Pages: 1 ... 5 6 [7] 8   Go Down

Author Topic: What are the essential adjustments that SHOULD be done in the raw processor?  (Read 39395 times)

AlterEgo

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1995

better

standard question = define "better" ?
Logged

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto

Thanks!

Erik

Yes, you created a new process recipe and for the ICC Profile you select Embed Camera Profile.

Cheers

Robert
Logged
Erik Kaffehr
 

AlterEgo

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1995

You can't use the Color Editor with those profiles, though ...
you can translate matrix + trc profile into LUT based though where the LUT will be doing the same color transform and then you can use color editor...

I am totally not Matlab expert, but something like this :

    Pmat = iccread('SonyA7RM2-CC24(M+TRC^TF).icm'); % profile after rawdigger + makeinputicc/argyll
    Plut = iccread('xxx.icc'); % using a template, some OEM profile from C1 distribution, to fill with our data and the rest will be from OEM profile - put the proper name instead of xxx.icc

    Plut.AToB0.InputTables = {Pmat.MatTRC.RedTRC, Pmat.MatTRC.GreenTRC, Pmat.MatTRC.BlueTRC};
    Plut.AToB0.OutputTables = {[0 65535], [0 65535], [0 65535]}; % linear curve
    Plut.AToB0.PreMatrix = [[1,0,0];[0,1,0];[0,0,1];[0,0,0]]'; % last column = 0,0,0

    m_rgb2xyz = [ Pmat.MatTRC.BlueColorant; Pmat.MatTRC.GreenColorant;  Pmat.MatTRC.RedColorant ]; % 3x3 in this order = BlueColorant is the top row, RedColorant is the bottom row

    n_3Dlut_Size = 33; % C1 uses something like 33x33x33 LUT
   
    m_3Dlut_Points = linspace(0, 32767, n_3Dlut_Size); % get even spacing for points from 0 to __32767__ (not 65535!) for cieXYZ luts

    for i = 1:n_3Dlut_Size
        for j = 1:n_3Dlut_Size
            for k = 1:n_3Dlut_Size
                m_RGB2XYZ_3Dlut(i,j,k,:) = round( [ m_3Dlut_Points(i), m_3Dlut_Points(j), m_3Dlut_Points(k) ] * m_rgb2xyz );
            end;
        end;
    end;

    Plut.AToB0.CLUT = m_RGB2XYZ_3Dlut; % replace 3Dlut converting RGB to cieLAB with 3Dlut converting RGB to cieXYZ
    Plut.Header.ConnectionSpace = 'XYZ';
    Plut.MediaWhitePoint = Pmat.MediaWhitePoint;
    Plut.MediaBlackPoint = Pmat.MediaBlackPoint;

    iccwrite(Plut, 'SonyA7RM2-CC24(3Dlut^XYZ).icm');
« Last Edit: March 23, 2016, 05:37:37 pm by AlterEgo »
Logged

AlterEgo

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1995

as long as the profile class is tagged correctly in the profile header ("scnr").
no, you don't need to have "scnr" for matrix input profiles (camera profiles)...
Logged

tho_mas

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1799

Quote
as long as the profile class is tagged correctly in the profile header ("scnr").
no, you don't need to have "scnr" for matrix input profiles (camera profiles)...
Long ago I've played around with matrix profiles (monitor profiles) and one way to make C1 read the profiles as "input" profiles was simply to rename "mntr" to "sncr" in the profile header. Maybe there is another solution making C1 recognize certain profiles as input profiles, I don't know ... but on my system (Mac) only "sncr" profiles are recognized by C1 as input profiles.
« Last Edit: March 23, 2016, 07:42:59 pm by tho_mas »
Logged

tho_mas

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1799
What are the essential adjustments that SHOULD be done in the raw processor?
« Reply #125 on: March 23, 2016, 07:42:04 pm »

Here is a raw file from P45+ back. In all conversions I make yellows in the blade go to the extreme limit of the RGB used:
http://echophoto.dnsalias.net/ekr/Articles/ColourSpaces/20150502-CF046583.iiq

Would be interesting what you find.
I don't understand what you actually did ... and also can't really decode the graph of color think.
Looks like you've massively boosted saturation of the greens in the Advanced Color Editor? This would explain the confusing shape of the (edited!) P45+ Flash profile.

