AVS Forum banner
421 - 440 of 1,661 Posts
Discussion starter · #421 ·
HDR10 21-Point Grayscale for specific 'nit' levels.

As we know in the HDR measurements world, 235 (8-bit) in LEGAL range represent 100% of the PQ EOTF with 10000 nits as a target.

When we want to measure the HDR grayscale performance, the standard method is to measure with 21-Point (16-235), so 0-10000 nits range.

However, movies mastered mainly using 1000/4000 nits mastering monitors (2000 nit also and soon 5000 nits), so any measurement above that peak output is not really required.

The 'regular/defacto' method we use to generate patterns is to send 1000 or 4000 nits mastering monitor peak metadata using an HDR10 Pattern Generator, PGenerator with HD Fury, or PGenerator with LG's 2020/2021 HDMI Override.

Since the typical method of evaluating HDR10 Grayscale has a 0-10000 nit range, the patch generator will display a lot of patterns above 1000 or 4000 nits, so we are using fewer points at the range we are interesting to measure.

For example, see a regular 0-10000 nit (16-235) patchset below:



The 16th patch (0 is black) will generate 180 code level, with a target Y of 973.123 nits. It's 74.89% of the PQ.

When we want to measure with 1000 nit metadata, we are interested to see how our TV responds to a 1000 nit movie metadata.

When we use a bright LCD/LED with a peak output higher than 1000 nits and the TV can count the incoming metadata from the HDMI input, it will not apply any roll-off to the PQ EOTF.

When we use a WRGB based OLED, the peak output is less than 1000 nits; the TV will apply a roll-off to avoid clipping of highlights until 1000 nits.

LG OLEDs 2019/2020/2021 have PQ Upload parameters; we can upload the calibrated peak output nits number and decide if we want to clip or roll-off or define the percentage the TV will start to roll off. (2018 LG's have a default 750 nit number baked, we can't change that)

However, when we take measurements of 21-Point Grayscale using regular methods, we waste five patches; as you can see, only 16 patches will measure the TV performance up to 1000 nits.

When we want to limit the peak level of patterns to a specific nit level, then we can use the Limit Range @ ColourSpace Settings.

Using ''Limit Range'' we can define the max 8-bit digital code values we want as our maximum generated patch, so our 21-Point Grayscale will measure only the grayscale steps only at the range we want to evaluate.

By enabling Limit Range with 181 value, we can take 21-Point Grayscale for 0-1000 nit range, so we will have more points to measure and see with more detail when the roll-off will start or how the display is tracking the PQ up to 1000 nits.

It's not only a waste of time with regular workflow, but we are also increasing the panel temperature because we will measure patches our display can't reproduce or we are interested in measuring.

Below you will see some examples of Limit Range values you can set to limit the measurements to specif nits, for various TVs evaluation:

~100 nits = code value 127, its the 50.7% of PQ (98.8 nits)
~200 nits = code value 143, its the 58% of PQ (201.5 nits)
~300 nits = code value 152, its the 62% of PQ (297.6 nits)
~400 nits = code value 159, its the 65.3% of PQ (401.5 nits)
~500 nits = code value 164, its the 67.6% of PQ (496.4 nits)
~600 nits = code value 168, its the 69.4% of PQ (587.8 nits)
~700 nits = code value 172, its the 71.2% of PQ (695.6 nits)
~750 nits = code value 174, its the 72% of PQ (756.6 nits)
~800 nits = code value 175, its the 72.6% of PQ (789.1 nits)
~850 nits = code value 177, its the 73.5% of PQ (858.1 nits)
~900 nits = code value 178, its the 74% of PQ (894.9 nits)
~950 nits = code value 179, its the 74.4% of PQ (933.2 nits)
~1000 nits = code value 181, its the 75.3% of PQ (1014.8 nits)
~2000 nits = code value 198, its the 82.6% of PQ (1982.8 nits)
~4000 nits = code value 214, its the 90.4% of PQ (4057.5 nits)

How to find the 8-bit value of the specific nit level we want?

Set @ ColourSpace ''Settings'' the 'ST.2084 REC.2020' as Target Gamut / EOTF.

