LG OLED's 3D LUT Profiling using LightSpace Thread - Page 25 - AVS Forum | Home Theater Discussions And Reviews
Forum Jump: 
 550Likes
Reply
 
Thread Tools
post #721 of 779 Old 01-08-2020, 10:02 AM
Member
 
fonsocm's Avatar
 
Join Date: Feb 2008
Posts: 44
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 18 Post(s)
Liked: 10
Quote:
Originally Posted by liberator72 View Post
Okay, so it was fast because you set Integration at 0.25. For an accurate dark room calibration I personally use 0.75 Integration as it is my understanding that will give a little more accuracy. I have used 0.25 for my day mode based on recommendations of Anger.Miki and IMO it is perfectly fine for my day mode. I haven’t tried it for my night mode calibrations yet as I am happy with the time it takes for 17^3 patch set using 0.75.

If you run another 1000 point profile, using 16-235 as your patch scale, you will get a better idea of the overall accuracy.

When you load a 3D LUT to your Pic Mode it will apply to all inputs using that same Pic Mode. You will have to manually change some settings within the Pic Mode though. If you calibrated COOL in the SM WB and used that for your calibration profile, you’ll have to apply the COOL WB to each input in the User Menu. Other post processing settings also need to be manually applied per input (e.g. TruMotion settings).
I made the quick profile, not 1000 point because I didn't have more time.

My main concern is the location of the colorimeter, using not contact make the things more difficult for me.
fonsocm is offline  
Sponsored Links
Advertisement
 
post #722 of 779 Old 01-09-2020, 03:41 AM
Member
 
fonsocm's Avatar
 
Join Date: Feb 2008
Posts: 44
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 18 Post(s)
Liked: 10
Hi.

Like I said my main concern is the location of the spectro and the colorimeter. I followed the recommendations to place in different distances to measure the same zone (they have different FOV) and the pairing be better for each channel. It's clear and logic.

However, I think there is a difference in the luminosity measure because the different locations.

In order to get a better pairing in luminosity and channel, I am thinking to follow using the method above to measure the channels and place to same distance the spectro and colorimeter to measure the luminosity.

Does it make sense or have I lost the way?

Thanks in advance.
fonsocm is offline  
post #723 of 779 Old 01-09-2020, 06:55 AM
Senior Member
 
liberator72's Avatar
 
Join Date: Mar 2018
Location: Sunny (when it's not raining) Devon, U.K.
Posts: 235
Mentioned: 11 Post(s)
Tagged: 0 Thread(s)
Quoted: 173 Post(s)
Liked: 263
Quote:
Originally Posted by fonsocm View Post
Hi.

Like I said my main concern is the location of the spectro and the colorimeter. I followed the recommendations to place in different distances to measure the same zone (they have different FOV) and the pairing be better for each channel. It's clear and logic.

However, I think there is a difference in the luminosity measure because the different locations.

In order to get a better pairing in luminosity and channel, I am thinking to follow using the method above to measure the channels and place to same distance the spectro and colorimeter to measure the luminosity.

Does it make sense or have I lost the way?

Thanks in advance.
Just to confirm because I may have missed it, but you are using i1Display Pro matching to an i1Pro2, correct?

If this is the case, it is normal for the two probes to measure different on luminance, but after probe matching using FCMM, they should read a lot closer to each other and ideally be pretty much perfectly matched reading those same patches with the correction applied. The exception is Red Luminance as the i1Display Pro has some issues reading full Red on these OLED's.

If after matching you're not getting very similar results between the two probes at least measuring RGBW, then I suspect either your meter placement was wrong, or there may be some external source contaminating the readings? If the later is the case then find some way to cover over the area while measurements are taken (drape some kind of cloth, towel anything really), or maybe just use the probes in contact mode and see what your matched results are then
fonsocm likes this.

LG OLED65C9MLB**Denon AVR-X3500H**Focal Sib Evo Atmos 5.1.2**Sky Q UHD**Xbox One X**Panasonic DP-UB820**ATV 4K**Harmony Elite
i1D3 OEM**i1D3 Retail**i1 Pro2 OEM**LightSpace HTP**CalMAN Home for LG**RPi Pgenerator**Ted's Disk & Media Files**S & M HD & UHD HDR Benchmark Disks**HD Fury Integral 2
liberator72 is online now  
Sponsored Links
Advertisement
 
post #724 of 779 Old 01-09-2020, 10:13 AM
Advanced Member
 
Anger.miki's Avatar
 
Join Date: Jan 2018
Posts: 754
Mentioned: 36 Post(s)
Tagged: 0 Thread(s)
Quoted: 470 Post(s)
Liked: 528
FOV matching is something that could be skipped if there are other difficulties. The most important things to check with WOLED are that the probes are perpendicular to the panel and that both reads the same patch after the same(ish) amount of time. That's why I suggest to beginners to place probes (one at a time) in contact when profiling. That's an easy and quick way to know that your probes are placed perpendicularly to the screen. OLEDs do not emit a heat that will compromise the readings. Anyway, unfortunately even if you are able to achieve a perfect FCCM, that doesn't mean that you'll have an accurate profile/LUT due to the non-additive nature of those panels. Multi-point volumetric probe matching is surely better, but there are some things/methodologies to be known even there...