As far as the actual (unedited) P45+ Flash profile goes attached you'll see a comparison to ProPhoto-RGB (top view and bottom view) which shows the areas where the P45+ Flash profile exceeds ProPhoto (bright blues and violets and dark, high saturated yellows and greens). ProPhoto is the colored profile, of course...

« Last Edit: March 23, 2016, 07:57:20 pm by tho_mas »
Logged

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto

Hi,

I just opened the file in C9 and exported it with embedded profile. No manipulation at all, at least as far as I know.

After that I opened the file in ColorThink 2.3 and exported the embedded profile, so I could read it into ColourThink. The image shows the following:

- Colour patches from the TIFF files
- The embedded ICC file in the TIFF image
- ProPhoto GRB

I have posted the raw image, and it can be worth a look. As far as I can see it is very close to the gamut limits and violently outside Adobe RGB. That applies to both LR 6.5 and C9.0.2.

ColorThink is by the way the tool that Andrew (the Digital Dog) uses. But this is the light version.

Best regards
Erik

I don't understand what you actually did ... and also can't really decode the graph of color think.
Looks like you've massively boosted saturation of the greens in the Advanced Color Editor? This would explain the confusing shape of the (edited!) P45+ Flash profile.

As far as the actual (unedited) P45+ Flash profile goes attached you'll see a comparison to ProPhoto-RGB (top view and bottom view) which shows the areas where the P45+ Flash profile exceeds ProPhoto (bright blues and violets and dark, high saturated yellows and greens). ProPhoto is the colored profile, of course...
« Last Edit: March 23, 2016, 11:48:31 pm by ErikKaffehr »
Logged
Erik Kaffehr
 

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto

Hi,

This is what I see in ColorSync. The first plot shows the Capture One P45 profile in /Library/ColorSync/Profiles/Phase One P 45 Flash.icm

The other one is the "camera profile" I have exported from Capture 9.0.2.

The wireframe volumes show the two Phase One Capture One profiles, while the translucent volume is Prophoto RGB. In this case I would say Prophoto RGB fully encloses the C1 profile for the P45/P45+.

I don't think colour spaces matter that much on the input side, as long that the input volume fully contains "Pointer's gamut". After all, a colour volume defines the RGB primaries and the white point and there are perfectly good mathematical conversions from one set of primaries to another set of primaries.

With output profiles the situation is different. In input profiles the XYZ values are really integrals of reflected colour multiplied with the spectral response curve for each channel, this is really what metamerism is about. With displays we have a set of well defined, narrow spectrum, primaries corresponding to "phosphors" or colour filters.

Best regards
Erik



I don't understand what you actually did ... and also can't really decode the graph of color think.
Looks like you've massively boosted saturation of the greens in the Advanced Color Editor? This would explain the confusing shape of the (edited!) P45+ Flash profile.

As far as the actual (unedited) P45+ Flash profile goes attached you'll see a comparison to ProPhoto-RGB (top view and bottom view) which shows the areas where the P45+ Flash profile exceeds ProPhoto (bright blues and violets and dark, high saturated yellows and greens). ProPhoto is the colored profile, of course...
Logged
Erik Kaffehr
 

tho_mas

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1799

I don't think colour spaces matter that much on the input side, as long that the input volume fully contains "Pointer's gamut". After all, a colour volume defines the RGB primaries and the white point and there are perfectly good mathematical conversions from one set of primaries to another set of primaries.
With 3D LUT profiles it's pretty different than just 3 primaries and a white point. Table based you can create profiles with - for instance - relatively high saturation in the midtones without the need to define too high saturated primaries. Table based you are not limited to the "triangle" - design of matrix profiles...
Logged

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto

Are you not mixing up gamut volumes, input profiles and output profiles?

