB
ID:	2581190   Click image for larger version

Name:	madtpg.png
Views:	91
Size:	294.9 KB
ID:	2581192  
3ll3d00d is online now  
post #5175 of 5181 Old 06-17-2019, 02:42 PM
AVS Forum Special Member
 
fhoech's Avatar
 
Join Date: Apr 2013
Location: Stuttgart, Germany
Posts: 1,047
Mentioned: 72 Post(s)
Tagged: 0 Thread(s)
Quoted: 572 Post(s)
Liked: 378
Quote:
Originally Posted by 3ll3d00d View Post
3dlut verification (~50 patches) with the projector in DCI mode & madvr set to "this display is already calibrated to DCI-P3"
When you want to verify the 3D LUT, you obviously have to select "via external 3d lut" under "this display is already calibrated" in madVR (and enable 3D LUT on the "verification" tab in DisplayCAL).

Quote:
I can see, according to the editor, that a patchset is defined in RGB terms so I guess this means there is no conversion done at all in madvr at this stage, the display just interprets this according to whatever gamut is selected in the projector at the time.
Correct.
3ll3d00d likes this.

DisplayCAL - Display Calibration and Characterization powered by ArgyllCMS
Current stable version 3.8.3 released 2019-07-02 | Changelog
DisplayCAL on Facebook

Last edited by fhoech; 06-17-2019 at 02:47 PM.
fhoech is offline  
post #5176 of 5181 Old 06-17-2019, 02:51 PM
AVS Forum Special Member
 
3ll3d00d's Avatar
 
Join Date: Sep 2007
Location: London, UK
Posts: 3,813
Mentioned: 239 Post(s)
Tagged: 0 Thread(s)
Quoted: 2286 Post(s)
Liked: 2372
Quote:
Originally Posted by fhoech View Post
When you want to verify the 3D LUT, you obviously have to select "via external 3d lut" under "this display is already calibrated" in madVR (and enable 3D LUT on the "verification" tab in DisplayCAL).

.
Yes that is what I did, the only difference was whether I profiled with some particular config in the madvr config page for calibration. Anyway now I can see this is not relevant, it is just a passthrough to the projector. Thanks for clearing this up for me.

I notice it reports I reach about 98.5% of DCI-P3, are there any other tweaks one should/can do to account for not quite reaching the target gamut?

Last edited by 3ll3d00d; 06-17-2019 at 02:57 PM.
3ll3d00d is online now  
post #5177 of 5181 Old 06-17-2019, 03:02 PM
AVS Forum Special Member
 
fhoech's Avatar
 
Join Date: Apr 2013
Location: Stuttgart, Germany
Posts: 1,047
Mentioned: 72 Post(s)
Tagged: 0 Thread(s)
Quoted: 572 Post(s)
Liked: 378
Quote:
Originally Posted by 3ll3d00d View Post
I notice it reports I reach about 98.5% of DCI-P3, are there any other tweaks one should/can do to account for not quite reaching the target gamut?
At 98.5% coverage, this is hardly a concern.

If the above verification results are with 3D LUT though, I'm certain that this could be improved - long-term display stability over the course of the measurements is probably the issue here, so reduce the profiling patch count, and enable white level drift comp.

DisplayCAL - Display Calibration and Characterization powered by ArgyllCMS
Current stable version 3.8.3 released 2019-07-02 | Changelog
DisplayCAL on Facebook
fhoech is offline  
post #5178 of 5181 Old 06-18-2019, 12:27 AM
AVS Forum Special Member
 
3ll3d00d's Avatar
 
Join Date: Sep 2007
Location: London, UK
Posts: 3,813
Mentioned: 239 Post(s)
Tagged: 0 Thread(s)
Quoted: 2286 Post(s)
Liked: 2372
Quote:
Originally Posted by fhoech View Post

If the above verification results are with 3D LUT though, I'm certain that this could be improved - long-term display stability over the course of the measurements is probably the issue here, so reduce the profiling patch count, and enable white level drift comp.
It is with a lut. I can try that, might be worth noting the same config (patch count) yields much lower dE with a rec.709 correction. The only difference is the iris is opened up for the dci-p3 run. Would this affect display stability?
3ll3d00d is online now  
post #5179 of 5181 Old 06-18-2019, 09:41 AM
Newbie
 
cezarL's Avatar
 
Join Date: Apr 2011
Posts: 5
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 0
Phew!
Figured out what the problem was with my attempts: the patterns generated by the Auto-optimized option.
Well, maybe not their contents (dunno how this is determined), but possibly 2060 was too large a number of patches for OLED and it became unstable.
After further digging back in the thread, I found an old post (#3360 - I'm not allowed to insert links yet) and went with the suggested "Large testchart for LUT profiles" (778 patches) at 15% window size with black background, plus full field pattern insertion for 10 seconds every 2 minutes.

Much better results now: DeltaE*76 average 0.05, max 1.60, RMS 0.16

Verification with "Extra Large verification testchart" produces a calibration report that I can use as a baseline for tinkering a bit with gamma adjustments and maybe further WB controls...
cezarL is offline  
post #5180 of 5181 Old 07-03-2019, 02:02 AM
AVS Forum Special Member
 
fhoech's Avatar
 
Join Date: Apr 2013
Location: Stuttgart, Germany
Posts: 1,047
Mentioned: 72 Post(s)
Tagged: 0 Thread(s)
Quoted: 572 Post(s)
Liked: 378
Quote:
Originally Posted by 3ll3d00d View Post
It is with a lut. I can try that, might be worth noting the same config (patch count) yields much lower dE with a rec.709 correction. The only difference is the iris is opened up for the dci-p3 run. Would this affect display stability?
That's likely. You can try and work around this with white level drift compensation to some extent, and/or use pattern insertion.
3ll3d00d likes this.

DisplayCAL - Display Calibration and Characterization powered by ArgyllCMS
Current stable version 3.8.3 released 2019-07-02 | Changelog
DisplayCAL on Facebook
fhoech is offline  
post #5181 of 5181 Old 07-04-2019, 12:32 AM
Member
 
Michael2907's Avatar
 
Join Date: Nov 2007
Posts: 31
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 20 Post(s)
Liked: 0
Good Morning,
I would like to use the MadVR Test Pattern Generator with HCFR, what do I have to adjust in saturation and Color Checker for it to work?
And one more question, how exactly are the values ​​that come out then.
Michael2907 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