T.U.C. Master | TVs: Pioneer PDP-LX5090H, LG OLED55C8PLA | AV Receiver: Pioneer VSX-921 | BD Player: Panasonic DMP-BDT260EG | External LUT box: Entertainment Experience eeColor | Softwares: Light Illusion Lightspace HTP, CM Enthusiast, HCFR, DisplayCAL | Probes: Klein K10-A, Jeti spectraval 1501, X-Rite i1 Pro 2 OEM, X-Rite i1 Display Pro OEM Rev. B-02, basICColor DISCUS | Test Pattern Generator: DVDO AVLab TPG
Anger.miki is online now  
post #725 of 779 Old 01-09-2020, 10:37 AM
Member
 
grizzledyoungman's Avatar
 
Join Date: Dec 2011
Posts: 102
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 72 Post(s)
Liked: 19
Considering LS after Calman Home for LG C9

Hello - after a few different calibrations with Calman Home for LG I've had very mixed feelings about the results. Here is the setup I used:

Calman Home for LG
LG iTPG with DDC
i1D3 retail rev B (from Nov 2018)
FSI WOLED EDR

Overall it's good in most situations, but posterization in the shadow areas in SDR modes is consistently noticeable. Undesirable, given how much content we watch these days is highly compressed SDR video.

As I understand, this issue is prevalent with Calman because of how the software calculates the 3D LUT, and that Lightspace is a significant improvement in that regard. As such, I'm curious to explore a supplemental calibration with Lightspace. My goal is a DIY calibration on my LG C9 that improves upon what Calman Home can do, ideally while leveraging LG's DDC to whatever degree possible.

Any advice on making the jump? Reading through the forums here and documentation elsewhere is useful, but the array of information and options is dizzying. As such, I'm hoping someone can point me in the right direction and help answer the following:
  1. What version of Lightspace is appropriate given the setup/goals described above?
  2. What EDR/metering mode is best for my meter? If I want to make a custom profile for my meter, what photospectrometers are supported?
  3. What TPG offers the best bang/buck for SDR and HDR calibration and verification?
  4. Is there a concise description of the SDR/HDR calibration workflows for that setup?
  5. Is enabling DDC really this complicated? https://displaycalibrations.com/lg_2...device_control Because that feels like a solid 4-5 hour procedure.
Thank you!
mrtickleuk likes this.
grizzledyoungman is offline  
post #726 of 779 Old 01-09-2020, 11:13 AM
Senior Member
 
liberator72's Avatar
 
Join Date: Mar 2018
Location: Sunny (when it's not raining) Devon, U.K.
Posts: 235
Mentioned: 11 Post(s)
Tagged: 0 Thread(s)
Quoted: 173 Post(s)
Liked: 263
Quote:
Originally Posted by grizzledyoungman View Post
Hello - after a few different calibrations with Calman Home for LG I've had very mixed feelings about the results.
Quote:
Originally Posted by grizzledyoungman View Post
Hello - after a few different calibrations with Calman Home for LG I've had very mixed feelings about the results.
Quote:
Originally Posted by grizzledyoungman View Post
Hello - after a few different calibrations with Calman Home for LG I've had very mixed feelings about the results.
WOW, you must be very eager to give this a try with 3 posts in just 2 minutes


I did my first ever 3D LUT calibration just a little over 2 months ago. While looking at the various guides/instructions/suggestions et al can seem a little daunting at first, once you actually get your hands on the software and all other relevant accompanying bit and bobs, it really couldn't be any more straight forward. So as a relative newcomer to LightSpace myself, I will answer your direct questions as simply as I can from that perspective. Other more experienced users will fill in the blanks for the more advanced side of things, but as far as getting set up to go....

1. To generate LUT's you will need at least Light Space HTL

2. All versions of LightSpace have a WRGB OLED EDR. LS HTL supports consumer meters (e.g. X-Rite Probes), so you could use an i1Pro/i1Pro2 to profile an i1 Display Pro. Higher end and Pro meters require a higher license level of LS HTP

3. The cheapest and most cost effective TPG is the Raspberry Pi based PGenerator. You can add a HD Fury device to inject HDR metadata should you wish.

4. There are guides to doing the SDR calibration (3D LUT creation/upload) and there is also lots of information here in this thread. HDR is a whole different ball game, you'd want to concentrate on SDR before going down the HDR road anyway.

5.Whilst setting up Device Control for connecting to the TV for DDC commands may look complicated, it really isn't. It takes very little time, and if you follow the steps laid out you can be all set up for DDC in 15 - 20 minutes. Whilst it isn't the "plug and play out of the box" approach, once set up it is just as plug and play as any other system.

