AVS Forum banner

161 - 180 of 201 Posts

·
Registered
Joined
·
1,235 Posts
Sorry but I don't believe anything regarding any testing when it's coming from SpectraCAL.
Oh, I'm not going to try arguing with you about any of your views of them. ;)
When you measure just grayscale, you will not see a difference if you compare 8-bit RGB vs. YCbCr 4:2:2 REC.2020.

When you measure pure primary colors, with 100% saturation, you will not have any difference between RGB 8-bit vs. 10-bit YCbCr 2020.
That's certainly reasonable theory, although I'm surprised you seemingly assert it as universally true of all displays without testing from verified contacts. ;)
The colorspace differences will appear when you will measure non 100% Saturation colors
Which I think is fairly common for HDR CMS calibration.

(I'm done.)
 

·
Registered
Joined
·
1,055 Posts
Discussion Starter #162 (Edited)
Since my TV watching is done via blu-ray player (UB9000) and it outputs YCbCr 4:4:4....I would need Ted's disc in order to confirm the conversion is still accurate within the player and TV, right?
Yes, using a disk from a player for checks and verification is a good way to check everything has gone as expected. It is what I do myself.

I know Sony’s handle the RGB Full signal just fine (with video levels configured from software).
LG OLED’s actually calibrate better with RGB Full and either video or video extended levels set, especially for LUT calibration. In fact, the iTPG within the 9 and X series TVs are full range RGB TPG’s.
Heck, even my old Panny Plasma that’s tucked away upstairs (obviously SDR only) handles RGB Full with levels configured in software correctly provided the HDMI input is set to receive a limited signal (which can be tricky as it always defaults to receiving a PC signal on one of my HDMI inputs).

With SDR you will be perfectly fine, it’s HDR where things start to get messy and that’s due to current displays being sorely lacking when it comes to HDR capabilities. There really isn’t much you can do with regards to accurate HDR adjustments. Lack of gamut coverage, lack of peak brightness, display instability all mean it’s pretty much guesswork at the moment until display technology progresses.

Sony, for example, know this. The HDR (and Dolby Vision) calibration is based from the underlying SDR calibration. If you SDR calibration is as good as you can get it, then the HDR (and Dolby Vision) calibration will be as good as it can be on that TV after the internal processing does its thing. Just read anything D-Nice has had to say about that on recent models and you will see how much he respects Sony for this approach.

LG on the other hand allow for independent adjustments in both HDR and Dolby Vision. But all that is done in either of these is at most a multipoint greyscale adjustment along with measurement of the the peak primary colours, which some of us advise against doing altogether in HDR anyway as the factory colour calibration is already very good. For Dolby Vision on LG, with it using the TV decoded version, both test patterns and actual content it is sent in an 8-bit RGB Full container anyway! Even from a Murideo SIX-G.

What Samsung do is unknown to me. I am not a fan of their products and do not have anything to do with them. But I was under the impression that the calibration controls on Samsung displays are somewhat funky in the first place from what I have read, even in SDR. I could be wrong on that, I have no idea and have no desire to find out.

The long and the short of it is that not everyone has the money or even the inclination to go out and buy an expensive external TPG that is accurate in every colour space output. There are solutions, such as streaming devices, that you can cast, mirror or even install an app onto that will provide automatic pattern generation, but many of these introduce significantly more errors than if you were to just use the HDMI output of your laptop. Many laptops output accurately when configured to RGB Full 8-bit, and it has been a highly recommended way to generate patterns for years, including by various software vendors themselves. But not all laptop HDMI outputs are accurate, and their accuracy can be affected by software updates.

Which brings us back to the point of the PGenerator. All it does is provide a known quantity. Accurate RGB Full patch generation in the same way that has been recommended for years, but without the risk of being affected by an update from Microsoft, or Intel etc.....Plus it has the ability for custom features to be added, such as the recent Ted's Disk Template and potentially other features in the future. It is not intended to replace or even compete with devices that cost hundreds, sometimes thousands of $$$ (or £££ in my case). It is an affordable solution that anyone can afford, and this has been made clear straight from the very first post so I find it highly disappointing that the discussion has now taken this tangent.
 

·
Registered
Joined
·
9,365 Posts
Oh, I'm not going to try arguing with you about any of your views of them.
They are not my views.

Since about 10 different people have tested with digital pixel analyzers and about 50 have posted more simplified tests with measurements, its the reality.

So it's not a personal view, as the data have posted to several threads.

