LG OLED's 3D LUT Profiling using LightSpace Thread - Page 11 - AVS Forum | Home Theater Discussions And Reviews
Forum Jump: 
 203Likes
Reply
 
Thread Tools
post #301 of 385 Old 08-22-2019, 08:34 AM
Advanced Member
 
Anger.miki's Avatar
 
Join Date: Jan 2018
Posts: 645
Mentioned: 18 Post(s)
Tagged: 0 Thread(s)
Quoted: 383 Post(s)
Liked: 410
After numerous tests carried out by me, @BlackJoker and @sat4all , we came to the conclusion that the new hybrid and hint features don’t work as we expected, at least on our C8s. Please test the next iteration more, because (especially “hint”) those are some great functions that would likely change the game. Thanks.

TVs: Pioneer PDP-LX5090H, LG OLED55C8PLA | SintoAmp: Pioneer VSX-921 | BD Player: Panasonic DMP-BDT260EG | External LUT box: Entertainment Experience eeColor | Softwares: Light Illusion Lightspace HTP, Portrait Displays CalMAN Home Enthusiast 2018 R3, HCFR, DisplayCAL | Probes: x-rite i1 Pro 2 - i1 Display Pro OEM B-02, basICColor DISCUS | Test Pattern Generator: DVDO AVLab TPG

Last edited by Anger.miki; 08-22-2019 at 01:53 PM.
Anger.miki is offline  
Sponsored Links
Advertisement
 
post #302 of 385 Old 08-22-2019, 01:59 PM
Advanced Member
 
Anger.miki's Avatar
 
Join Date: Jan 2018
Posts: 645
Mentioned: 18 Post(s)
Tagged: 0 Thread(s)
Quoted: 383 Post(s)
Liked: 410
To be more specific: at the moment peak chroma gives better results than hybrid on OLED.

While a LUT created on a QP Primary + Secondary (or even a 10^3 profile) with a 21^3 profile (measured with a Klein K10A) as “Hint” didn’t increase the accuracy of a LUT based on the same QP with no hint.

As soon as I get the time, I’ll test both features with my Kuro.

TVs: Pioneer PDP-LX5090H, LG OLED55C8PLA | SintoAmp: Pioneer VSX-921 | BD Player: Panasonic DMP-BDT260EG | External LUT box: Entertainment Experience eeColor | Softwares: Light Illusion Lightspace HTP, Portrait Displays CalMAN Home Enthusiast 2018 R3, HCFR, DisplayCAL | Probes: x-rite i1 Pro 2 - i1 Display Pro OEM B-02, basICColor DISCUS | Test Pattern Generator: DVDO AVLab TPG
Anger.miki is offline  
post #303 of 385 Old 08-23-2019, 06:39 AM
Senior Member
 
ebr9999's Avatar
 
Join Date: Jun 2013
Posts: 387
Mentioned: 8 Post(s)
Tagged: 0 Thread(s)
Quoted: 226 Post(s)
Liked: 86
Following my previous post, I have decided not to use automatic 1DLUT generation. Doing that I have had a chance of using the new feature Hint, for my case with full success.

I have prepared everything for generating a 17 cube (4/5 hours with my I1D3). My panel (oled light 25) stays off till BR=52. Anyhow I have started with BR=51.
I have got good result, but poor near black details. I have had to rise BR to 52 and increase luminance at 05 IRE to 15.
So I decided: let's have another test with BR=52, but I have made a mistake: I have forgotten 05 Luminance at 15. The result was overall good, but with a near black disaster (you can easily undesrstand why).
So, I wondered: why spending another night for that? Let's try to use hint, using as base a Large grey scale, done on the correct 3DLUT settings. I was surprised of results: nice near black. I have checked for artefacts and banding, nothing to report.

Here measurements:
RGB Balance and Diff Gamma

Finally something in line with what I can achieve manually.
Overall dE2000 on a profile with saturations and a color checker

You can see how only a point with dE=1.37 has been measured (grey scale at 17 IRE). All the other ones below 1.

Next step: I will generate a 3Dcube to check if I get the same results. Another night ...
Another thought for discussion: Calman achieve good grey scale balance. I understand this is somehow related to their interactive approach, that makes them having good result on any pannel. I don't expect LS doing the some, but something going in that direction, e.g. a two pass process (something working better than augmented data), would be, in my opinion, needed.
L30Z3N and mrtickleuk like this.

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  
Sponsored Links
Advertisement
 
post #304 of 385 Old 08-23-2019, 07:50 AM
AVS Forum Special Member
 
Light Illusion's Avatar
 
Join Date: Aug 2010
Posts: 1,806
Mentioned: 27 Post(s)
Tagged: 0 Thread(s)
Quoted: 764 Post(s)
Liked: 1123
You can already do a 2-stage 1D LUT (or actually x-stage 1D LUT).

Do the first 1D LUT as standard.
Re-profile with the 1D LUT active in the TV.
Concatenate the first 1D LUT with the 2nd 1D LUT, and upload the resulting 1D LUT into the TV.
Repeat as needed...

Unlike 3D LUTs, where re-profiling with the LUT active to make a 2nd 3D LUT will not work, due to Gamut reduction, a 1D LUT has no such issues, so can be done basically as many times as you want.

Steve

Steve Shaw
LIGHT ILLUSION

Light Illusion is offline  
post #305 of 385 Old 08-23-2019, 12:07 PM
AVS Forum Special Member
 
Light Illusion's Avatar
 
Join Date: Aug 2010
Posts: 1,806
Mentioned: 27 Post(s)
Tagged: 0 Thread(s)
Quoted: 764 Post(s)
Liked: 1123
Quote:
Originally Posted by Anger.miki View Post
To be more specific: at the moment peak chroma gives better results than hybrid on OLED.

While a LUT created on a QP Primary + Secondary (or even a 10^3 profile) with a 21^3 profile (measured with a Klein K10A) as “Hint” didn’t increase the accuracy of a LUT based on the same QP with no hint.

As soon as I get the time, I’ll test both features with my Kuro.
The LG WOLEDs are very unstable, so using a profile that is 'new' (as in recent) for a direct Peak Chroma LUT generation will likely generate the best results with any unstable display.

Using an older profile as a Hint will likely not work as the display will have changed a lot in the duration between the two profiles.
(Especially when the Hint profile is from a totally different LG display, as is the case here, as there is no guarantee two different WOLEDs will have anywhere near the same volumetric response.)

Steve

Steve Shaw
LIGHT ILLUSION

Light Illusion is offline  
post #306 of 385 Old 08-24-2019, 01:01 AM
Senior Member
 
ebr9999's Avatar
 
Join Date: Jun 2013
Posts: 387
Mentioned: 8 Post(s)
Tagged: 0 Thread(s)
Quoted: 226 Post(s)
Liked: 86
Quote:
Originally Posted by Light Illusion View Post
You can already do a 2-stage 1D LUT (or actually x-stage 1D LUT).