At the end of the day, if you are not happy with what you have now, it's up to you whether you want to go to what amounts to a little extra effort to get what is basically the best available option out there. Add to the fact the advice and support you get from everyone surrounding LightSpace is exemplary, if you find yourself getting a bit stuck, there will always be someone to chirp up with help and advice.

It should also be noted that if you were to decide to purchase LightSpace now, (AFAIK) you would be eligible for the free upgrade to ColourSpace when it releases some time early this year.
L30Z3N likes this.
liberator72 is online now  
post #727 of 779 Old 01-09-2020, 11:21 AM
Advanced Member
 
Make73's Avatar
 
Join Date: Jan 2010
Location: Finland
Posts: 509
Mentioned: 9 Post(s)
Tagged: 0 Thread(s)
Quoted: 225 Post(s)
Liked: 155
Quote:
Originally Posted by grizzledyoungman View Post
As I understand, this issue is prevalent with Calman because of how the software calculates the 3D LUT, and that Lightspace is a significant improvement in that regard. As such, I'm curious to explore a supplemental calibration with Lightspace. My goal is a DIY calibration on my LG C9 that improves upon what Calman Home can do, ideally while leveraging LG's DDC to whatever degree possible.
https://displaycalibrations.com/lightspace_order.html
I suggest you check that page and contact owner, Ted Aspiotis.
He have some ready writen basic instructions for newbies like me (actually he is working to publish one to web but not yet ready) and was very helpful also for other clients.

btw, you posted same three times, maby you could remove two additionals
dwaleke likes this.
Make73 is offline  
post #728 of 779 Old 01-09-2020, 11:35 AM
Member
 
grizzledyoungman's Avatar
 
Join Date: Dec 2011
Posts: 102
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 72 Post(s)
Liked: 19
Quote:
Originally Posted by Make73 View Post
https://displaycalibrations.com/lightspace_order.html
I suggest you check that page and contact owner, Ted Aspiotis.
He have some ready writen basic instructions for newbies like me (actually he is working to publish one to web but not yet ready) and was very helpful also for other clients.

btw, you posted same three times, maby you could remove two additionals

Will do! Removed the triple post - sorry. Browser issue!
grizzledyoungman is offline  
post #729 of 779 Old 01-13-2020, 09:03 AM
Member
 
Join Date: Dec 2019
Location: Finland
Posts: 22
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 13 Post(s)
Liked: 8
After enough (stupid) questions, managed to find enough time for a second calibration.
As the white balance takes less time than display profiling, I decided to run it twice. Now I have cool for HRD and DV measured with brightness at 50 and medium with 52 for SDR.


Included the verification .bcs file if someone is interested to see it.
Attached Files
File Type: zip C8 2nd verification.zip (39.5 KB, 10 views)
Miksu2 is offline  
post #730 of 779 Old 01-13-2020, 11:44 AM
Senior Member
 
ebr9999's Avatar
 
Join Date: Jun 2013
Posts: 412
Mentioned: 15 Post(s)
Tagged: 0 Thread(s)
Quoted: 239 Post(s)
Liked: 123
Use of Volumetric Colorimeter Correction using Jeti/I1d3 – My conclusions

I have tested the volumetric probe correction in my environment: LG Oled E8, I1D3, Jeti 1501, Murideo PG. Comparisons have been run between various volumetric corrections and FCMM.
The approach I have used was:
  1. Measure for corrections based on the native (wide) gamut of my E8
  2. Validate the correction in the same scenario
  3. Test the corrections on a Rec.709 Gamut previously generated with LS.
For all verifications I have measured the distance from Jeti to the corrected I1D3, both in math terms (i.e. %dY - normalised on 255,255,255 for easier comparison, dx, dy, dxy) and perceptive terms (dE2000). 1000 (10^3) patches have used.


My conclusions:
  1. You start having significant improvements with volumetric is based on 4^3 profiles. Volumetric based on Quick Profiles has not given me good results. I see FCMM limitation due the non additive nature of WRGB OLED display has been solved.
  2. I have tried measuring for volumetric having reference and colorimeter reading at the same time (i.e. using DIP mode). This method has not given me significant improvements, so I have switched to the usual readings, keeping care to do preroll and warming up.
  3. Using the CMF basic I1D3 profile works badly for Reds. It works using the WRGB Oled EDR profile.
More detail about my conclusions


About item 1:
Here below a snapshot comparing Primary Only QP (with a trick I explain later), 3^, 4^, 5^ and FCMM;


Starting from 4^ the correction is interesting

Here also a snapshot of primary and secondary colour down to level 193:

The quality of 4^ is impressive!

The limitation in correction appear down to around 1.5 Nits. This is a difficult area also for FCMM, e and clearly, 11D3 limits take their role. To have an idea about you have to open the attached excel and have a look.