Best regards
Erik

With 3D LUT profiles it's pretty different than just 3 primaries and a white point. Table based you can create profiles with - for instance - relatively high saturation in the midtones without the need to define too high saturated primaries. Table based you are not limited to the "triangle" - design of matrix profiles...
Logged
Erik Kaffehr
 

tho_mas

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1799

Are you not mixing up gamut volumes, input profiles and output profiles?

Best regards
Erik
not that I know of...
Logged

Robert Ardill

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 658
    • Images of Ireland

For anyone who has i1Publish and is interested in camera profiling, I got this information from X-Rite (it's not an officially supported solution, but they say it works):

"Creating Camera Profiles for Capture One

Phase One’s Capture One software is rightly very widely used by photographers for processing their images. Its colour management has always been excellent and the built-in camera profiles are accurate enough for many types of photography. However, I have had a few customers in the fashion and fine art reproduction areas who needed to improve upon the standard level of accuracy and get results closer to the original garment or art work. Capture One’s own Color Editor module allows you to edit the supplied profiles and save them under new names but this tutorial will take you through the steps you need to create a completely new custom camera profile for your camera and lighting set up.

Camera Profiling Requirements

In order to create camera profiles you need a special test target and profile creation software. I used a X-Rite ColorChecker Passport and X-Rite i1 Profiler software. You could also use the larger ColorChecker or other target and there are plenty of alternative software packages. The ColorChecker Passport does come with its own profiling software but the profiles it creates are only compatible with Adobe Photoshop and Lightroom. Whichever target you use you need to photograph it under the lighting that you intend to use for the rest of the shoot. It needs to be lit evenly and exposed correctly. It also makes it easier if it is reasonably large in the frame and straight.

Processing the Raw File

Camera profiling applications generally require 8-bit TIFF files and so you will need to process the image through Capture One before you create the camera profile. However, this needs to be done in a very specific way if you want to get a good profile. Flick to the Color tab and under ICC Profile select Show All at the bottom of the drop down menu. Then go back into the ICC Profile menu and choose Effects/No color correction. This disables the default profiles. You will also need to set the correct Curve. Choosing the camera Linear curve usually gives the best results. Don’t worry if the image looks awful. That shows you how much work the camera profile does in creating a decent image.

Click on the Output tab and start by choosing TIFF 8-Bit Full Size (Adobe RGB) as the Process Recipe. But then change the ICC Profile to Embed camera profile. You can now set the destination and process the file.

Creating Camera Profiles for Capture One
The process of creating the camera profile in i1 Profiler is pretty quick and easy. You select Scanner Profiling and select your target type, in the case X-Rite ColorChecker 24. You can then drag and drop the Tiff file you processed out of Capture One. If the exposure, cropping etc is good it should automatically find the 24 colour patches in the image. If it doesn’t you can manually set the target cropping. It will warn you if it thinks the exposure is out of the acceptable range. Then click Next. The Reference file for the ColorChecker target is included in the i1 Profiler software so you can just click Next again. Give the profile a sensible name; include camera model, lighting, location and anything else that will help you identify exactly what situation the profile was created for. Choose ICC Profile Version 2 and either User level (Mac) or System level (Windows) and click Create and save profile. i1 Profiler will then create the profile and give you a report on the profile. It lists a series of delta E values. Delta E is a measurement of colour difference. The lower the numbers are the better the profile. You can experiment with different exposures or film curves to get you lower numbers. In my test I got an average delta E (CIE 1976) of 0.03 with 100% of the patches in gamut.

Applying a Custom Profile in Capture One

Capture One reads which profiles are on your system when it launches. So if you want it to see the new profile you will have to Quit and then start the application again. In the Color tab you should be able to find your new profile under ICC Profile/Other. The supplied ICC profiles tend to made to give pleasing results as are the curves other than linear so you may see the image is a bit flatter than you were used to but you can now adjust the image using the Exposure tab as normal. Camera profiles are about getting the colours right, not necessarily the exposure. Any adjustments you make and the profile can be applied to the rest of the images in the series by using Copy and Apply adjustments commands.