Do the first 1D LUT as standard.
Re-profile with the 1D LUT active in the TV.
Concatenate the first 1D LUT with the 2nd 1D LUT, and upload the resulting 1D LUT into the TV.
Repeat as needed...

Unlike 3D LUTs, where re-profiling with the LUT active to make a 2nd 3D LUT will not work, due to Gamut reduction, a 1D LUT has no such issues, so can be done basically as many times as you want.

Steve
Please, Steve a clarification:
When I generate 1DLUT I apply Sub Black. As for concatenating I think to use "Use Existing" (Color Space Conversion) I understand I should save (or re-convert) a copy of the previous 1DLUT without Sub Black applied, and then, at the end, apply Sub Black, as usual. Is that the correct procedure?

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 #307 of 385 Old 08-24-2019, 01:18 AM
AVS Forum Special Member
 
Light Illusion's Avatar
 
Join Date: Aug 2010
Posts: 1,806
Mentioned: 27 Post(s)
Tagged: 0 Thread(s)
Quoted: 764 Post(s)
Liked: 1123
I assume you actually mean you apply one of the 'VideoScale' options?

You can concatenate any LUTs, with/without VideoScale applied - so long as all the LUTs have the same VideoScale function applied.
Or, you can concatenate them before applying VideoScale, and then apply it to the final LUT.

I would probably use the second option personally.

Steve

Steve Shaw
LIGHT ILLUSION

Light Illusion is offline  
post #308 of 385 Old 08-25-2019, 04:26 AM - Thread Starter
AVS Forum Special Member
 
ConnecTEDDD's Avatar
 
Join Date: Sep 2010
Location: Athens, Greece
Posts: 8,149
Mentioned: 199 Post(s)
Tagged: 1 Thread(s)
Quoted: 3123 Post(s)
Liked: 3865
Quote:
Originally Posted by ebr9999 View Post
My conclusion: honestly I was expecting more from this 1DLUT Autocal. What concerns me the most is the color shift at high IRE. I have done some tests with Calman, based on 04.10.55 and I have got results in line with my manual calibration. It looks like the way my pannel reacts is not expected by LS software, and augmented data don't fix it. I am not a fan of Autocal, for its absenze of back-up, but combining its 1DLUTgeneration with LS 3DLUT generation looks, at least on my pannel, the most effective approach.

My next step will be 3DLUT generation, based on my manual calibration
Hi,

The issue you see, its not related with LightSpace.

There issues LG's have internally with 1D LUT processing, we had seen the same when the LG 2018 Template was ready, about May 2018, but the reason we (we = the calibration enthusiasts I mean) LG 2018 Template released at March 2019 it was because we were seeing a lot of more CalMAN's LUT issues by using LG's (exclusive from CalMAN only) LUT capabilities to be reported here in AVSForum.

Below you can find several posted CalMAN LUT engine problems using LG 2018 OLED TV's, a lot of these issues explained using LightSpace tools at many AVForum.com posts:

1 - 2 - 3 - 4 - 5 - 6 - 7 - 8 - 9 - 10 - 11 - 12 - 13 - 14 - 15 - 16 - 17 - 18 - 19 - 20 - 21 - 22 - 23 - 24 - 25 - 26 - 27 - 28 - 29 - 30 - 31 - 32 - 33 - 34 - 35 - 36 - 37 38 - 39 - 40 - 41 - 42 - 43 - 44 - 45 - 46 - 47 - 48 - 49 - 50 - 51.

The analysis of the issues from me or other LightSpace users for these CalMAN problems, helped CalMAN to resolve partially some issues, which was generally good for all CalMAN customers (those from Hollywood also where they never seen anything). Tyler (wifi-spy) has Thanked me and LightSpace users for that help.







No of the data above came from my personal measurements, all these are users data.

Since we saw that people have spend already a lot of their personal free time (which is limited for all), we released the LG Template to public for free, for LG TV users to be able to have other calibration options, to resolve their issues and be able to spend more of their free time on watching films than spending time with measurements.

There several posted reports @ AVSForum.com from users who used LG 2018 OLED TV's LUT capabilities with LightSpace, you can see below:

1 - 2 - 3 - 4 - 5 - 6 - 7 - 8 - 9 - 10 - 11 - 12 - 13 - 14 - 15 - 16 - 17 - 18 - 19.

The same happened for LG 2019 Template also, its was ready from May 2019 but we released it last week.

To give you an example, the before some days to the Warner Bros Shoutout, the LG C9 there had been calibrated using LG 2019 Template and LightSpace.

Another example is that available LG 2019 Template for PQ Curve Tool Upload, just we haven't released it yet, because I'm not ready to support the questions or haven't written the instructions, but it will be free for all users, since that tool is not affecting calibration, the user can use whatever software he is happy to use for his setup HDR10 calibration, since the PQ curve upload is just a parametric control of tone mapping of a menu which is not available to normal TV menu or service menu.



Light Illussion has contact LG many times to ask for integration with LG, but LG has never send any single reply, while the most important document related for AutoCAL and LG capabilities which I posted (world's first) here... talking about contacting with LG asking for API, so because we see that LG is not interesting, 'we' (the enthusiasts) released the LG Template to public. (since even CalMAN 3rd party file upload capability was broken for at least months).

I'm giving these details above to showcase the the group of enthousiast/friends we have, they have tested virtually everything, aiming to have the best possible picture, at this point we (the enthusiast team I mean) don't care what calibration solution we will use, the main priority is to have the best possible picture, all team members have both CalMAN and LightSpace software, but for the LG we use only LightSpace with DeviceControl, since that way provides the best end results with measurements and visually looking also.

The real problem for the issue you see now, is related with internal LG 1D LUT problem, its happening to most of the models from various different countries. The problem is that seems there many variations of electronics around the world or TV versions per region or manufactured dates.

The main problems are 3, and all related with internal 1D LUT.

1) When you will Enable calibration, using CalMAN or LG Template for DeviceControl for that, the TV acts differently. A CalMAN user has upload a video about this problem from a 2019 LCD model, but his finding have been totally ignored.
Quote:
Originally Posted by [email protected] View Post
I have a question about...
Below there 2 APNG files (Animated PNG), 2 pictures will swap with 1 sec interval to show the difference with Enabled and Disabled Calibration:

RGB Balance:



Gamma:



You can say....OK, since there available that difference, if we take measurements with Disabled the calibration it will be OK, Yes, but there also the 2 following problems.....

2) There difference with the 1D LUT table you upload internally to the TV from the one that is actually generated, due to some LG processing they are not matching....

To realize that, instead of loading the 1D LUT internally to the LG, load it as LightSpace's Active LUT and perform a Grayscale Large measurement run.

So LightSpace, will load the correction 1D LUT as an offset to your pattern generator output. This is the exact procedure used for years in professional market to locate similar issues.

