Pages: [1]   Go Down

Author Topic: Check dE from an input profile  (Read 519 times)

jejes

  • Newbie
  • *
  • Offline Offline
  • Posts: 46
Check dE from an input profile
« on: January 28, 2020, 05:16:50 pm »

Hello, i want to check dE from my camera profile that i have made. I have made 3 profiles with DcamPro, Argyll and Xrite. And i want to check them to see what dE they have. I use Capture One 20. Once i have the profile installed what i should to do.

1. Export a TIF 8 bits with the Input profile embedded. Auto Curve - myprofile.icc  + (embedded)
2. Export a TIF 8 bits with a generic RGB colour space. AutoCurve myrpfile.icc -> sRGB

Wich free tools can i use to check my profiles? I use 3 but i'm not sure
1. profcheck -v2 -k -I r -w cc24.ti3 myprofile.icc
2. dcamprof test-profile cc24.ti3 myprofile.icc
3. https://deltae.picturae.com/ -- Upload JPG with sRGB or myprofile.icc embedded
cc24.ti3=read from my export.tif

Am I right?. Please need some colour genius help.

Sorry for my english it's not my main language.
Logged

Doug Gray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1966
Re: Check dE from an input profile
« Reply #1 on: January 29, 2020, 01:10:38 am »

Hello, i want to check dE from my camera profile that i have made. I have made 3 profiles with DcamPro, Argyll and Xrite. And i want to check them to see what dE they have. I use Capture One 20. Once i have the profile installed what i should to do.

1. Export a TIF 8 bits with the Input profile embedded. Auto Curve - myprofile.icc  + (embedded)
2. Export a TIF 8 bits with a generic RGB colour space. AutoCurve myrpfile.icc -> sRGB

Wich free tools can i use to check my profiles? I use 3 but i'm not sure
1. profcheck -v2 -k -I r -w cc24.ti3 myprofile.icc
2. dcamprof test-profile cc24.ti3 myprofile.icc
3. https://deltae.picturae.com/ -- Upload JPG with sRGB or myprofile.icc embedded
cc24.ti3=read from my export.tif

Am I right?. Please need some colour genius help.

Sorry for my english it's not my main language.


Are you trying to make a scene referred camera profile? These are the only types subject to accuracy evaluation. But these are designed for duplication work (picture of a picture, for instance), and not normally used.

Scene referred profiles are as colorimetric as possible subject to variations from the deviations of the camera's CFA from CIE standard observer functions. See Computing XYZ From Spectral Data at http://www.brucelindbloom.com/

One can get some idea of how well things match by using something like the Colorchecker CG which has a diverse set of colors with different substrates. However, there is no way to perfectly match things. Unlike a printer, colors that appear exactly the same can and do differ in RGB response because the CFAs are not very good matches to human vision. This is called metameric failure. How close you get depends almost entirely on the camera's CFA responses.

That said, scene referred camera profiles don't produce very attractive images. Camera makers, and RAW converters almost always use output-referred profiles in order to produce more pleasing images and these vary a lot in a dE sense even if a camera had perfect CFAs that matched CIE standard observer functions.

For output referred profiles people just try different settings, take pictures, and use what they prefer.
Logged

jejes

  • Newbie
  • *
  • Offline Offline
  • Posts: 46
Re: Check dE from an input profile
« Reply #2 on: January 29, 2020, 10:38:08 am »

My goal is to check how good is my profile. I want to know dE of my profile. But i don't know in wich colour space do i have to export the Colourchecker photo from C1 and if i have to upload for example here  https://deltae.picturae.com/  in sRGB, adobeRGB.

I want to know that info, i know that with ColourThinkPro you can do it, but i don't have this program.