Patchscale Legal / 8bit:



Tick from ''Manual Measure'' the radio buttons of the R/G/B color channels to see grayscale triplets.

Use the up/down arrow keys from the keyboard to change the +-1 digital 8-bit value while you watch the Target Y.

The 170 8-bit code value has 639 nits target Y:



How to perform manual measurements for 21-Point Grayscale for specific nit level?

If you want to limit the displayed patches to 750 nits, find the closer 8-bit digital code value using the arrow keys, the 174 (756 nits), see the animated PNG:



Go again to the ColourSpace ''Settings'' and enable 'Limit Range' 174.



Go to 'Manual Measure window:



Define the steps you want, for example, 21-Point.



Move the RGB sliders at the lowest position (16.16.16).

Press to measure; it will measure black.

You can select the next patch using a left mouse click at the area with the red dot.

You can select the next/previous point using single left mouse clicks at the area above/below the three sliders.

How to perform HDR10 21-Point Grayscale sweep using a patchlist for specific nit levels?

Using the ''Limit Range'' function of ColourSpace, I have created 15 HDR10 patchlists for 21-Point Grayscale.

ColourSpace users can load any of those patchlists to measure 21-Point HDR10 Grayscale with a limit to the nits they want.

It will not require to enable Limit Range.

Select the ST.2048 REC.2020 as target colorspace, keep the Limit range unticked, select LEGAL patchscale for PGenerator users (or other pattern generator users with video level output) @ ColourSpace Settings.



Add metadata using HDFury GUI to the HD Fury Device you have connected at the output of PGenerator or use LG 2020/2021 HDMI Override menu to force/setup HDR10 metadata details.

Use the Manual Measure Window and load each patchset; I have posted picture examples about how to load a patchlist in this post's ''instructions'' area.

HDR10 21-Point Grayscale Patchlist Download Link

You can download the ZIP with 15 HDR10 21-Point Grayscale Patchsets here.
 
Discussion starter · #422 · (Edited)
HDR10 4-Point Saturation (P3-D65 inside REC.2020) Patchsets for specific nits

HDR10 movies mastered mainly using 1000/4000 nits capable mastering monitors (only a few with 2000 nit and soon 5000 nits).

Currently, the available reference HDR10 broadcast monitors used for the color grade of HDR10 movies can cover the P3D65 gamut.

These monitors are calibrated for P3D65 1000/4000 nits, and when the movie is ready, they map P3D65 1000/4000 nits inside a REC.2020 10000 nit container.

When we want to evaluate the gamut performance of a calibrated HDR10 TV, we are interested to see how the TV is tracking the Saturation levels of P3D65 inside REC.2020 colorspace.

The typical Saturation Sweep mainly used has a 50% Luminance level (~100) nits.

I have created patchlists for ColourSpace for the users to take 4-Point Saturation sweeps for different Luminance levels.



The Saturation sweep measures the x-Steps from 100% Saturation with a specific % luminance level of the primary/secondary colors towards to White.

About the Luminance levels of the Saturation sweeps:

~100 = 50% of PQ (94.4 nits) usually this one is the standard measurement run.
~100 = 50% of PQ (94.4 nits50.7% (98.8 nits) this one is closer to 100 nits.
~200 = 58% (201.5 nits)
~300 = 62% (297.6 nits)
~400 = 65.3% (401.5 nits)
~1000 = 75.3% (1014.8 nits)
~2000 = 82.6% (1982.8 nits)
~4000 = 90.4% (4057.5 nits)

The recommendation for WRGB OLEDs, due to the problem with low color volume due to WRGB panel structure, is to use 100/300 or 400 nits Saturation Sweeps.

For LCD/LED high-brightness TVs, they can use all options, except 4000 nits.

Projector users who use the HDR-Multiplier calibration method with ColourSpace can use the 1000/4000 nits.

Setup Instructions

Select the ST.2048 REC.2020 as target colorspace, keep the Limit range unticked, select LEGAL patchscale for PGenerator users (or other pattern generator users with video level output) @ ColourSpace Settings.