When a company wants to sell software, they can recommend that they have bit-perfect patch generation solutions available (which will require to buy software to use them).

After some thousands of people bought, users have found that they are not bit-perfect, so the person who said that he has verified with multiple analyzers was laying to make enthusiasts buy specific software.

That's certainly reasonable theory, although I'm surprised you seemingly assert it as universally true of all displays without testing from verified contacts. ;)

Which I think is fairly common for HDR CMS calibration.
RGB-Video vs. YCbCr 2020, as you see to your measurements, the RGB balance/gamma is nearly identical (if you count the instability of the panels in HDR mode.

Someone familiar with the conversion matrix of RGB to YCrCr will understand better.

For example, YCbCr 709 with YCbCr 2020, they have the same YCbCr triplets when you measure Grayscale or 100% Saturation of Primary Colors.

You can use a 'dirty trick' to have reference YCbCr 2020, if you want just to perform grayscale calibration for HDR.

If you use a Blu-ray player with YCbCr 4:4:4 output (any OPPO) and playback a calibration disk, which is compatible/bit-perfect for CalMAN (let's say Ted's disk as AVSHD is not bit-perfect for Grayscale with CalMAN/ChromaPure for 5%-50% patches):



...then the player will output YCbCr REC.709 triplets, add upscale to 2160p24 if you like, or use 1080p24.

If you add to the output of that OPPO player an HD Fury device and inject HDR10 metadata and a special AVI info frame that you are sending YCbCr REC.2020 4:4:4, the display will trigger the REC2020 gamut.

For 1080p24 YCbCr REC.2020:

82:02:0d:e3:40:c8:64:20:00:00:00:00:00:00:00:00:00

For 2160p24 YCbCr REC.2020:

82:02:0d:a6:40:c8:64:5d:00:00:00:00:00:00:00:00:00

Using that trick, you can have bit-perfect YCbCr REC.2020 for Grayscale calibration only, not for gamut measurement, ColorChecker, saturation sweep of P3 inside REC2020 etc.

Only if you measure 100% Saturation with 100% Stimulus, for Red, Green, Blue, only these 3 color patches will match the YCbCr 2020 triplet.

For most of the displays, we don't touch CMS controls, as the TV internal tone mapping is responsible for the gamut mapping in HDR.

That 'dirty' trick is an example of what you can do when you know some more advanced details about how stuff works.
 

·
Registered
Joined
·
9,365 Posts
Not to sidetrack the conversation, but want to make sure I'm understanding correctly.

OK so I calibrated my Sony A9G using Autocal feature with Calman from Home....using the bit-perfect (RGB Full) PGenerator from RPi, and my settings with Calman and TV were the 16-235/Limited per their instructions.

Since my TV watching is done via blu-ray player (UB9000) and it outputs YCbCr 4:4:4....I would need Ted's disc in order to confirm the conversion is still accurate within the player and TV, right?
Hi,

When you are performing AutoCAL, it will require to have a verified for its accuracy pattern generator.

The adjustments the AutoCAL will perform to your display, its the same adjustments you can apply manually, using the TV remote, or using the DDC of CalMAN for these specific points.

The point of AutoCAL is to speed up the procedure when you have limited time.

However, when you perform 10x AutoCAL, you will see that the adjustment values will be different; your reports will be different also.

People who don't have time to learn how to calibrate manually, they prefer AutoCAL as its faster procedure.

But when you have bought a meter and CalMAN for Sony and want to calibrate a different TV brand to another room, a Samsung or a Panasonic, you will have to purchase CalMAN for Panasonic and CalMAN for Samsung.

When you want to calibrate a different TV, like VIZIO, Philips, TCL, since the user will luck the basic and the procedure of calibration, he will not be able to calibrate any display manually (or projectors).

So I'm advising the users, even when they have AutoCAL functions for their TV to start learning the manual calibration procedure, to improve their understanding and knowledge about calibration.

After some practice, they will know much better to calibrate, and they will have better results with manual calibration.

If you have that time and interest to learn, spend some time to use your CalMAN for Sony to calibrate the calibration controls manually using the DDC controls from CalMAN.

When you are calibrating using an external generator (any model/brand, 1$, or 10K$ model), the movie player's sources have to be verified after the end of calibration to see if they have correct output.

Some player enhancements (enabled by default to players, can alter the player output.

For example, the Sony players have a Theater mode. This function will affect the player output, making the calibration you have performed using an external generator invalid for that player.