Regarding the use of quick profile, Steve (@Light Illusion) has clarified me that they must be valid Quick Profile and must be read from the Calibration Interface. So, as I wanted to remove RGB level 13 from Primary & Secondary QP (the validation of Primary only has returned poor results), I have had to run it from Display Characterisation. Also important, for anything it is not a quick profile, you have to start from DC. If interested I have files I can share with you.


About item 2:
Here below testing outcomes (over 987 most luminance patches) generation comparing readings done at the same and at different time:

There is a little advantage for data taken at different times. If you to have a deeper look go into the attached file.


About Item 3:
The herein snapshot, shows how, using the I1D3 CMF profile reflects in a significant Red chromaticity errors, whilst that is not happening using EDR correction and FCMM.

This has been also been confirmed by @liberator72 testing on the same subject.

Achievements: @Anger.miki (dE2000 genius), @bobof (his patch generator and his advices) and @liberator72 for sharing his testing results
dwaleke, bobof, L30Z3N and 3 others 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

Last edited by ebr9999; 01-22-2020 at 07:13 AM.
ebr9999 is offline  
post #731 of 779 Old 01-13-2020, 01:41 PM
Senior Member
 
liberator72's Avatar
 
Join Date: Mar 2018
Location: Sunny (when it's not raining) Devon, U.K.
Posts: 235
Mentioned: 11 Post(s)
Tagged: 0 Thread(s)
Quoted: 173 Post(s)
Liked: 263
Quote:
Originally Posted by ebr9999 View Post
I have tested the volumetric probe correction in my environment: LG Oled E8, I1D3, Jeti 1501, Murideo PG. Comparisons have been run between various volumetric corrections and FCMM.
The approach I have used was:
  1. Measure for corrections based on the native (wide) gamut of my E8
  2. Validate the correction in the same scenario
  3. Test the corrections on a Rec.709 Gamut previously generated with LS.
For all verifications I have measured the distance from Jeti to the corrected I1D3, both in math terms (i.e. %dY - normalised on 255,255,255 for easier comparison, dx, dy, dxy) and perceptive terms (dE2000). 1000 (10^3) patches have used.


My conclusions:
  1. You start having significant improvements with volumetric is based on 4^3 profiles. Volumetric based on Quick Profiles has not given me good results. I see FCMM limitation due the non additive nature of WRGB OLED display has been solved.
  2. I have tried measuring for volumetric having reference and colorimeter reading at the same time (i.e. using DIP mode). This method has not given me significant improvements, so I have switched to the usual readings, keeping care to do preroll and warming up.
  3. Using the CMF basic I1D3 profile works badly for Reds. It works using the WRGB Oled EDR profile.

What an excellent write up on the conclusions from all the recent testing @ebr9999

If I may add, although you and I both are using an i1d3 colorimeter, your Volumetric Corrections are being made with a Jeti Sprectro, whereas I am profiling in combination with an i1Pro2.

While it has been previously suggested that little benefit may be seen from using an i1Pro2 for Volumetric Correction profiles due to its limitations in low light measurements, some real benefit can be seen to achieved provided extra care is taken during setup.

As stated, with the WRGB OLED EDR correction in place when using the i1d3, corrections are much better than using Generic CMF and my own results are very good until below around 2 nits in the Blue region where luminance errors begin to get higher the lower the luminance gets. I guess this is to be expected due to the limitations of the i1Pro2 to measure low luminance patches, and the amount of time it takes to read them.

Using a relatively small patchset of just 47 points returns improvements that can be seen below. The problem with using the i1Pro2 for running larger validations once again comes down to low luminance limitations and the speed of measurements. A 968 point validation takes almost an hour, it's just 20 minutes with an i1d3. Due to this, the further you get into the longer validations, the more the panel drifts, and the larger the differences in measurements between the probes become. Looking to the screenshots below confirms this. The 47 point volumetric correction validation is very good match. As the validation profiles get larger, the time difference between the different probes reading becomes longer as the measurements are being taken consecutively and not concurrently, and errors will begin to creep in and fluctuate. Take a look to the random snip from the middle of a 180 point validation where the luminance errors begin to rise because of this. But considering the time between readings, it is somewhat understandable. When we get to the lengthy 968 point validation, it begins well considering an hour has passed since the i1Pro2 started its readings, but as it progresses through (the screenshot taken approximately 2/3rd way through the patchset), the differences in measurements being to rise somewhat. Again this, I can only assume, is due to panel drift between the length of time between readings of the two slower consumer meters.

But when compared to a freshly created FCMM correction the next day (to give the panel a short break) followed by a new i1Pro2 968p validation run and an i1d3 968p validation run, the difference in results is quite staggering!

There is some methodology in the requirements for setting up for measurements to create the Volumetric Corrections, and these, I'm sure, can and will be explained in more detail at a later date. But suffice to say, Volumetric Corrections even on low end consumer probes, gives substantially better results on OLED panels than the regular FCMM Corrections.