Once you will have generated the 1D LUT correction table (...and you have that table opened) without applying any videoscale filter, go to LightSpace -> Calibration Interface -> Quick Profile, and select that LUT file as 'Active LUT':



... and LightSpace will load your LUT values and modify the patch generator colors based to your LUT from LightSpace generated values, this is a procedure they performing a lot in post-production to verify that kind of issues stuff....so you can load any external 1D/3D LUT file you want... which is ideal for LUT checking before uploading into a display or external 3D LUT box, to easier locate where is any issue... to the setup/hardware or level mismatch issue or internal processing etc.

With that way, you can verify (without having bought any external 3D LUT Box, without uploading anything internally to a LUT capable LG, or to test other TV/Projector you are thinking to buy a LUT Box how it will improve your color accuracy....before buying it.

This is not a new feature of LightSpace, its added before about 10 years.

Measuring the Grayscale as Active LUT and then mearing the same grayscale with the 1D LUT table loaded internally to the TV, you will find differencies.

These differences, we have found when initially tested internal LUT capabilities, this was one reason the 'Aug' feature added.

The issue its been explained there also:

https://www.lightillusion.com/error.html#signal_path

https://www.lightillusion.com/devicecontrol_manual.html

...and Steve Shaw has mentioned that problem again here:



3) When you reset the 1D LUT of TV from DeviceControl or CalMAN, there is a visible banding to the screen, at least to 4x 2018 OLED we have tested, to one was not visible.

To do that test, just display a grayscale ramp pattern before you will click to reset 1D LUT, when you will reset, you will see banding to the grayscale ramp, since with UNITY values it has banding, its impossible to use the internal 1D LUT table for calibration, at least for now.

So its better to not reset the 1D LUT table, do manual cal of 100% White using RGB-High and set Brightness to a value that is correct for the TV near black performance and then perform a 3D LUT, where inside to the 3D LUT it will contain 33-Point Grayscale also. Also when you take any measurement leave disabled the calibration from DeviceControl.

Ted's LightSpace CMS Calibration Disk Free Version for Free Calibration Software: LightSpace DPS / CalMAN ColorChecker / HCFR
S/W: LightSpace CMS, SpaceMan ICC, SpaceMatch DCM, CalMAN 5, CalMAN RGB, ChromaPure, ControlCAL
V/P: eeColor 3D LUT Box - P/G: DVDO AVLab TPG
Meters: JETI Specbos 1211, Klein K-10A, i1PRO2, i1PRO, SpectraCAL C6, i1D3, C5
ConnecTEDDD is online now  
post #309 of 385 Old 08-25-2019, 06:05 AM - Thread Starter
AVS Forum Special Member
 
ConnecTEDDD's Avatar
 
Join Date: Sep 2010
Location: Athens, Greece
Posts: 8,149
Mentioned: 199 Post(s)
Tagged: 1 Thread(s)
Quoted: 3123 Post(s)
Liked: 3865
I will post some post calibration results of an enthusiast which don't have AVSForum account, but its a member of the 'hardcore enthusiast team' and has tested virtually everything to get the best possible picture from his C8.

Because the verification is not typical like most of the users are using, he used 1000 points to verify the end results, so he performed a 10-Point Cube for verification to see how its C8 performs after uploading only a 3D LUT correction table using LightSpace (Peak Chroma) and LG Template of DeviceControl.
The pre-calibration procedure was an RGB-Gain adjustment of 100% White using White Balance of Normal OSD menu.

21-Point Cube measurements (with Drift Comp 50 value) used with Discus Colorimeter, but using a special patchset from another enthusiast, which based to the latest info about LG's color utilization of WRGB panel, he has wrote a utility where its calculating and sort the patches for better balance of both the RGBW utilization and the power over the course of the patch set.

Its attempting to balance both the pixel and sub-pixel power.

The end results is that by sorting the patches with that way, the panel has less drifting during the cube measurement, the lowest drift plot for WRGB OLED we have ever seen.

Lets star with how the 3D LUT correction looks from LightSpace's 3D LUT Preview:





Its smooth, without anomalies/noise/issues.

Now I will use ColourSpace to showcase the dE analysis since its easier to evaluation of such large verification run (1000 measured color points), dE2000 Avg: 0.3930 and Max: 1.3462

Lets see how was the panel at native gamut, as its measured from LightSpace, the following picture show 9261 measured colors, looking the CIE1931:



Its normal to have so large errors since its measurement of native panel.

The white line triangle is the REC.709.

The Green dots show errors 0-1 dE2000

The Yellow dots show errors 1-2.3 dE2000

The Red dots show errors higher from 2.3dE2000

The only Green errors as you can see is close to grayscale mainly only.

Now, its been added the tangent lines, gray lines where are showing the distance from where the color should be, with no error (0dE).

As CIE1931 is 2D chart, here is the 3D CIE1931, at top of the 'like x-mas tree' chart is the 100% White, and you see the 21-steps of measured luminance levels, as the Y dimension was missing from classic CIE1931 chart.



...now with added tangent lines:



How its time for the post calibration verification, of 1000 points:



991 points had dE 0-1dE and 9 only 1-1.5dE with maximum error for one point with 1.3462 dE

Now, its the CIE chart with filter to show only the errors above 1dE:



3D CIE:



3D CIE with Tangent lines:

bobof and mrtickleuk like this.

Ted's LightSpace CMS Calibration Disk Free Version for Free Calibration Software: LightSpace DPS / CalMAN ColorChecker / HCFR
S/W: LightSpace CMS, SpaceMan ICC, SpaceMatch DCM, CalMAN 5, CalMAN RGB, ChromaPure, ControlCAL
V/P: eeColor 3D LUT Box - P/G: DVDO AVLab TPG
Meters: JETI Specbos 1211, Klein K-10A, i1PRO2, i1PRO, SpectraCAL C6, i1D3, C5

Last edited by ConnecTEDDD; 08-25-2019 at 07:55 AM.
ConnecTEDDD is online now  
post #310 of 385 Old 08-25-2019, 06:29 AM - Thread Starter
AVS Forum Special Member
 
ConnecTEDDD's Avatar
 
Join Date: Sep 2010
Location: Athens, Greece
Posts: 8,149
Mentioned: 199 Post(s)
Tagged: 1 Thread(s)
Quoted: 3123 Post(s)
Liked: 3865
Now again using the same patchset and 21-Point Cube also, the only change was instead of using normal menu WB controls to pre-cal 100%, to use only the Service Menu RGB-Gain controls (without touching RGB balance of normal WB menu).

This provides less LG processing directly from the panel with result a 'cleaner' picture after the calibration with smoother gradations.


The reason for that is because the normal menu White Balance settings combined with White Balance of Service Menu settings and work as offset from the native panel, and when you have active two controls which are performing the same job, this can make the internal processing algorithm more complex, so its better idea one White Balance menu to be used only to adjust from native panel response and not from both menus (normal TV + service) the same time. Performing only Service Menu White Balance adjustments to pre-cal 100% White only, it will provide smoother gradations, as about other 10 users have reported to me also.