Add metadata using HDFury GUI to the HD Fury Device you have connected at the output of PGenerator or use LG 2020/2021 HDMI Override menu to force/setup HDR10 metadata details.

Use the Manual Measure Window and load each patchset; I have posted picture examples about how to load a patchlist in this post's ''instructions'' area.



HDR10 4-Point Saturation (P3-D65 inside REC.2020) Patchsets Download link

You can download the ZIP with 8 4-Point Saturation (P3-D65 inside REC.2020) Patchsets here.
 
Discussion starter · #423 ·
Windows 11

For users that will upgrade their current Windows to W11, before starting the procedure, de-activate the ColorSpace License and re-activate it after the W11 upgrade.
 
I finally got around to trying out ColourSpaceZRO + PGenerator + Lightspace CMS Calibration Disk (thanks for all the hard work on this, Ted!) with my Sony XBR-Z9F, and calibrated for the SDR 20-point grayscale values with the Sony grayscale CSV. (attached are my before/after calibration images)

I did have a few noob questions:

ColourSpaceZRO - Recommended settings as starting point
...
Profile Luma Auto, tick when you want from ColourSpace to continuously update the targets for black/white from these measurements.
  1. I think I forgot to tick this setting today. Does this Profile Luma setting affect anything for the RGB u/v calibration, or is it just for adjusting the Y luminance target?
  2. I was trying to calibrate SDR to 200 nits instead of 100 nits (because I'm in a fairly bright room for viewing - so I set the TV's Brightness setting to 10 instead of 2 for Peak White). Am I doing this correctly by setting the brightness first and calibrating for that, or should I calibrate for 100 nits first then adjust the brightness after?
  3. I'm still trying to figure out how to do the Gamut/color calibration ("Per color adjustment" settings on the Sony settings). I see the calibration disk files include patchlists for colour gamut (75/100% Saturation, 75/100% Stimulus Level). I saw the ZRO guide says to use the 75% patches - is the idea similar to the grayscale calibration, where I change the Hue and Saturation values to lower the deviation for u/v and change the Lightness value to lower the deviation for Y?
 

Attachments

Discussion starter · #425 ·
I finally got around to trying out ColourSpaceZRO + PGenerator + Lightspace CMS Calibration Disk (thanks for all the hard work on this, Ted!) with my Sony XBR-Z9F, and calibrated for the SDR 20-point grayscale values with the Sony grayscale CSV. (attached are my before/after calibration images)

I did have a few noob questions:


  1. I think I forgot to tick this setting today. Does this Profile Luma setting affect anything for the RGB u/v calibration, or is it just for adjusting the Y luminance target?
  2. I was trying to calibrate SDR to 200 nits instead of 100 nits (because I'm in a fairly bright room for viewing - so I set the TV's Brightness setting to 10 instead of 2 for Peak White). Am I doing this correctly by setting the brightness first and calibrating for that, or should I calibrate for 100 nits first then adjust the brightness after?
  3. I'm still trying to figure out how to do the Gamut/color calibration ("Per color adjustment" settings on the Sony settings). I see the calibration disk files include patchlists for colour gamut (75/100% Saturation, 75/100% Stimulus Level). I saw the ZRO guide says to use the 75% patches - is the idea similar to the grayscale calibration, where I change the Hue and Saturation values to lower the deviation for u/v and change the Lightness value to lower the deviation for Y?
Hi,

1) ColourSpace will need to use the measured black/white level when you are calibrating with a non-absolute colorspace target, a non-ST.2084 HDR-based colorspace, like REC.709.

When the AUTO @ Profile Luma is ticked, every time you will measure your black/white, the software will use these values to calculate targets and plot the charts.

Keep AUTO enabled for the whole calibration session or a verification session.

When you want to start your calibration and pre-calibrate your 100% Reference White (235.235.235 with Legal patchscale), then you can disable AUTO and set the values of peak output you want manually.

Enter the nits values you want for your MAX and then use the Manual Measure and look at the RGB Widgets to help you reach your target Y and fix the RGB balance of your reference white.