Thank you
Logged

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 16005
    • http://www.digitaldog.net/
Re: Check dE from an input profile
« Reply #3 on: January 29, 2020, 10:47:39 am »

Nearly impossible to do what you ask without tools like CTP.
To give you an example of how this is done when you have a true reference and output you measure (as with printer profiles) the steps are:
We'll use a target we call 216pixel_RGBs.tif as an example, but you can use the iStar or whatever patch set you wish:
So there are a number of ways to tackle this depending on what you're looking to judge.
One way is to simply compare the reference* values that created the target to the measured values of the target run through the printer. But that only provides an overall accuracy report, not specifics. But the idea is, you've got an RGB value that's a reference and an RGB value from the measurement of the target itself. Conduct a dE report in ColorThink.
But here's a way to test the differing tables (it's kind of complicated). We'll use a target we call 216pixel_RGBs.tif as an example, use the iStar or whatever patch set you wish where 1 pixel is one color patch:
1. Open the “216pixel_RGBs.tif file in Photoshop. This file needs to be kept exactly as it is (individual pixels). Duplicate the file. Size the image using Nearest Neighbor to an appropriate print size. Call this “Reference Print File”. You'll print this out and measure.
2. Print this file out to the profiled printer. (You're sending the RGBs in the file to the printer.)
3. Let the print stabilize, and then measure it and save the data as “Reference Labs.txt”. The 216 RGB.TXT file (a reference file that built this target) needs to be installed for MeasureTool or PatchTool or whatever software you'll use to access access the data. This data measured is the bottom line -- what happens when you send these RGBs to this device.
4. Now you need to create predicted Lab files from the RGB. You need two versions -- one in pixels to compare in ColorThink, one upsampled to print -- for each profile.
5. Take the RGB pixel file and Assign the printer profile. Convert Abscol to Lab. Save the results with the profiling package somehow identified in the filename (eg “i1P 288 Luster/Predicted”).
I suggest saving the pixels versions as ProfilePackageName Predicted Labs.
6. Take the upsampled RGB file. Assign the printer profile. Convert Abscol to Lab. Convert from LAB to printer profile being tested using absocl and print.
7.Let the print stabilize and measure the results. Save the measurements as ProfilePackageName Roundtrip Labs (“i1p 288 Luster Round Trip”).
You now have three things to compare for each package:
A. The Reference File (“Reference Labs.txt”). The Predicted Labs (“i1P 288 Luster”). The Roundtrip Labs (“i1P 288 Luster Round Trip”).
B. Reference-to-Predicted (“Reference Labs.txt”/“i1P 288 Luster”) shows you the accuracy of the AtoB Colorimetric table.
C. Predicted-to-Roundtrip (“i1P 288 Luster”/“i1P 288 Luster Round Trip”) shows you the accuracy of the BtoA Colorimetric table.
Reference-to-Roundtrip (“Reference Labs.txt”/“i1P 288 Luster Round Trip”) gives you a decent measure of the overall profile accuracy.
You'll typically see that the AtoB and BtoA errors tend to cancel each other somewhat because they go in opposite Directions-Reference-to-Roundtrip shows that nicely.
*You’d need a reference of the scene (target) + Illuminant which ain’t easy hence my first sentence including ‘nearly’.
Logged
Andrew Rodney
Author “Color Management for Photographers"

Doug Gray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1966
Re: Check dE from an input profile
« Reply #4 on: January 29, 2020, 11:51:26 am »

My goal is to check how good is my profile. I want to know dE of my profile. But i don't know in wich colour space do i have to export the Colourchecker photo from C1 and if i have to upload for example here  https://deltae.picturae.com/  in sRGB, adobeRGB.

Read the Wiki at that site. That site is designed to provide dE info on reproduction type profiles. These are also called scene referred profiles. Scene referred profiles are for things like scanners or cameras that are trying to duplicate a painting, document, or physical photograph. They are pretty awful for normal photography.

Here's a description of scene referred v output referred profiles at the ICC site. There's a wide range of different color/tone shifts that various output profiles produce to account for dynamic range as well as to simulate the look of films back in the day when film processing was done. Basically, dE applies to scene referred profiles but not output referred ones.

http://color.org/ICC_white_paper_17_ICC_profiles_with_camera_images.pdf

Logged

DP

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 625
Re: Check dE from an input profile
« Reply #5 on: January 30, 2020, 12:03:41 am »

Read the Wiki at that site. That site is designed to provide dE info on reproduction type profiles. These are also called scene referred profiles. Scene referred profiles are for things like scanners or cameras that are trying to duplicate a painting, document, or physical photograph. They are pretty awful for normal photography.