The dE are slightly improved, but visually looks better from previous calibration. This is the recommended method of pre-cal.

The 1000 measured color points had dE2000 Avg: 0.3390 and Max: 1.1453



993 points had dE 0-1dE and 7 only 1-1.5dE with maximum error for one point with 1.1453 dE



3D CIE with tangent lines:



Normalized RGB:


Ted's LightSpace CMS Calibration Disk Free Version for Free Calibration Software: LightSpace DPS / CalMAN ColorChecker / HCFR
S/W: LightSpace CMS, SpaceMan ICC, SpaceMatch DCM, CalMAN 5, CalMAN RGB, ChromaPure, ControlCAL
V/P: eeColor 3D LUT Box - P/G: DVDO AVLab TPG
Meters: JETI Specbos 1211, Klein K-10A, i1PRO2, i1PRO, SpectraCAL C6, i1D3, C5
ConnecTEDDD is online now  
post #311 of 385 Old 08-25-2019, 06:31 AM - Thread Starter
AVS Forum Special Member
 
ConnecTEDDD's Avatar
 
Join Date: Sep 2010
Location: Athens, Greece
Posts: 8,149
Mentioned: 199 Post(s)
Tagged: 1 Thread(s)
Quoted: 3123 Post(s)
Liked: 3865
Quote:
Originally Posted by ebr9999 View Post
Here outcome of my testing, done on SM, decreasing for each SM the default RGB by 20 points and verifying the result on Dark Room mode with Oled Light set to 25.
The outcome is ('=>' means 'affects')

SM Cold => Cold

SM Medium => Medium, Warm1

SM Warm => Warm1, Warm2, Warm3.
Thanks for the processional level test you have performed, I will update my guide with that valuable info.

Ted's LightSpace CMS Calibration Disk Free Version for Free Calibration Software: LightSpace DPS / CalMAN ColorChecker / HCFR
S/W: LightSpace CMS, SpaceMan ICC, SpaceMatch DCM, CalMAN 5, CalMAN RGB, ChromaPure, ControlCAL
V/P: eeColor 3D LUT Box - P/G: DVDO AVLab TPG
Meters: JETI Specbos 1211, Klein K-10A, i1PRO2, i1PRO, SpectraCAL C6, i1D3, C5
ConnecTEDDD is online now  
post #312 of 385 Old 08-25-2019, 06:36 AM - Thread Starter
AVS Forum Special Member
 
ConnecTEDDD's Avatar
 
Join Date: Sep 2010
Location: Athens, Greece
Posts: 8,149
Mentioned: 199 Post(s)
Tagged: 1 Thread(s)
Quoted: 3123 Post(s)
Liked: 3865
Quote:
Originally Posted by ebr9999 View Post
I have prepared everything for generating a 17 cube (4/5 hours with my I1D3). My panel (oled light 25) stays off till BR=52. Anyhow I have started with BR=51.
I have got good result, but poor near black details. I have had to rise BR to 52 and increase luminance at 05 IRE to 15.
So I decided: let's have another test with BR=52, but I have made a mistake: I have forgotten 05 Luminance at 15. The result was overall good, but with a near black disaster (you can easily undesrstand why).
So, I wondered: why spending another night for that? Let's try to use hint, using as base a Large grey scale, done on the correct 3DLUT settings. I was surprised of results: nice near black. I have checked for artefacts and banding, nothing to report.
When you are using the internal 1D LUT, the BR should be set to 50, there no choice to set it to other setting. The 1D LUT will try to fix your issues in near black, because each panel is different, to some users the problem is larger to some others is smaller. When 1D LUT will try to fix the near black and there no enough space for adjustments there...then you have to skip performing the 1D LUT and use BR setting to the value you will that resolve the problem.

But after the latest recommendation, we don't suggest from any user to use 1D LUT because of the LG's internal issues with processing of that table.

Ted's LightSpace CMS Calibration Disk Free Version for Free Calibration Software: LightSpace DPS / CalMAN ColorChecker / HCFR
S/W: LightSpace CMS, SpaceMan ICC, SpaceMatch DCM, CalMAN 5, CalMAN RGB, ChromaPure, ControlCAL
V/P: eeColor 3D LUT Box - P/G: DVDO AVLab TPG
Meters: JETI Specbos 1211, Klein K-10A, i1PRO2, i1PRO, SpectraCAL C6, i1D3, C5
ConnecTEDDD is online now  
post #313 of 385 Old 08-25-2019, 08:34 AM
Senior Member
 
ebr9999's Avatar
 
Join Date: Jun 2013
Posts: 387
Mentioned: 8 Post(s)
Tagged: 0 Thread(s)
Quoted: 226 Post(s)
Liked: 86
Quote:
Originally Posted by ConnecTEDDD View Post
Hi,

The issue you see, its not related with LightSpace.
...............................
Many thanks for you full of useful info answer (I am not quoting all of it on the sake of brevity).
Some comments:
  • As you have seen in my further post I have got good outcomes by not generating 1DLUT.
  • I don't blame LS. I had put 1DLUT generation out my testing, but following Steve indication about how achieving an n-pass approach by LUT concatenation, as being closer to the iteractive approach of CM. I will test that again and report here.
  • About reading with calibration enabled: I have noted that when you enable calibration, only on Cinema mode, gamut is set to wide and gamma to 2.2. Quite annoying for me as I have there a manual calibration I use for comparing the autocal ones. Note that when you enable enable calibration the Device Control template is set on Cinema and you graphs look related to Cinema. Maybe only somthing is out of our control in a specific mode. Not noted anything like that on the other modes and, as far as I can remember, on previous sw (I am 05.10.03).
  • I full agree LG should make available their API's. It is the base for a fair and fruitful competition.
ConnecTEDDD likes this.

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 #314 of 385 Old 08-27-2019, 09:29 AM
Senior Member
 
ebr9999's Avatar
 
Join Date: Jun 2013
Posts: 387
Mentioned: 8 Post(s)
Tagged: 0 Thread(s)
Quoted: 226 Post(s)
Liked: 86
^^^^^^^Here some follow up of my work:
  • About readings with Calibration Enabled: Every time I switch to a different mode, a see the TV saying "calibration enabled". It looks like that status is not general, but valid for each mode. I guess it has to be a little investigate by the Device Control Template author.
  • I have tried concatenating 1DLUT's. After 2 iterations I have stopped as I could not see improvments:

    The near Black is definitely good, but, as spotted by the graph, I can see color banding at high IRE. I am open to suggestions about how improving.
  • I have run a 3DLUT with luminance at 05 IRE set to 0 (you remind my mistake I have fixed with hint). Anyhow to get a good nearblack I have had to set L at 5 at 05 IRE. I have also tried to compare peak with hybrid. With Hybryd the near black is darker, but RGB balance a little better. I will not bother with these graphs.
    Anyhow the final result:

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 #315 of 385 Old 08-27-2019, 12:28 PM - Thread Starter
AVS Forum Special Member
 
ConnecTEDDD's Avatar
 
Join Date: Sep 2010
Location: Athens, Greece
Posts: 8,149
Mentioned: 199 Post(s)
Tagged: 1 Thread(s)
Quoted: 3123 Post(s)
Liked: 3865
Quote:
Originally Posted by ebr9999 View Post
I have also tried to compare peak with hybrid. With Hybryd the near black is darker, but RGB balance a little better. I will not bother with these graphs.
Hi, currently the best results we have seen its by skipping entirely the 1D LUT, pre-cal only 100% White from SM, set Brightness where its better for the display and the run 17-Point Cube with 16-255 patch generation (for Murideo set PC Level and @ LightSpace Options set patch scale 16-255) and not videoscale filter.