After that procedure, you can enable AUTO for the remaining calibration.

Usually, many display models have both 2-Point and 10/20-Point RGB balance controls.

When 2-Point controls are available, you can't fix gamma errors.

You can disable the Y errors from the RGB Widget by enabling the 'Chroma Only'.

Chroma Only will plot only RGB balance errors to the RGB bars of the RGB Widget.



After the 2-Point calibration, you can disable Chroma Only and continue with the 10/20-Point calibration.

2) Pre-define your Peak output levels while you perform your Peak White calibration.

After the end of the calibration, don't change Peak output levels.

3) You can enable the HSL for the CMS calibration from the overlay of the RGB Widget.

The 75% patterns you see at the LightIllusion site have 100% Saturation and 75% Luminance level.

Usually, when you are using 75% Stimulus level patterns, it may provide better results because the Luminance of 75% Stimulus Patterns has about 50% of the luminance of 100% Stimulus, so your adjustment point is at the center of the color luminance range.

Using 100% Stimulus Patterns, you will have your adjustment point is at the highest end of the luminance range, but you will probably have more significant errors at lower luminance levels.

See there: CMS Calibration With Color Checker
 
Thanks for all the great info :)

After the 2-Point calibration, you can disable Chroma Only and continue with the 10/20-Point calibration.
Just to confirm here - I should be doing a 2-point calibration first (with the Sony's "R/G/B-Gain" and "R/G/B-Bias), then doing the 20-point calibration on top of that? (instead of ignoring 2-point and just using the 20-point controls)
 
Just to confirm here - I should be doing a 2-point calibration first (with the Sony's "R/G/B-Gain" and "R/G/B-Bias), then doing the 20-point calibration on top of that? (instead of ignoring 2-point and just using the 20-point controls)
On Sony, you can use RGB Gain to get close to your desired white point at 100%, then use the 20p for fine tuning. It’s not strictly necessary - skipping RGB Gain adjustments means you will just have to adjust 20p sliders more.

A consideration is that 2p controls apply universally to all picture modes for a given input. And subsequent adjustments to any color temp, on top of that 2p “pre calibration”, will carry across all inputs and modes.

Generally, there’s no reason to use the RGB Bias controls at all, as that can introduce additional issues because it affects only the low range.
 
Discussion starter · #428 ·
ColourSpaceZRO 1.0.0.1254 (20 October 2021)

Release Notes


Fixed ST2084 Sub Space error.

Added Automated Extra Delay function.

Cleaned-up Hardware Options menus.

Removed need for User Access License key.

Download Instructions

Please read about the new download instructions here.

Installation Notes

ColourSpace Installer will completely overwrite any previously installed version, not just update it.

It will not require to un-install older releases before installing a newer version

Extra Delay 'AUTO' function



The Auto button can be used to set the Extra Delay automatically, by first opening a patch window on the display to be profiled, placing the probe on it, and pressing the Auto button.

Extra Delay sets a delay time after the patch has changed before probe measurement starts, and manages signal path delays in the image pipeline.

When using the Auto option, Intelligent Integration & Average Low Light are normally not used.
 
Removed need for User Access License key.
@ConnecTEDDD what is meant by this? What is the CS licensing mechanism now? I’m wondering if it’s easier for me to use CS on both my laptop and HTPC, as before I would have to deactivate and reactivate the license.

EDIT: got a reply from Steve over at the LI forum. This only affected the user access key and had nothing to do with CS licensing. I don’t think I ever had whatever the user access key was.
 
Discussion starter · #430 ·
@ConnecTEDDD what is meant by this? What is the CS licensing mechanism now? I’m wondering if it’s easier for me to use CS on both my laptop and HTPC, as before I would have to deactivate and reactivate the license.

EDIT: got a reply from Steve over at the LI forum. This only affected the user access key and had nothing to do with CS licensing. I don’t think I ever had whatever the user access key was.
Before that release, we were using two codes to activate the software, the first code initially for the IntelliProtect window and then the actual license key to the ColourSpace license window.

The first code will not be required anymore; users will input just their license key.