Here's a description of scene referred v output referred profiles at the ICC site. There's a wide range of different color/tone shifts that various output profiles produce to account for dynamic range as well as to simulate the look of films back in the day when film processing was done. Basically, dE applies to scene referred profiles but not output referred ones.

http://color.org/ICC_white_paper_17_ICC_profiles_with_camera_images.pdf

dear, dear

all "camera profiles" used by raw converters are scene referred - as they deal with raw files where the data are the numbers that are not coordinates in any formal color space and they ("camera profiles") guide the raw converter code how to convert that data into the proper coordinates in some formal colors space - that is how "color" appears... now those profiles might be reproduction or not (plus the way you actually use a raw converter - I mean adjustments you select in its UI - also plays a role - profile is not working alone) - but they are all scene referred
« Last Edit: January 30, 2020, 12:08:47 am by DP »
Logged

Doug Gray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1966
Re: Check dE from an input profile
« Reply #6 on: January 30, 2020, 02:38:56 am »

dear, dear

all "camera profiles" used by raw converters are scene referred - as they deal with raw files where the data are the numbers that are not coordinates in any formal color space and they ("camera profiles") guide the raw converter code how to convert that data into the proper coordinates in some formal colors space - that is how "color" appears... now those profiles might be reproduction or not (plus the way you actually use a raw converter - I mean adjustments you select in its UI - also plays a role - profile is not working alone) - but they are all scene referred

Yeah, that's true for the C1 ICC profile front end prior to sending on the image data for output rendering. My mistake. I'm too used to the Adobe Raw converter which is quite difficult to get scene referred final images from.

I stand corrected.
Logged

jejes

  • Newbie
  • *
  • Offline Offline
  • Posts: 46
Re: Check dE from an input profile
« Reply #7 on: January 30, 2020, 09:21:58 am »

Please, let's suppose then that i want to know dE00 of my  reproduction profile/scene referred from C1. How I can do it?

This is my "supposed" Workflow

1. Shoot ColourChecker. Inside C1 RGB D4 Gray 121 121 121
2. Assign My-Profile.icc and Curve Auto
3. Export to Colourspace (wich colourspace? sRGB AdobeRGB). With sRGB i get usually a dE00 <4. With AdobeRGB i get dE00 >12.
4. Measure with wich tools (A. profcheck -v2 -k -I r -w cc24.ti3 myprofile.icc B. dcamprof test-profile cc24.ti3 myprofile.icc  C. https://deltae.picturae.com/)

Many thanks for your help and your comentars
Logged

Doug Gray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1966
Re: Check dE from an input profile
« Reply #8 on: January 30, 2020, 11:35:33 am »

Use linear curve, not auto curve.
Logged

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 16005
    • http://www.digitaldog.net/
Re: Check dE from an input profile
« Reply #9 on: January 30, 2020, 12:49:23 pm »

Not all "camera profiles" used by raw converters are scene referred but should be  ;). Some folks do (still  :-X) build output referred profiles from rendered raw. The Adobe paper is a bit dated but in that  respect still pertinent.
But this doesn’t seem to help the OP. Examine his last question about assigning a working space which would be on output referred data. What I think he needs, again depending on the converter, how the profiles are built etc, is a reference and then the results to produce a dE report. But for that to be ‘useful’ in terms of accuracy, indeed the data (and profile) must be scene referred. So again, this goes back to the nearly impossible comment depending again on a lot of the OP’s workflow constraints or lack thereof. There’s a target to build the profile, there’s an illuminant used to capture it. Is there Lab values available of this data? What then is used with the profile upon the raw data to produce what? A rendered image in what color space that can be compared (Lab in vs. Lab out)?
I’d submit this is a nasty and mostly unnecessary rabbit hole. I’d submit that raw converters produce rendered color based on the desires of the person rendering it; subjective. I’d submit accuracy isn’t useful expect in very, very rare conditions (I want to colorimetrically match the original scene and all such scenes). I’d submit outside of this need, build a suite of profiles, apply them to a group of representative images and pick the one who’s rendering without any other adjustments appear ‘best’ and move on.
Logged
Andrew Rodney
Author “Color Management for Photographers"

GWGill

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 554
  • Author of ArgyllCMS & ArgyllPRO ColorMeter
    • ArgyllCMS
Re: Check dE from an input profile
« Reply #10 on: January 30, 2020, 05:58:53 pm »

Not all "camera profiles" used by raw converters are scene referred but should be  ;).
Really ?