Massive thanks go out to all involved in the testing of this. A huge amount of time invested by everyone!
Attached Thumbnails
Click image for larger version

Name:	47p Volumetric Correction Validation.PNG
Views:	34
Size:	141.2 KB
ID:	2669440   Click image for larger version

Name:	180p Validation Snip.PNG
Views:	34
Size:	136.9 KB
ID:	2669442   Click image for larger version

Name:	968p Validation Snip.PNG
Views:	34
Size:	146.4 KB
ID:	2669444   Click image for larger version

Name:	968p Validation Part 2.PNG
Views:	31
Size:	128.6 KB
ID:	2669446   Click image for larger version

Name:	FCMM 968p Validation Snip.PNG
Views:	24
Size:	137.6 KB
ID:	2669448  

liberator72 is online now  
post #732 of 779 Old 01-17-2020, 06:34 AM
Member
 
Join Date: Dec 2019
Location: Finland
Posts: 22
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 13 Post(s)
Liked: 8
Although everything looks ok, i'm getting de of around 100 when verifying my LG C8 with i1Display pro (retail). Looking at the LUT, the cube looks ok, making a lut of verification profile looks like a complete cube in cube viewer...


I've triple checked all the settings (and completely calibrated 3 times) and always get the same results on verification run. The first two runs were with Teds recommended settings, the latest (files included) with Intelligent Integration=2 and finseqs 4913 CSV and augmented with Gray only large (the post verification is before augmentation, run out of time to verify augmented LUT). The PRE is after 3dLUT reset and white balance set via service menu (target 123nits).


Settings:
TV: all processing off, all power savings off including from service menu, gamma 2.2, br 52, cr 85 , expert dark PM and the TV was on long before calibration.

Drift Comp. 50
AIO mode
patch scale is 16-255 (PGenerator RGB FULL), TV back level Low.
Pre roll 2.0 s with Teds supplied CSV
Intelligent Integration 2 (was 1 for the firs two runs).
Bobofs Finseqs 20180603 (was Teds supplied CSV for the first two runs).


Verification with the same settings, except:
Drift Com. 0
patch scale is 16-235
Teds supplied CSV for verification with no pre roll.



LUT created to REC.709 (gamma 2.4) with peak chroma.


Updating the LUT via DeviceControl gave no errors and I even "rebooted" the TV for over two minutes for good measure the last time.


Everything on wired ethernet, rpi2b, all software latest versions (not todays LighSpace).
Any help appreciated.


Edit: Problem resolved. The calibration was fine, but the verification patchset was faulty.
Attached Files
File Type: zip C8.zip (891.0 KB, 12 views)

Last edited by Miksu2; 01-18-2020 at 06:02 AM. Reason: Resolved
Miksu2 is offline  
post #733 of 779 Old 01-17-2020, 09:49 AM
Senior Member
 
ebr9999's Avatar
 
Join Date: Jun 2013
Posts: 412
Mentioned: 15 Post(s)
Tagged: 0 Thread(s)
Quoted: 239 Post(s)
Liked: 123
I have given a look to "C8 1st verification".bcs. What patch set have you used to produce it? I see strange stuf inside .... Sparse patches as you have generated a bpd.

Try first with a stardard greyscale or anothere QP to check if you are OK, and then try with esotic patchsets ....
Miksu2 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

Last edited by ebr9999; 01-17-2020 at 10:45 AM.
ebr9999 is offline  
post #734 of 779 Old 01-17-2020, 10:58 AM
Member
 
Join Date: Dec 2019
Location: Finland
Posts: 22
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 13 Post(s)
Liked: 8
Quote:
Originally Posted by ebr9999 View Post
I have given a look to "C8 1st verification".bcs. What patch set have you used to produce it? I see strange stuf inside .... Sparse patches has you have generated a bpd.

Try first with a stardard greyscale or anothere QP to check if you are OK, and then try with esotic patchsets ....
Sorry about that, accidentally added new files to an old zip file instead of creating a new one. That was the first calibration I made (settings in the first post). The patch set was from Ted (bought LightSpace on the lasts days of last year) so about the most recent recommendation from his starter guide.


Only the newest files dated 16.1.2020 are using something other then recommended. Haven't even had the opportunity to run a verification on the augmented LUT. Although don't expect it to look any different. de 100 just isn't possible without seeing something odd. I've taken the augmented LUT in use and it's not showing anything odd either in normal use. So the meter has to return usable numbers or the LUT isn't uploaded...? Either or de of 100 is too far of to not be seen.



Added the grayscale large that I used for augmenting (can't run any new calibrations as the TV is in an open concept kitchen/dining/living room...


I'll add all the information from existing data that you need, but taking new measurements might take a while. So just let me know what measurements you need and I'll run them as fast as I can (can't run on nights what must be the most popular option).