YCbCr 4:4:4 16-235 will work also with videoscale sub-black.

Peak Chroma provides better results currently.

Ted's LightSpace CMS Calibration Disk Free Version for Free Calibration Software: LightSpace DPS / CalMAN ColorChecker / HCFR
S/W: LightSpace CMS, SpaceMan ICC, SpaceMatch DCM, CalMAN 5, CalMAN RGB, ChromaPure, ControlCAL
V/P: eeColor 3D LUT Box - P/G: DVDO AVLab TPG
Meters: JETI Specbos 1211, Klein K-10A, i1PRO2, i1PRO, SpectraCAL C6, i1D3, C5
ConnecTEDDD is online now  
post #316 of 385 Old 09-06-2019, 01:02 PM
Member
 
sat4all's Avatar
 
Join Date: Apr 2013
Posts: 101
Mentioned: 4 Post(s)
Tagged: 0 Thread(s)
Quoted: 69 Post(s)
Liked: 28
Quote:
Originally Posted by ConnecTEDDD View Post
21-Point Cube measurements (with Drift Comp 50 value) used with Discus Colorimeter, but using a special patchset from another enthusiast, which based to the latest info about LG's color utilization of WRGB panel, he has wrote a utility where its calculating and sort the patches for better balance of both the RGBW utilization and the power over the course of the patch set.

Its attempting to balance both the pixel and sub-pixel power.

The end results is that by sorting the patches with that way, the panel has less drifting during the cube measurement, the lowest drift plot for WRGB OLED we have ever seen.
The guys behind this work didn't stop there, but they kept chasing perfection... the results are piece of art.



H/W:
- LG OLED65C8
- i1D3 rev.b profiled against i1Pro2 rev.e
- madVR HTPC for PG

Profiling Steps:
- Before Profiling you should warm up your TV for at least 1 hour with normal content.
- Set TV Gamma to 2.2/Brightness 50/Contrast 85
- Skip 1DLUT
- Reset only 3DLUT slot using DCC
- TV Service Menu White Balance gain adjustment using 192 as a fixed data point for one of the RGB channels while reducing the two others, while displaying 100% grey 10% pattern window with the desired OLED Light for your target nit.
- LS Patch Scale 0-255 / GPU 0-255 RGB / madTPG 16-235 (10% window pattern) / TV Black Laevel: Low

(In the final step we will apply videoscale sub-black filter on the 3DLUT)
(In my opinion even if you are tragetting data range it's better to calibrate using video range then after uploading the lut to the tv switch your tv to data range aka black level high, i've always achieved better results this way).




- LS Drift value was set to 50.
- WOLED optimized 10^3 patchset was used for pre-roll.
- Peak Chroma was used for LUT generation.

Here is my results with the new woled optimized patchset and it's only 4196p.
















@ConnecTEDDD
I've sent you my 1000p verfication file, would you please post ColourSpace analysis.


Cheers

LG OLED65C8 / Denon AVR-X3200W / KEF E305+ONKYO SKH-410 / Synology DS2415+ / Logitech Harmony 950
ZOTAC MAGNUS EN1060K: Win 10 x64 + Kodi DSPlayer x64

Last edited by sat4all; 09-06-2019 at 03:06 PM.
sat4all is online now  
post #317 of 385 Old 09-06-2019, 01:10 PM
AVS Forum Special Member
 
mrtickleuk's Avatar
 
Join Date: May 2016
Location: Birmingham, UK - you know, the original one!
Posts: 6,911
Mentioned: 179 Post(s)
Tagged: 0 Thread(s)
Quoted: 4741 Post(s)
Liked: 7041
Quote:
Originally Posted by sat4all View Post
The guys behind this work didn't stop there, but they kept chasing perfection... the results are piece of art.
Here is my results with their new woled optimized patchset and it's only 4196p.
P.S: Before profiling we skip 1dlut and do reset only 3dlut slot using free dcc soft, followed by SM WB.
Very impressive. "A piece of art" I agree!
sat4all likes this.

_______________
Denon AVR-X4200W, Arcam Alpha 8P; 5.1.4 setup: Mission 702e, M7C1i, 77DS, 731. Rel T5 Sub. Monitor Audio CT165 4 Tops | LG OLED55C8PLA TV | Samsung UBD-K8500 UHD Blu-Ray

HDMI 2.0 4K modes | Dolby & DTS core+outer audio tracks on (UHD) Blu-Rays | Hello to Jason Isaacs
mrtickleuk is online now  
post #318 of 385 Old 09-06-2019, 01:25 PM - Thread Starter
AVS Forum Special Member
 
ConnecTEDDD's Avatar
 
Join Date: Sep 2010
Location: Athens, Greece
Posts: 8,149
Mentioned: 199 Post(s)
Tagged: 1 Thread(s)
Quoted: 3123 Post(s)
Liked: 3865
Quote:
Originally Posted by mrtickleuk View Post
Very impressive. "A piece of art" I agree!
The verification was 1000 color points, 10-Point Cube, so is an advanced analysis of real performance with too many different areas of the gamut measured, for example ColorChecker SG which is 96 colors but not capturing the whole colorspace performance, there a lot of areas not measured:

This is ColorChecker SG in RGB space:



Here is 10-Point Cube:

Guz911 and sat4all like this.

Ted's LightSpace CMS Calibration Disk Free Version for Free Calibration Software: LightSpace DPS / CalMAN ColorChecker / HCFR
S/W: LightSpace CMS, SpaceMan ICC, SpaceMatch DCM, CalMAN 5, CalMAN RGB, ChromaPure, ControlCAL
V/P: eeColor 3D LUT Box - P/G: DVDO AVLab TPG
Meters: JETI Specbos 1211, Klein K-10A, i1PRO2, i1PRO, SpectraCAL C6, i1D3, C5
ConnecTEDDD is online now  
post #319 of 385 Old 09-06-2019, 02:08 PM - Thread Starter
AVS Forum Special Member
 
ConnecTEDDD's Avatar
 
Join Date: Sep 2010
Location: Athens, Greece
Posts: 8,149
Mentioned: 199 Post(s)
Tagged: 1 Thread(s)
Quoted: 3123 Post(s)
Liked: 3865
Quote:
Originally Posted by sat4all View Post
The guys behind this work didn't stop there, but they kept chasing perfection... the results are piece of art.