I thought the whole purpose of DNG profiles was to cook the camera values into "pleasing" output (just like film is expected to do), not to render accurate scene referred color values.
Logged

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 16005
    • http://www.digitaldog.net/
Re: Check dE from an input profile
« Reply #11 on: January 30, 2020, 06:01:40 pm »

Really ?
I thought the whole purpose of DNG profiles was to cook the camera values into "pleasing" output (just like film is expected to do), not to render accurate scene referred color values.
They are indeed. They are used with the various rendering controls in the product to produce pleasing colors. And I'd not want to be forced into working otherwise.
Logged
Andrew Rodney
Author “Color Management for Photographers"

DP

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 625
Re: Check dE from an input profile
« Reply #12 on: January 30, 2020, 07:48:31 pm »

Yeah, that's true for the C1 ICC profile front end prior to sending on the image data for output rendering. My mistake. I'm too used to the Adobe Raw converter which is quite difficult to get scene referred final images from.

I stand corrected.

DCP profiles for ACR/LR are also scene ref'd in the same sense ...
Logged

DP

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 625
Re: Check dE from an input profile
« Reply #13 on: January 30, 2020, 07:50:46 pm »

Use linear curve, not auto curve.

use "linear scientific" curve ... regular "linear" .fcrv files have some flag inside that tell C1 code to use non linear operations near clipping... it might matter - so better to be safe than sorry
Logged

DP

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 625
Re: Check dE from an input profile
« Reply #14 on: January 30, 2020, 08:01:08 pm »

;). Some folks do (still  :-X) build output referred profiles from rendered raw.

you intend to argue that using .tiff output from C1 to prepare camera profile based on how P1 direct makes it not scene-referred ? I argue they are still scene ref'd because when said camera profile in ICC container is applied by C1 it still deals with the data that is not in some proper color space (even by that time the data represens something that was already demosaicked, WB'd and hell lot of various linear and non linear operations performed following what you do in C1 UI - but no COLOR TRANSFORM happened yet) - hence it can't be output ref'd

 
Logged

DP

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 625
Re: Check dE from an input profile
« Reply #15 on: January 30, 2020, 08:06:25 pm »

Really ?

I thought the whole purpose of DNG profiles was to cook the camera values into "pleasing" output (just like film is expected to do), not to render accurate scene referred color values.

the whole  purpose of DNG (DCP) profiles was to try to have something more suitable for a practical raw conversion pipeline as Adobe imagined that ... no camera profile can render anything accurate by itself as any camera profile does not know what the actual illumination was for a given raw file w/o raw converter telling it (for example by performing WB - per channel exposure correction) before camera profile is applied
Logged

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 16005
    • http://www.digitaldog.net/
Re: Check dE from an input profile
« Reply #16 on: January 30, 2020, 08:11:34 pm »

you intend to argue that using .tiff output from C1 to prepare camera profile based on how P1 direct makes it not scene-referred ?
I don’t intend to argue. I intend we speak clearly and accuracy. We’re talking about color accuracy so let’s be accurate in our language too. Not all camera profiles used by raw converters are scene referred. The statement is more accuracy written: “Raw converters that can use scene referred profiles use scene referred profiles“. There are all kinds of workflows historically and today where ICC camera profiles are made from output referred data.
You’re now specifically speaking of C1 and TIFFs yet you spoke of all raw converters.
What C1 may or may not output may or may not be scene referred but that doesn’t speak of all raw converters either. And it’s useful to define scene referred and then be certain that what’s being output from ANY raw converter IS scene referred. That is IF you want to speak accuracy here.  :o
A scene is defined in ISO 22028-1 as follows:
scene
spectral radiances of a view of the natural world as measured from a specified vantage point in space and at a specified time
A scene may correspond to an actual view of the natural world or to a computer-generated virtual scene simulating such a view.
A scene-referred image is an image where the image data is an encoding of the colors of a scene (relative to each other), as opposed to a picture of a scene. In a picture, the colors are typically altered to make them more pleasing to viewers when viewed using some target medium.