/Miksu2
Attached Files
File Type: zip Augment grayscale.zip (2.3 KB, 3 views)
Miksu2 is offline  
post #735 of 779 Old 01-17-2020, 11:26 AM
Advanced Member
 
Make73's Avatar
 
Join Date: Jan 2010
Location: Finland
Posts: 509
Mentioned: 9 Post(s)
Tagged: 0 Thread(s)
Quoted: 225 Post(s)
Liked: 155
Quote:
Originally Posted by Miksu2 View Post
Although everything looks ok, i'm getting de of around 100...
Not look good at all, none of Your verifications.
Did you looked RGB Sep, DifGamma and Gamma ? Those are mess.

Integration time = 0.75
Extra Delay Time = 0.50
Stabilisation = 0.350

?
Miksu2 likes this.
Make73 is offline  
post #736 of 779 Old 01-17-2020, 11:45 AM
Member
 
Join Date: Dec 2019
Location: Finland
Posts: 22
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 13 Post(s)
Liked: 8
Quote:
Originally Posted by Make73 View Post
Not look good at all, none of Your verifications.
Did you looked RGB Sep, DifGamma and Gamma ? Those are mess.

Integration time = 0.75
Extra Delay Time = 0.50
Stabilisation = 0.350

?
Sorry for the incomplete settings, yes all correct as you guessed.
Everything else looks good except verifications. Added the earlier .bcs:s


Omitting the pre roll should only add minor errors, and above few de should be noticeable, so I would expect everything to look horrible if verification runs are correct. If there were problems with setup, that would also cause horrible results, but the picture if fine.


/Miksu2
Attached Files
File Type: zip C8 1st verification.zip (39.5 KB, 3 views)
File Type: zip C8 2nd verification.zip (39.5 KB, 3 views)
Miksu2 is offline  
post #737 of 779 Old 01-17-2020, 11:52 AM
Advanced Member
 
Make73's Avatar
 
Join Date: Jan 2010
Location: Finland
Posts: 509
Mentioned: 9 Post(s)
Tagged: 0 Thread(s)
Quoted: 225 Post(s)
Liked: 155
Quote:
Originally Posted by Miksu2 View Post
Sorry for the incomplete settings, yes all correct as you guessed.
Everything else looks good except verifications. Added the earlier .bcs:s


Omitting the pre roll should only add minor errors, and above few de should be noticeable, so I would expect everything to look horrible if verification runs are correct. If there were problems with setup, that would also cause horrible results, but the picture if fine.


/Miksu2
Okay so picture looks great but measures not.

I suppose you have rev.B version of probe which support AIO mode.

I had also same kind of mess with my first measures and wondered how good nerves Ted had when I always done some setting wrong
I´m sure you will figure it out.
What Ted said? Did he looked it with ColourSpace?

edit: One thing I had...faulty cat5 cable from PC to router so maby it´s worth of trying. It affected exactly same kind of mess. But since your profiling looks okay, I believe that´s not the case.
Miksu2 likes this.

Last edited by Make73; 01-17-2020 at 11:56 AM.
Make73 is offline  
post #738 of 779 Old 01-17-2020, 12:08 PM
Member
 
Join Date: Dec 2019
Location: Finland
Posts: 22
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 13 Post(s)
Liked: 8
Quote:
Originally Posted by Make73 View Post
Okay so picture looks great but measures not.

I suppose you have rev.B version of probe which support AIO mode.

I had also same kind of mess with my first measures and wondered how good nerves Ted had when I always done some setting wrong
I´m sure you will figure it out.
What Ted said? Did he looked it with ColourSpace?

edit: One thing I had...faulty cat5 cable from PC to router so maby it´s worth of trying. It affected exactly same kind of mess. But since your profiling looks okay, I believe that´s not the case.
The meter is a new (just double checked, it’s rev.B).

Ted has been a tremendous help, but as I don’t have the chance of running calibration runs that often, he recently haven’t had time (ColorSpace documentation) and I don’t want to distract him from that as picture is quite nice so this is low effort.

I’m sure the problem is quite academic and will soon be discovered.

/Miksu2
Make73 likes this.
Miksu2 is offline  
post #739 of 779 Old 01-17-2020, 12:15 PM
Senior Member
 
liberator72's Avatar
 
Join Date: Mar 2018
Location: Sunny (when it's not raining) Devon, U.K.
Posts: 235
Mentioned: 11 Post(s)
Tagged: 0 Thread(s)
Quoted: 173 Post(s)
Liked: 263
Quote:
Originally Posted by Miksu2 View Post
Sorry for the incomplete settings, yes all correct as you guessed.
Everything else looks good except verifications. Added the earlier .bcs:s


Omitting the pre roll should only add minor errors, and above few de should be noticeable, so I would expect everything to look horrible if verification runs are correct. If there were problems with setup, that would also cause horrible results, but the picture if fine.