Creating camera profiles for Capture One, or any other software, can mean a little experimentation and testing of different settings. Some camera profiling software offers choices for how the profile is made and what is prioritised – accuracy or a more “pleasing” result. I suggest that you put some test shots through the built-in profiles and then compare the results to the custom profiles. Don’t forget when you process shots out of Capture One you will want to choose a destination colour space, such as Adobe RGB, sRGB or ProPhoto. Once you get a good camera profile you may be very pleased how much more colour accurate the images are

"

Cheers

Robert
Logged
Those who cannot remember the past are condemned to repeat it. - George Santayana

AlterEgo

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1995

For anyone who has i1Publish and is interested in camera profiling
I think better to procure the old ProfileMaker (v5.x.x) though, rather then the current software X-Rite has
Logged

Robert Ardill

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 658
    • Images of Ireland

I think better to procure the old ProfileMaker (v5.x.x) though, rather then the current software X-Rite has

Why is that?  Are there features in it that are not in i1Publish?  Or does it give better quality profiles?  I have to say that I've been very happy with the printer profiles made with i1Publish.

There is also the option (free) of ArgyllCMS if you don't mind a command-line tool.

Robert
Logged
Those who cannot remember the past are condemned to repeat it. - George Santayana

Robert Ardill

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 658
    • Images of Ireland

standard question = define "better" ?

Well I'm not talking about 'more pleasing' because that is something that I want to control myself by making the adjustments that I want.

I am talking about maximizing the quality of the image, in other words to retain as much of the image information as possible with as little distortion as possible.  After all, there isn't much point in buying expensive lenses if all of the potential richness is thrown away in raw processing.

And I am also therefore talking about color fidelity.  Without getting bogged down in all of the techno-speak, what I mean by that is that if I take a photo of a scene and I have a well-calibrated output device, I want the image of the scene to be as close as possible to the scene.  So I want the blue of the sky to be the same blue, the green of the leaves to be the same green etc.,  all within reason and the limits of the technology,of course.

I know that this is a lot to ask and I'm sure it isn't easy to achieve (after all, matching print to monitor isn't easy either), but it does seem to me to be a worthwhile objective.

Cheers

Robert
Logged
Those who cannot remember the past are condemned to repeat it. - George Santayana

AlterEgo

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1995

Why is that? 
is "click Create" is the only option for tuning the profiles for cameras in i1Publish ?
Logged

AlterEgo

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1995

I am talking about maximizing the quality of the image, in other words to retain as much of the image information as possible with as little distortion as possible.

And I am also therefore talking about color fidelity.  Without getting bogged down in all of the techno-speak, what I mean by that is that if I take a photo of a scene and I have a well-calibrated output device, I want the image of the scene to be as close as possible to the scene.  So I want the blue of the sky to be the same blue, the green of the leaves to be the same green etc.,  all within reason and the limits of the technology,of course.

so you want profiles that are essentially reproduction profiles and the whole conversion to be reproduction work like... very good - do you haul spectrophotometer (or better yet - lab grade spectroradiometer) with you when you are shooting ? you want to record the spectral data  ;D of your leaves then... people do that
Logged

tho_mas

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1799
Logged

Robert Ardill

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 658
    • Images of Ireland

so you want profiles that are essentially reproduction profiles and the whole conversion to be reproduction work like... very good - do you haul spectrophotometer (or better yet - lab grade spectroradiometer) with you when you are shooting ? you want to record the spectral data  ;D of your leaves then... people do that

People also photograph a colorchecker or similar target that gives them a reference.

Why would you think that the aim of color fidelity is stupid?  We may not achieve it, but we may go quite a long way towards it.  White-balance for example gets us quite a long way in the right direction ... is this something you think is a waste of time?
Logged
Those who cannot remember the past are condemned to repeat it. - George Santayana

Robert Ardill

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 658
    • Images of Ireland

http://www.basiccolor.de/basiccolor-input-en/

Thanks ... have you tried it and if so how successful have you been at creating high quality camera profiles?

Cheers

Robert
Logged
Those who cannot remember the past are condemned to repeat it. - George Santayana
Pages: 1 ... 5 6 [7] 8   Go Up