@ConnecTEDDD
I've sent you my 1000p verfication file, would you please post ColourSpace analysis.
The problem I see is that you have raised the perfection limit...so now we have to catch you with better results from these! lol

ColourSpace visuals now, 1000 color points evaluation.

APNG (Animated PNG file) of 2D CIE1931 Chart, the 3 frame loop will display your results filtered to display the dE2000 points with 0 - 0.5 dE2000 as frame 1, 0.5 - 1.0 dE2000 as frame 2 and 1 - 1.5 dE2000 as frame 3. The 2 points with highest errors was one with 1.06 and one with 1.07



ColourSpace display as green dots the 0-1 dE2000, while with orange dots the 1-2.3 dE2000.

APNG (Animated PNG file) of 3D CIE1931 Chart, the 3 frame loop will display your results filtered to display the dE2000 points with 0 - 0.5 dE2000 as frame 1, 0.5 - 1.0 dE2000 as frame 2 and 1 - 1.5 dE2000 as frame 3.



The larger dots are 100W and 100% Luminance of RGBCMY, in 3D Space.

APNG (Animated PNG file) of Normalized RGB Cube, the 3 frame loop will display your results filtered to display the dE2000 points with 0 - 0.5 dE2000 as frame 1, 0.5 - 1.0 dE2000 as frame 2 and 1 - 1.5 dE2000 as frame 3.


Ted's LightSpace CMS Calibration Disk Free Version for Free Calibration Software: LightSpace DPS / CalMAN ColorChecker / HCFR
S/W: LightSpace CMS, SpaceMan ICC, SpaceMatch DCM, CalMAN 5, CalMAN RGB, ChromaPure, ControlCAL
V/P: eeColor 3D LUT Box - P/G: DVDO AVLab TPG
Meters: JETI Specbos 1211, Klein K-10A, i1PRO2, i1PRO, SpectraCAL C6, i1D3, C5
ConnecTEDDD is online now  
post #320 of 385 Old 09-06-2019, 02:18 PM
Member
 
sat4all's Avatar
 
Join Date: Apr 2013
Posts: 101
Mentioned: 4 Post(s)
Tagged: 0 Thread(s)
Quoted: 69 Post(s)
Liked: 28
Quote:
Originally Posted by ConnecTEDDD View Post
The problem I see is that you have raised the perfection limit...so now we have to catch you with better results from these! lol




Early christmas tree


Thanks Ted
Guz911 likes this.

LG OLED65C8 / Denon AVR-X3200W / KEF E305+ONKYO SKH-410 / Synology DS2415+ / Logitech Harmony 950
ZOTAC MAGNUS EN1060K: Win 10 x64 + Kodi DSPlayer x64
sat4all is online now  
post #321 of 385 Old 09-06-2019, 02:46 PM - Thread Starter
AVS Forum Special Member
 
ConnecTEDDD's Avatar
 
Join Date: Sep 2010
Location: Athens, Greece
Posts: 8,149
Mentioned: 199 Post(s)
Tagged: 1 Thread(s)
Quoted: 3123 Post(s)
Liked: 3865
Quote:
Originally Posted by sat4all View Post
Early christmas tree
The early X-mas tree was your native display gamut performance, the Animated PNG below is filtering dE with 12 frames per 1 sec, from 0 to >5dE2000.

Green dots are 0-1dE2000
Orange dots are 1-2.3dE2000
Red dots are 2.3dE2000 and higher.

Guz911 likes this.

Ted's LightSpace CMS Calibration Disk Free Version for Free Calibration Software: LightSpace DPS / CalMAN ColorChecker / HCFR
S/W: LightSpace CMS, SpaceMan ICC, SpaceMatch DCM, CalMAN 5, CalMAN RGB, ChromaPure, ControlCAL
V/P: eeColor 3D LUT Box - P/G: DVDO AVLab TPG
Meters: JETI Specbos 1211, Klein K-10A, i1PRO2, i1PRO, SpectraCAL C6, i1D3, C5
ConnecTEDDD is online now  
post #322 of 385 Old 09-06-2019, 03:27 PM - Thread Starter
AVS Forum Special Member
 
ConnecTEDDD's Avatar
 
Join Date: Sep 2010
Location: Athens, Greece
Posts: 8,149
Mentioned: 199 Post(s)
Tagged: 1 Thread(s)
Quoted: 3123 Post(s)
Liked: 3865
Some LUT Previews using your 3D LUT correction as offset to my patterns using Animated PNG.

Frame 1 is the original pattern, frame 2 is your 3D LUT correction applied to the original pattern.....No issues.

Color Ramp:



Grayscale Ramp:


Grayscale Linearization:

Guz911 and sat4all like this.

Ted's LightSpace CMS Calibration Disk Free Version for Free Calibration Software: LightSpace DPS / CalMAN ColorChecker / HCFR
S/W: LightSpace CMS, SpaceMan ICC, SpaceMatch DCM, CalMAN 5, CalMAN RGB, ChromaPure, ControlCAL
V/P: eeColor 3D LUT Box - P/G: DVDO AVLab TPG
Meters: JETI Specbos 1211, Klein K-10A, i1PRO2, i1PRO, SpectraCAL C6, i1D3, C5
ConnecTEDDD is online now  
post #323 of 385 Old 09-12-2019, 12:01 AM - Thread Starter
AVS Forum Special Member
 
ConnecTEDDD's Avatar
 
Join Date: Sep 2010
Location: Athens, Greece
Posts: 8,149
Mentioned: 199 Post(s)
Tagged: 1 Thread(s)
Quoted: 3123 Post(s)
Liked: 3865
Quote:
Originally Posted by ConnecTEDDD View Post
Calibration Notes for LG 2018/2019 Profiling

Grade-1 SDR Reference Monitor/TV Calibration Targets

The viewing of accurate images depends on an accurate calibration of the display, control over the viewing environment (lightning and room decor), as well as the appropriate placement of the observer relative to the screen.

The Reference Viewing Environment can be considered one where color critical decisions are made, while the Home Viewing Environment is where finalized deliverables are viewed, with the intent to best match the original artistic intent, as defined by the director and colorist within the Reference Viewing Environment.



The goad of home TV calibration is the Home Viewing Environment to match the Reference Viewing Environment image.

Within the professional industries reference monitors (Grade-1) are the standard for color critical work.



EBU TECH 3320 (Version 4.1 - September 2019) - User Requirements for Video Monitors in Television Production, defines the technical characteristics for video broadcast monitors used in a professional TV production environment for evaluation and control of the images being produced.

Its describing the definition of Grade-1 SDR Reference Monitor with Standard Dynamic Range capabilities.

Grade-1 Monitors are devices for high-grade technical quality evaluation of images at key points in a color grading production workflow.

They are used for critical evaluation during post-production.