/Miksu2
There's something wrong here with the grey scale patches for your verification??? On the left is your verification profile with only the grey scale patches selected. On the right is one of my 1000 point verification profiles, again with just the grey scale selected.

Sorry, I may have missed it, but what patchset have you used for verification?
Attached Thumbnails
Click image for larger version

Name:	Greyscale..PNG
Views:	28
Size:	449.6 KB
ID:	2671438  
Miksu2 likes this.
liberator72 is online now  
post #740 of 779 Old 01-17-2020, 12:30 PM
Advanced Member
 
Make73's Avatar
 
Join Date: Jan 2010
Location: Finland
Posts: 509
Mentioned: 9 Post(s)
Tagged: 0 Thread(s)
Quoted: 225 Post(s)
Liked: 155
There´s no black measure at all, 0.0646 is lowest luminance measure.
Miksu2 likes this.

Last edited by Make73; 01-18-2020 at 07:41 AM.
Make73 is offline  
post #741 of 779 Old 01-17-2020, 12:31 PM
Member
 
Join Date: Dec 2019
Location: Finland
Posts: 22
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 13 Post(s)
Liked: 8
Quote:
Originally Posted by liberator72 View Post
There's something wrong here with the grey scale patches for your verification??? On the left is your verification profile with only the grey scale patches selected. On the right is one of my 1000 point verification profiles, again with just the grey scale selected.

Sorry, I may have missed it, but what patchset have you used for verification?
Verification patchset Ted sent when I bought LightSpace a few weeks ago (not sending it online without his permission). So, taking a run with build in set would show if I got a bad file?

I’ll try persuade that eating breakfast in dark, but flashing is good for you and we should make it a habit on weekends.
Miksu2 is offline  
post #742 of 779 Old 01-17-2020, 12:42 PM
Member
 
Join Date: Dec 2019
Location: Finland
Posts: 22
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 13 Post(s)
Liked: 8
Quote:
Originally Posted by Make73 View Post
There´s no black measure at all, 0.0646 is lowest luminance measure.

Here is proper 1000p .csv for verifying.
Thanks, I’ll run as soon as possible. Just a thought as we’re talking about de of 100 here, would it be blasphemy to run with fast settings (you tested from Anger Miki) without that much warmup time with TV? That would only cause “a little more” error.

That might give me an early morning run ability.

/Miksu2
Miksu2 is offline  
post #743 of 779 Old 01-17-2020, 12:44 PM
Senior Member
 
liberator72's Avatar
 
Join Date: Mar 2018
Location: Sunny (when it's not raining) Devon, U.K.
Posts: 235
Mentioned: 11 Post(s)
Tagged: 0 Thread(s)
Quoted: 173 Post(s)
Liked: 263
Quote:
Originally Posted by Miksu2 View Post
Verification patchset Ted sent when I bought LightSpace a few weeks ago (not sending it online without his permission). So, taking a run with build in set would show if I got a bad file?

I’ll try persuade that eating breakfast in dark, but flashing is good for you and we should make it a habit on weekends.
Don't worry, I have all the sets Ted provides, I don't need you to post them. But I can tell you that the set you used for verification is not that set because the grey scale patches you see in my verification to the left are from that patchset. Yours looks more to me like the Pre Roll patchset he provides as the 240.240.240 patches are drift patches included within that set.
Miksu2 likes this.

LG OLED65C9MLB**Denon AVR-X3500H**Focal Sib Evo Atmos 5.1.2**Sky Q UHD**Xbox One X**Panasonic DP-UB820**ATV 4K**Harmony Elite
i1D3 OEM**i1D3 Retail**i1 Pro2 OEM**LightSpace HTP**CalMAN Home for LG**RPi Pgenerator**Ted's Disk & Media Files**S & M HD & UHD HDR Benchmark Disks**HD Fury Integral 2
liberator72 is online now  
post #744 of 779 Old 01-17-2020, 12:45 PM
Advanced Member
 
Make73's Avatar
 
Join Date: Jan 2010
Location: Finland
Posts: 509
Mentioned: 9 Post(s)
Tagged: 0 Thread(s)
Quoted: 225 Post(s)
Liked: 155
Quote:
Originally Posted by Miksu2 View Post
Thanks, I’ll run as soon as possible. Just a thought as we’re talking about de of 100 here, would it be blasphemy to run with fast settings (you tested from Anger Miki) without that much warmup time with TV? That would only cause “a little more” error.

That might give me an early morning run ability.

/Miksu2
You can use faster settings but I would keep integration time at 0.5 ,extra delay 0.5 ,stabilisation 0.150 (or greater to avoid meter errors with RPi)
For lowest possible dE´s I would use 2 for 1000p pre-roll.
In time you can see what´s best for your meter.
Miksu2 likes this.
Make73 is offline  
post #745 of 779 Old 01-17-2020, 12:49 PM
Member
 
Join Date: Dec 2019
Location: Finland
Posts: 22
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 13 Post(s)
Liked: 8
Quote:
Originally Posted by liberator72 View Post
Don't worry, I have all the sets Ted provides, I don't need you to post them. But I can tell you that the set you used for verification is not that set because the grey scale patches you see in my verification to the left are from that patchset. Yours looks more to me like the Pre Roll patchset he provides as the 240.240.240 patches are drift patches included within that set.
So just an error in transmission. I’ll run the set (from post above) as soon as possible.