Further it’s useful to consider:
3.36
scene-referred image state
image state associated with image data that represents estimates of the colour-space coordinates of the elements of a scene
Note 1 to entry: When the phrase “scene-referred” is used as a qualifier to an object, it implies that the object is in a scene-referred image state. For example, scene-referred image data are image data in a scene-referred image state.
Note 2 to entry: Scene-referred image data can be determined from raw DSC image data before colour-rendering is performed. Generally, DSCs do not write scene-referred image data in image files, but some may do so in a special mode intended for this purpose. Typically, DSCs write standard output-referred image data where colour-rendering has already been performed.
Note 3 to entry: Scene-referred image data typically represent relative scene colourimetry estimates. Absolute scene colourimetry estimates may be calculated using a scaling factor. The scaling factor can be derived from additional information such as the image OECF, FNumber or ApertureValue, and ExposureTime or ShutterSpeedValue tags.
Note 4 to entry: Scene-referred image data may contain inaccuracies due to the dynamic range limitations of the capture device, noise from various sources, quantization, optical blurring and flare that are not corrected for, and colour analysis errors due to capture device metamerism. In some cases, these sources of inaccuracy can be significant.
Note 5 to entry: The transformation from raw DSC image data to scene-referred image data depends on the relative adopted whites selected for the scene and the colour space used to encode the image data. If the chosen scene adopted white is inappropriate, additional errors will be introduced into the scene-referred image data. These errors may be correctable if the transformation used to produce the scene-referred image data are known and the colour encoding used for the incorrect scene-referred image data has adequate precision and dynamic range.
Note 6 to entry: The scene may correspond to an actual view of the natural world or may be a computer-generated virtual scene simulating such a view. It may also correspond to a modified scene determined by applying modifications to an original scene to produce some different desired scene. Any such scene modifications should leave the image in a scene-referred image state and should be done in the context of an expected colour-rendering transform.


Now, if you’d like to accuracy explain to the OP exactly how to produce an accuracy metric from his profiles, the actual question, by all means try to do so accurately.  ;) 
Logged
Andrew Rodney
Author “Color Management for Photographers"

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 16005
    • http://www.digitaldog.net/
Re: Check dE from an input profile
« Reply #17 on: January 30, 2020, 08:14:53 pm »

... no camera profile can render anything accurate by itself as any camera profile does not know what the actual illumination was for a given raw file w/o raw converter telling it (for example by performing WB - per channel exposure correction) before camera profile is applied
So now we are back on topic for the OP and we’re back to my points that, this is a big rabbit hole that’s also largely unnecessary to dig into for all? (no) but nearly most users.
Logged
Andrew Rodney
Author “Color Management for Photographers"

Doug Gray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1966
Re: Check dE from an input profile
« Reply #18 on: January 30, 2020, 08:53:06 pm »

So now we are back on topic for the OP and we’re back to my points that, this is a big rabbit hole that’s also largely unnecessary to dig into for all? (no) but nearly most users.

Yeah, a large rabbit hole.  Sticking with scene referred profiles, even with a perfect D50 illuminant (good luck with that) and a profile made using said illuminant, things being photographed have diverse spectral reflectances and the target used to create a profile only has a subset of these. To the degree that the camera's CFA is not a linear transform of color matching functions variations in dE will occur. And some objects that are metamers but with differing spectral reflectances will still produce different responses hence an image with different colors. They are difficult to predict or even quantify absent spectral response curves of the camera sensor, CFA filters, and spectral reflectance of objects being photographed.

I recall some academic studies where captures were taken with a camera using multiple additional filters. The combined images were then processed to reconstruct a significantly closer match to Luther-Ives.

But all of this is mostly of academic interest and perhaps a few museums interested in archival processes. Rabbit hole it is.
Logged

digitaldog

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 16005
    • http://www.digitaldog.net/
Re: Check dE from an input profile
« Reply #19 on: January 30, 2020, 08:57:12 pm »

With respect to this rabbit hole, and along with the idea of simply taking images and viewing them with a group of profiles instead, let me add:
There seems to be some perverse human characteristic that likes to make easy things difficult.” -Warren Buffett
Perhaps Mr. DP can provide a step by step procedure for the OP, with consideration he lacks tools like CTP available instead.
I’ve got other things to do now....  ::)
Logged
Andrew Rodney
Author “Color Management for Photographers"
Pages: [1]   Go Up