As a minimum requirement, these monitors shall have the quality properties of the image system they are used to evaluate.

It is expected that all applied technologies are state-of-the-art at this level as the Grade 1 monitor is a 'measuring instrument' for visual evaluation of image quality.

Grade-1 SDR Reference Monitor should be been calibrated and capable to produce a reference luminance level of 100 cd/m2 (nits) for 100% White (235 level @ 8-bit) patch on the screen.

Automatic Brightness Limiter (ABL) functions shall not be used for Grade-1 SDR Reference Monitors, this means that the monitor need to be capable to display 100 cd/m2 with a full field 100% Reference White pattern also.

100% luminance on the screen corresponds to a 10-bit luma signal of digital level 940, and the black level corresponds to a 10-bit luma signal of digital level 64.

100% luminance on the screen is defined as the luminance of a luma signal of digital level 940, but levels 941 through 1019 should also be correctly displayed.

The highest value of 10-bit luma signal is digital level 1019. The luma level 1019 is called 'Super-White' or '109% White'.

For the luminance gamma characteristic (Electro-Optical Transfer Function) of the screen, its recommended that a nominal value of 2.4 gamma to be used.
The EBU TECH 3320 (Version 4.1 - September 2019) I was talking to my 03 August 2019 post has been released to public @ 11 September 2019: https://tech.ebu.ch/publications/tech3320

Ted's LightSpace CMS Calibration Disk Free Version for Free Calibration Software: LightSpace DPS / CalMAN ColorChecker / HCFR
S/W: LightSpace CMS, SpaceMan ICC, SpaceMatch DCM, CalMAN 5, CalMAN RGB, ChromaPure, ControlCAL
V/P: eeColor 3D LUT Box - P/G: DVDO AVLab TPG
Meters: JETI Specbos 1211, Klein K-10A, i1PRO2, i1PRO, SpectraCAL C6, i1D3, C5
ConnecTEDDD is online now  
post #324 of 385 Old 09-12-2019, 08:01 AM
Advanced Member
 
Anger.miki's Avatar
 
Join Date: Jan 2018
Posts: 645
Mentioned: 18 Post(s)
Tagged: 0 Thread(s)
Quoted: 383 Post(s)
Liked: 410
I finally calibrated my new LG OLED55C8PLA. The results are stunning, let me accompany you in the description of the settings, of the hardware/software and methodology used and of the results:

Settings

- Picture mode: Expert (Bright)
- Gamma 2.2
- Color gamut: Wide
- OLED Light: 26
- Contrast: 85
- Brightness: 50
- Color: 50
- Sharpness 0
- All image elaboration features OFF
- Energy saving and burn-in prevention functions OFF


Hardware/Software and methodology

- Spectro: x-rite i1 Pro 2
- Colorimeter: basICColor Discus (mode 0, 0.35 seconds extra delay)
- TPG: DVDO AvLab TPG
- Laptop: Dell Inspiron 17-7000 series
- Software: Light Illusion LightSpace CMS HTP 10.0.0.2983, Device Control - LG 2018 Template.

1) did reset Expert (Bright) 3DLUT with Device Control
2) adjusted 2 points high white balance with RGB 16-235 12 bit 100% white with spectro into TV SERVICE MENU
3) created FCCM with RGB 16-255 12 bit patches
4) ran a LightSpace display characterization with special OLED patch set: 4566 points measured with RGB 16-255 (0.35 seconds black patch insertion). It took 2 hours and 28 minutes.
5) converted color space: source rec709 gamma 2.2
6) exported 3DLUT in dcl format
7) uploaded dcl file through Device Control
8) ran a 1000 points verification RGB 16-235 with an average dE 2000 of 0.2622(!) and max of 0.9573(!!)

A special mention goes to the patchset. I've bought it from a guy who did one specially for OLED panels. As you will see, the results are outstanding! For further details about that PM me.


Results

Profile:

1DLUT graph




The Cube









Verification:

1DLUT graph




The Cube





Charts/Graphs/Summary















Not bad uh!?! Panasonic and Sony should really think deeply about adding 3DLUT capabilities to their 2020 sets.

Unfortunately, I don't have the new ColorSpace but if @ConnecTEDDD could be so kind to do his magic with CS, I'll be glad. I leave my verification bcs attached. Thanks.
Attached Files
File Type: zip C8 FE OPT RGB16-235 verif.zip (39.5 KB, 7 views)

TVs: Pioneer PDP-LX5090H, LG OLED55C8PLA | SintoAmp: Pioneer VSX-921 | BD Player: Panasonic DMP-BDT260EG | External LUT box: Entertainment Experience eeColor | Softwares: Light Illusion Lightspace HTP, Portrait Displays CalMAN Home Enthusiast 2018 R3, HCFR, DisplayCAL | Probes: x-rite i1 Pro 2 - i1 Display Pro OEM B-02, basICColor DISCUS | Test Pattern Generator: DVDO AVLab TPG

Last edited by Anger.miki; 09-16-2019 at 02:43 AM.
Anger.miki is offline  
post #325 of 385 Old 09-12-2019, 10:02 AM
Senior Member
 
Vishwa Somayaji's Avatar
 
Join Date: Jun 2001
Location: Edmonton,Alberta Canada
Posts: 490
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 64 Post(s)
Liked: 71
Wow! those cubes look perfect. Now when you watch the TV you know that you are looking at the picture as the director intended. Thanks for these pictures. It shows it is possible to achieve near perfect results.
Anger.miki likes this.
Vishwa Somayaji is online now  
post #326 of 385 Old 09-12-2019, 12:45 PM
AVS Forum Special Member
 
sillysally's Avatar
 
Join Date: Jun 2006
Location: Chicago
Posts: 5,152
Mentioned: 20 Post(s)
Tagged: 0 Thread(s)
Quoted: 1178 Post(s)
Liked: 1052
Your Gamut coverage is low. What filters in lightspace did you use?
If you didn't use filters or max luminance to convert, then you may want to check your adjustments in the service menu.

ss

"Don't worry be happy"
LG77C8, Klein K10-A, Jeti 1211, CMS Lightspace, Calman home plus licence for K10-A, Murideo 6G.
sillysally is offline  
post #327 of 385 Old 09-12-2019, 01:38 PM
AVS Forum Special Member
 
mrtickleuk's Avatar
 
Join Date: May 2016
Location: Birmingham, UK - you know, the original one!
Posts: 6,911
Mentioned: 179 Post(s)
Tagged: 0 Thread(s)
Quoted: 4741 Post(s)
Liked: 7041
Please could someone explain in layman's terms why some of the cubes show a small "wedge" of dots, which only takes up half the cube, but the others show a full cube?

Is the former ("display characterization") representing what the panel was measured as, with the latter ("verification") showing the results of applying the LUT? If so, how can the display suddenly be capable of emitting the colours at the edges of the cube, when it couldn't before?

Perhaps they are at different scales, and the verification cube is filled up with the gamut that has been selected for the TV? So what's the scale of the cube in the first stage - the full rec.2020?