Thank you all for the help.
/Miksu2
Miksu2 is offline  
post #746 of 779 Old 01-17-2020, 12:51 PM
Member
 
Join Date: Dec 2019
Location: Finland
Posts: 22
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 13 Post(s)
Liked: 8
Quote:
Originally Posted by Make73 View Post
You can use faster settings but I would keep integration time at 0.5 ,extra delay 0.5 ,stabilisation 0.150 (or greater to avoid meter errors with RPi)
For lowest possible dE´s I would use 2 for 1000p pre-roll.
In time you can see what´s best for your meter.
Thank you, much appreciated!
/Miksu2
Miksu2 is offline  
post #747 of 779 Old 01-17-2020, 12:58 PM
Senior Member
 
liberator72's Avatar
 
Join Date: Mar 2018
Location: Sunny (when it's not raining) Devon, U.K.
Posts: 235
Mentioned: 11 Post(s)
Tagged: 0 Thread(s)
Quoted: 173 Post(s)
Liked: 263
Quote:
Originally Posted by Miksu2 View Post
So just an error in transmission. I’ll run the set (from post above) as soon as possible.

Thank you all for the help.
/Miksu2
That's all it is. If you think about it, everything is going to based off of your Black & White levels. So all the luminance values of every colour is going to be calculated from that, and if you have no 0.0.0 and no 255.255.255 patches then the whole thing is a waste of time.

What you have is grey scale starting at 15.15.15 and a bunch load of repeated drift patches, so Gamma can't be calculated, and neither can colour luminance. This is why your dE values are all over the place. You actual profile looks good though so there isn't anything major to worry about, but I would just make sure you are using the correct patchset the next time you verify.
Miksu2 likes this.

LG OLED65C9MLB**Denon AVR-X3500H**Focal Sib Evo Atmos 5.1.2**Sky Q UHD**Xbox One X**Panasonic DP-UB820**ATV 4K**Harmony Elite
i1D3 OEM**i1D3 Retail**i1 Pro2 OEM**LightSpace HTP**CalMAN Home for LG**RPi Pgenerator**Ted's Disk & Media Files**S & M HD & UHD HDR Benchmark Disks**HD Fury Integral 2
liberator72 is online now  
post #748 of 779 Old 01-17-2020, 01:03 PM
Advanced Member
 
Make73's Avatar
 
Join Date: Jan 2010
Location: Finland
Posts: 509
Mentioned: 9 Post(s)
Tagged: 0 Thread(s)
Quoted: 225 Post(s)
Liked: 155
I got that wrongly named file also and reported back.
Ted have so much work to do and peoples to help so accidents may happen sometime
Miksu2 likes this.
Make73 is offline  
post #749 of 779 Old 01-17-2020, 01:10 PM
Member
 
Join Date: Dec 2019
Location: Finland
Posts: 22
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 13 Post(s)
Liked: 8
Quote:
Originally Posted by liberator72 View Post
That's all it is. If you think about it, everything is going to based off of your Black & White levels. So all the luminance values of every colour is going to be calculated from that, and if you have no 0.0.0 and no 255.255.255 patches then the whole thing is a waste of time.

What you have is grey scale starting at 15.15.15 and a bunch load of repeated drift patches, so Gamma can't be calculated, and neither can colour luminance. This is why your dE values are all over the place. You actual profile looks good though so there isn't anything major to worry about, but I would just make sure you are using the correct patchset the next time you verify.
That sounds so right. Everything is looking alright except the verification, there’s no need to look further. I’ll test the “hypothesis” as soon as possible to make sure.

The amount of knowledge in here is invaluable to firstimers. Thank you so much for bearing with us.
/Miksu2
Miksu2 is offline  
post #750 of 779 Old 01-17-2020, 01:14 PM
Member
 
Join Date: Dec 2019
Location: Finland
Posts: 22
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 13 Post(s)
Liked: 8
Quote:
Originally Posted by Make73 View Post
I got that wrongly named file also and reported back.
Ted have so much work to do and peoples to help so accidents may happen sometime
Sure are waiting for the first release of ColorSpace so much that don’t want to bother Ted with my problems. Actually looking these threads, I might be able to see errors better with ColorSpace as I’m used to seeing those graphs
/Miksu2


Edit:
Have to save one .bcs so I can have a an very exotic christmas tree next year. Right color for the season (not for the tree), maybe an nice overlay on top of a new calibration?
Make73 likes this.

Last edited by Miksu2; 01-17-2020 at 01:27 PM.
Miksu2 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