The basic test the users have to do is to check the player's levels if they match the levels they see from the external generator. (contrast/brightness/color clipping...no meter required)

For more advanced testing, measuring with a meter for grayscale/color gamut will be an excellent test to verify all working as expected.

Usually, users have to copy the settings to other HDMI inputs when they have multiple sources. That test can confirm that the settings have been transferred with success.
 

·
Registered
Joined
·
45 Posts
Hello Everyone!

I have an opportunity to purchase a new HDFury Linker 4K60 4:4:4 600MHZ or a new Integral 4K60 4:2:0 300MHz for the same price. Is there a preferred choice for which would be better to use now and in the long run with Pgenerator?

Any pros and cons to either one? Thanks in advance!
 

·
Registered
Joined
·
9,365 Posts
Hello Everyone!

I have an opportunity to purchase a new HDFury Linker 4K60 4:4:4 600MHZ or a new Integral 4K60 4:2:0 300MHz for the same price. Is there a preferred choice for which would be better to use now and in the long run with Pgenerator?

Any pros and cons to either one? Thanks in advance!
Hi,

HD Fury Linker is a better option, as it has upscaling also.

I have verified with the digital pixel analyzer that the upscale will not alter the pattern's digital values.
 

·
Registered
Joined
·
9,520 Posts
The Raspberry Pi PGenerator was already the absolute best “bang for your buck” bit accurate TPG. The extremely affordably cost makes it an essential piece of kit for any enthusiast. The addition of Ted's LightSpace CMS Calibration Disk, PGen Edition takes things up to a whole new level with built in, verified accurate Pre/Post-Calibration test patterns. The fact that the PGenerator is fully working with LightSpace/ColourSpace, CalMAN and HCFR calibration software solutions makes Ted's LightSpace CMS Calibration Disk, PGen Edition an absolutely essential addition.
Just a small post to say a large "thank you" to @ConnecTEDDD , Riccardo, @liberator72 and of course everyone else who was involved in bringing the Raspberry Pi version of Ted's disc into existence!

I've been using these PGen patterns over the last few days. One thing I didn't expect to be so useful that I really like is the patterns without text. For example, the "Advanced Brightness Flashing Bars" pattern, without text. To make a judgement by eye of what is perfect black, and what is not black, can be very "challenging" (is "16" visible or not; have I just raised the whole background level to be above black, etc). ;)
Even though I've used Ted's normal disc patterns for years now, I have found that I can be more accurate more easily, using the non-text versions, which wasn't possible before. This is great! Without the normal text versions nearby to refer to, it would be tricky - so having them both available is perfect :)
 

·
Registered
Joined
·
123 Posts
I would like to thank @ConnecTEDDD @liberator72 for the wonderful job and of course to PGenerator creator Biasi...
Everything is working great, all patterns on display interacting with the calibration software without disconnect them ever. Is also the first time that I can see a very "clean" color ramp horizontal&vertical pattern using 3DLUT on eecolor aswell. blacktunnel static can show me a very delicate level 17 Gray dificult to see in a dark pitch room on my older B6 Oled ^__^ and also Blue tunnel looks without any color banding everyting looks very clean and smooth, sing the 3DLUT is working properly, not too much my Sony Blu-ray player USB port using the same pattern to test after calibration job.
So, I have to say this upgrade is a MUST, to any enthusiast or even professional calibrator...
A big applause all of you guy's and thank you for this fantastic job and of course for all your support anytime!!!!

THANKS!!!
 

·
Registered
Joined
·
9,365 Posts
I've been using these PGen patterns over the last few days. One thing I didn't expect to be so useful that I really like is the patterns without text. For example, the "Advanced Brightness Flashing Bars" pattern, without text. To make a judgement by eye of what is perfect black, and what is not black, can be very "challenging" (is "16" visible or not; have I just raised the whole background level to be above black, etc). ;)
Even though I've used Ted's normal disc patterns for years now, I have found that I can be more accurate more easily, using the non-text versions, which wasn't possible before. This is great! Without the normal text versions nearby to refer to, it would be tricky - so having them both available is perfect :)
I'm glad you find them useful.

Flashing Bars Patterns without Text Indication


They were 'last-minute addition' these variations of the patterns without text.

They were necessary with modern displays with deep blacks like OLEDs.

Any other display tech panel or projector user will appreciate that addition and found them useful also.

They are available all the patterns with flashing bars for brightness without text.

The user will have the capability to select the regular version (as static bars or flashing bars) with gray text indicators of the bars (as were available to the standard calibration disk):



or to use the new version with ''No Text''.