_______________
Denon AVR-X4200W, Arcam Alpha 8P; 5.1.4 setup: Mission 702e, M7C1i, 77DS, 731. Rel T5 Sub. Monitor Audio CT165 4 Tops | LG OLED55C8PLA TV | Samsung UBD-K8500 UHD Blu-Ray

HDMI 2.0 4K modes | Dolby & DTS core+outer audio tracks on (UHD) Blu-Rays | Hello to Jason Isaacs
mrtickleuk is online now  
post #328 of 385 Old 09-12-2019, 02:00 PM
AVS Forum Special Member
 
Rolls-Royce's Avatar
 
Join Date: Apr 2003
Location: Victorville, CA
Posts: 4,132
Mentioned: 38 Post(s)
Tagged: 0 Thread(s)
Quoted: 1131 Post(s)
Liked: 1235
Quote:
Originally Posted by mrtickleuk View Post
Please could someone explain in layman's terms why some of the cubes show a small "wedge" of dots, which only takes up half the cube, but the others show a full cube?

Is the former ("display characterization") representing what the panel was measured as, with the latter ("verification") showing the results of applying the LUT? If so, how can the display suddenly be capable of emitting the colours at the edges of the cube, when it couldn't before?

Perhaps they are at different scales, and the verification cube is filled up with the gamut that has been selected for the TV? So what's the scale of the cube in the first stage - the full rec.2020?
They are indeed at different "scales". The first should be the display's native gamut, closer to DCI P-3, while the second appears to be for Rec.709. The fact that Miki used his AVLab TPG (SDR-only) with no other device such as an HDFury in the chain tends to bear this out.

...Royce...

"I never drink...wine."
Bela Lugosi, DRACULA, 1931
Rolls-Royce is online now  
post #329 of 385 Old 09-12-2019, 02:22 PM - Thread Starter
AVS Forum Special Member
 
ConnecTEDDD's Avatar
 
Join Date: Sep 2010
Location: Athens, Greece
Posts: 8,149
Mentioned: 199 Post(s)
Tagged: 1 Thread(s)
Quoted: 3123 Post(s)
Liked: 3865
Quote:
Originally Posted by sillysally View Post
Your Gamut coverage is low. What filters in lightspace did you use?
If you didn't use filters or max luminance to convert, then you may want to check your adjustments in the service menu.
Hi SS,

The percentage is the volumetric value of gamut coverage... and its correct, as its a verification with 1000 point profile, 93-94% is expected.

LightSpace uses normalized volumetric values for gamut coverage for SDR (for PQ are absolute).

When you try to check the coverage, the value will be lower, as it is already ‘on the edge’ of the target colorspace, so many 'edge' points will be 'slightly' under gamut.

Only 70 of the 1000 points need to be slightly under gamut for a value of 93% and for a volumetric profile; that can be easily happen for a verification.

When you check the native gamut profile, the figure will always be larger, as the figure always shows the coverage 'After' calibration.

So applying to a verification profile, which is already 'after' calibration it will always produce a lower value, as it assumes another calibration process is to be applied.

Ted's LightSpace CMS Calibration Disk Free Version for Free Calibration Software: LightSpace DPS / CalMAN ColorChecker / HCFR
S/W: LightSpace CMS, SpaceMan ICC, SpaceMatch DCM, CalMAN 5, CalMAN RGB, ChromaPure, ControlCAL
V/P: eeColor 3D LUT Box - P/G: DVDO AVLab TPG
Meters: JETI Specbos 1211, Klein K-10A, i1PRO2, i1PRO, SpectraCAL C6, i1D3, C5
ConnecTEDDD is online now  
post #330 of 385 Old 09-12-2019, 03:07 PM - Thread Starter
AVS Forum Special Member
 
ConnecTEDDD's Avatar
 
Join Date: Sep 2010
Location: Athens, Greece
Posts: 8,149
Mentioned: 199 Post(s)
Tagged: 1 Thread(s)
Quoted: 3123 Post(s)
Liked: 3865
Quote:
Originally Posted by mrtickleuk View Post
Please could someone explain in layman's terms why some of the cubes show a small "wedge" of dots, which only takes up half the cube, but the others show a full cube?

Is the former ("display characterization") representing what the panel was measured as, with the latter ("verification") showing the results of applying the LUT? If so, how can the display suddenly be capable of emitting the colours at the edges of the cube, when it couldn't before?

Perhaps they are at different scales, and the verification cube is filled up with the gamut that has been selected for the TV? So what's the scale of the cube in the first stage - the full rec.2020?
There is explanation of what to look by looking a Cube Viewer of LightSpace here: https://www.avsforum.com/forum/139-d...l#post56183576

It has example pictures of how you can see issues by only looking a cube preview (without taking any measurement with a meter) so to understand and evaluate the 'quality' of the correction table. If there issues there, then you will able to notive problem when you will evaluate patterns also, just LightSpace has tools for such evaluation so you save time even from displaying patterns to your specific TV, you can load reference pictures to evaluate first as quick test, and later you can confirm by loading the same patterns to the real display. That tool called LUT Preview, I have posted examples of loading my patterns to LightSpace LUT preview, some posts earlier evaluating the results of the user 'sat4all'.

Below you can find several posted CalMAN 3D LUT engine problems using LG 2018 OLED TV's, a lot of these issues explained using LightSpace tools at many AVSForum.com posts (none of these are my measurements results, some are my deep technical analysis only...from my home):

1 - 2 - 3 - 4 - 5 - 6 - 7 - 8 - 9 - 10 - 11 - 12 - 13 - 14 - 15 - 16 - 17 - 18 - 19 - 20 - 21 - 22 - 23 - 24 - 25 - 26 - 27 - 28 - 29 - 30 - 31 - 32 - 33 - 34 - 35 - 36 - 37 - 38 - 39 - 40 - 41 - 42 - 43 - 44 - 45 - 46 - 47 - 48 - 49 - 50 - 51.

LightSpace tools used, as for locating the issues... just taking post cal measurement its the 50% of the success of a calibration, the end preserved result matters more.

If the dE are low but the picture has issues, then you have to ignore the LUT..... so this is exact what the users of the above links did. (the users saying that, not me....just users reported the problems, and I explained to them the problems, suggesting them where to look and what patterns to display to their setup to realize the problems with real content.

Calibration is not only dE...
mrtickleuk likes this.

Ted's LightSpace CMS Calibration Disk Free Version for Free Calibration Software: LightSpace DPS / CalMAN ColorChecker / HCFR
S/W: LightSpace CMS, SpaceMan ICC, SpaceMatch DCM, CalMAN 5, CalMAN RGB, ChromaPure, ControlCAL
V/P: eeColor 3D LUT Box - P/G: DVDO AVLab TPG
Meters: JETI Specbos 1211, Klein K-10A, i1PRO2, i1PRO, SpectraCAL C6, i1D3, C5
ConnecTEDDD 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