The purpose of the Activation/De-activation button is to use in case of PC format, Windows upgrade (10 -> 11), or when you buy a new PC, to de-activate from the old.

The number of these de-activations is specific, and data is gathered from your system specs each time you are using these functions, so it's not for changing the license from one PC to another.

You will see after some time that you have reached the limit of activation, and you will have to contact LightIllusion.

LightIllusion will check your license records and see that you are changing the license multiple times between two systems.

When you have two systems, it makes sense to get an additional license for the second system; Ted's calibration disk users have a discount for the ZRO license order.
 
Hi, I wanted to calibrate the Gamma with Autocal and ColourSpaceZRO to my N5 using Spyder5 off screen, instead of SpyderX on screen, with the OMARDRIS patch. Unfortunately when I launch the test to recognize the probe it gives me an error. Is it possible to use the Spyder 5 with the OMARDRIS patch?
Thanks
 
Hi,
to use the JVC-AutoCal with ColourSpace, you first have to download the special adaptation for ColourSpace and copy it into the JVC-AutoCal program directory. Then you can use any sensor supported by ColourSpace and in JVC-AutoCal you just select - depending on the version - the Spyder 5 or the SpyderX.
Please have a look here:

The customizations (DLLs) for ColourSpace can not be used together with my other DLLs, because the DLLs must always have the same name, so that JVC-AutoCal recognizes them. ;)
Peter
 
You may want to take that question to the JVC autocal thread.
Yes, I think that a discussion about "Autocal" happening in a ColourSpace thread, isn't something Light Illusion would be thrilled about. :ROFLMAO:
 
Discussion starter · #436 ·
Yes, I think that a discussion about "Autocal" happening in a ColourSpace thread, isn't something Light Illusion would be thrilled about. :ROFLMAO:
It's possible because ColourSpace has unique 'Remote Control' and 'Secondary Execution' capabilities.
 
So with Autocal 12 for JVC RS1000 I cannot use the Spyder5 probe to measure the Gamma off screen with your Spyder5_as_SpyderX Patch. Is that so? And how could I do?
Are you sure that the original driver for the Spyder5 is installed and not the ArgyllCMS driver? :unsure:
Then neither my DLL nor ColourSpace will find the Spyder5.

You basically have to decide if you want to use ColourSpace or not.
The different JVC AutoCal adjustments cannot be used at the same time.

If you use the JVC AutoCal adjustment for ColourSpace, then it doesn't matter which sensor you connect.
ColourSpace ZRO supports the Spyder5 and also the EODIS3 and then you can do all measurements with ColourSpace.
You only need one solution, not two. If you have ColourSpace ZRO, you don't need another one.
Peter
 
  • Like
Reactions: ConnecTEDDD
<Are you sure that the original driver for the Spyder5 is installed and not the ArgyllCMS driver? > Yes
<Then neither my DLL nor ColourSpace will find the Spyder5> ColourSpace will find the Spyder5

When I do the test with Test_dccmtr_dll.exe a dos window opens and after a few seconds the message appears on the screen: Error connection to localhost: 20102
If I launch ColourSpaceZRO by setting the required parameters and then Autocal12 a window appears indicating that the connection between Autocal and ColourSpace has occurred. In Autocal the Spyder5 probe is recognized. After the calibration procedure is completed regularly, in ColourSpace I have only a series of errors on the various screens except for those relating to CIE xy and CIE uv. On the others only error messages
Thanks
Giuseppe
 
Hi,
The remote service of ColourSpace must be active before the test program is started. Otherwise it cannot find the service.
ColourSpace has only the task to deliver the measured values of the sensor to my DLL. It only gets the order to make a measurement, but beyond that it doesn't get any info about the measurements and therefore can't evaluate any results. Important are only the results that JVC-AutoCal shows.
So ColourSpace is only used to control the sensor.
Peter
 
Thanks Peter, I missed this step (The remote service of ColourSpace must be active before the test program is started). Now the connection happens regularly. Tomorrow I'll try Autocal again with ColouSpaceZRO.
Thanks
Joseph
 
421 - 440 of 1,661 Posts
Top