You can see that small 10x10 pixel box indicator to that animated PNG:



''No Text'' edition will have only a 10x10 pixel indicator of about 5% gray at the top and bottom of the Reference Black bar and no other text.

Selectable Flashing Bars Timing Ratio

The Delay Flashing Pattern value allows the user to define the timing ratio of flashing bars.

Using a 240 ms setting (the default), it will match the 1-sec interval flashing rate of the same pattern as available to 24p video stream edition of the pattern, calibration disk or media files.



But that function is providing the user to capability to the user to define and change the speed.

I believe it's a good idea to provide increased flexibility, and a feature is not available on costly generators.

Its available button to display the static version (with bars frame) or inject a black frame, to stop the animation.

Black Frame Injection


The black frame injection is essential during a calibration, after an HDR measurement, for the panel to cool-down or after the end of some measurements, when the user wants to open the device OSD to change settings, etc.

Custom Pattern Library


That new development (thanks to the developer of PGenerator) is opening some further capabilities to enthusiasts.

The 'Custom Pattern Library' provides the capability to the user to add his custom images.

That function has no limit, and it's up to the SD card empty space capability.

For example, a 16GB SD card (which costs about 5$) will have space for about 1000 TGA uncompressed images to upload some thousand PNG custom images.

We recommend TGA files to initialize lower CPU/GPU resources for the PGenerator to draw it.

TGA files are uncompressed while PGN are compressed but lossless.

For example, a lot of times, users are using their own reference movies for color evaluation.

Since commercial hardware solutions will have Copywrite problems to include screenshots from commercial movies to generators, users can load their own still images to their Library.

It will require attention for these screenshots; if the user will playback a blu-ray movie to his HTPC, the software player needs to be configured to Video Levels, not to expand to PC Levels.

The black level or the 2.35:1 bars of the movie should look like Gray.

We do this for capturing video levels, as by default, software players will expand the video to PC levels.

It will provide quick access to inject reference still pictures; various content users may be more familiar to check for skin tones etc.

Users can create their own pattern using Photoshop or with any other software they prefer.

They can use 0-255 values, but the 16 should be reference black and 235 as reference white.

If you have found a reference image to the web (like this nice image from Burosch), reference images from a company, standards, reference images from manufactures of devices, these patterns should be full range most of the time.

You can convert them to video range using PhotoShop, using the Levels window.



You can check with the color picker some white or black pixels to see if they use 0 for black and 255 for white and then use Levels to compress them to 16-235.

Then save as TGA 24-bit or TIFF, BMP, PNG 8bit 1920x1080.

For uploading or deleting custom uploaded images, these actions will be available from inside the Template.



Browse for an image to upload or delete from the drop-down list the image you want.

 

·
Registered
Joined
·
9,365 Posts
Is also the first time that I can see a very "clean" color ramp horizontal&vertical pattern using 3DLUT on eecolor aswell. blacktunnel static can show me a very delicate level 17 Gray dificult to see in a dark pitch room on my older B6 Oled ^__^ and also Blue tunnel looks without any color banding everyting looks very clean and smooth, sing the 3DLUT is working properly, not too much my Sony Blu-ray player USB port using the same pattern to test after calibration job.
So, I have to say this upgrade is a MUST, to any enthusiast or even professional calibrator...
A big applause all of you guy's and thank you for this fantastic job and of course for all your support anytime!!!!
Thank you for your good words.

The patterns look more detailed, as they are uncompressed versions.

Uncompressed patterns are available to expensive external pattern generation also.

Since the consumer industry is using 8-bit YCbCr 4:2:0 compression for blu-ray, they look 'less detailed' after the compression.

You will see for sure improvement to gradation patterns with colors, to advanced sharpness or other patterns in uncompressed versions.

But it's important to evaluate with YCbCr 4:2:0 patterns from the movie sources players also for checking the upsampling the TV or player is doing.

I mean, while to have uncompressed patterns is good capability, it will not make 'unless' the YCbCr pattern evaluation from blu-rays or Media files.
 

·
Registered
Joined
·
51 Posts
Thanks @liberator72 for the detailed setup guide for PGenerator.
I have been struggling the whole day in getting all to work with Colourspace, I have eventually found the problem, I am sad to say that it seems to be with the image for PGen that you linked to or that my image perhaps become corrupt. After ironing out a lot of issues I eventually got Colourspace to connect to PGen, but the patterns were not updating. I eventually flashed the SD card with the image link provided by Lightillusion and bang it works perfectly.
Just one thing I will suggest considering when installing and/or running colour management software and the PGen software, change the user type to administrator under any windows version, doing this should provide access to set different networking options, firewall rules, etc...
 

·
Registered
Joined
·
1,055 Posts
Discussion Starter #172
Thanks @liberator72 for the detailed setup guide for PGenerator.
I have been struggling the whole day in getting all to work with Colourspace, I have eventually found the problem, I am sad to say that it seems to be with the image for PGen that you linked to or that my image perhaps become corrupt. After ironing out a lot of issues I eventually got Colourspace to connect to PGen, but the patterns were not updating. I eventually flashed the SD card with the image link provided by Lightillusion and bang it works perfectly.
Just one thing I will suggest considering when installing and/or running colour management software and the PGen software, change the user type to administrator under any windows version, doing this should provide access to set different networking options, firewall rules, etc...
Could you please tell me which version you downloaded. By that, I mean when you downloaded from the link in the first post? The image I was hosting worked perfectly fine for several users so was not corrupt, the current image is not hosted by me so if it is that one I will need to check it.
 

·
Registered
Joined
·
9,365 Posts
A good idea is for Leon to add to the first page a link for the following site:


The MD5 hash checksum key of the 'correct' ISO file.

After a download, users can check if the file is 'good' by performing that on-line check.
 

·
Registered
Joined
·
51 Posts
Could you please tell me which version you downloaded. By that, I mean when you downloaded from the link in the first post? The image I was hosting worked perfectly fine for several users so was not corrupt, the current image is not hosted by me so if it is that one I will need to check it.
As I mentioned it could have been corrupted somehow (after download), perhaps if someone else has the same issue and posts here then there will be need for concern, but after Ted's post we now know of a way to check the download. I also changed my user type to admin and restarted the laptop - Windows 10, that could also have been the fix.
 

·
Registered
Joined
·
2,722 Posts
As I mentioned it could have been corrupted somehow (after download), perhaps if someone else has the same issue and posts here then there will be need for concern, but after Ted's post we now know of a way to check the download. I also changed my user type to admin and restarted the laptop - Windows 10, that could also have been the fix.
Hi,
I just used it the other day with no issues.
 

·
Registered
Joined
·
1,055 Posts
Discussion Starter #176
As I mentioned it could have been corrupted somehow (after download), perhaps if someone else has the same issue and posts here then there will be need for concern, but after Ted's post we now know of a way to check the download. I also changed my user type to admin and restarted the laptop - Windows 10, that could also have been the fix.
I have just downloaded the image from the first post, flashed it to a seperate SD Card and can confirm it is working correctly. It must have somehow become corrupted on your system.
 

·
Registered
Joined
·
426 Posts
So I'm having a hard time getting PGenerator to connect to Colourspace using a wired connection to PC. I looked at mrtickluk's problem and fix and am still not sure if that is my problem or not. I took pics of what I'm getting in colourspace and firewall rules.

Edit: I am just running through mrtickleuk's suggestions for firewall rules and so on. In Calman I had to manually enter IP address, do I need to do that somewhere in Colourspace as well?
 

Attachments

·
Registered
Joined
·
1,055 Posts
Discussion Starter #179 (Edited)
I believe Ted already replied to you in the ColourSpace thread, but looking at your Firewall Rules briefly (sorry I am not home and on my phone browser it's not easy to see), it looks like you have two instances of ColourSpace and two instances of Colour Management System.

I'm not really familiar with Windows Defender Firewall rules/exceptions as I use a third party security system, but they seem to be conflicting. One of each is set to Private, and one of each is set to Public. Either try changing the ones currently set to Public over to Private, or removing the one ones set to Public altogether.

PGenerator needs to be connected in a Private Network when operating via the Direct LAN connection method (ethernet connection between computer and Raspberry Pi).

Edit: I have just re-read your other post In the ColourSpace thread. You seem to be able to successfully connect with Device Control and use Ted's patterns okay if I understand it correctly? If that is the case then it seems like your Firewall is indeed blocking the ColourSpace software connection, so the above suggestion could indeed resolve your issue.

Colour Management System will most likely relate to the LightSpace Software, whereas ColourSpace is obviously self explanatory.

Please also see this post a few above yours, where the user had to change his Windows account type to Administrator to access options for Firewall rules.

As soon as I find some free time I will consolidate some connection troubleshooting tips and create a more detailed post which I can link at the end of the first post.
 
  • Like
Reactions: ConnecTEDDD
161 - 180 of 201 Posts
Top