10 bit RAW converters and editors

[digitaldog]

Now I learn to my horror that something on output side has to 'know' what the input side was.

It doesn't have to know as experienced from years of the current implementation, but it would be useful. Think about the same image in sRGB and ProPhoto RGB being treated the same and possibly without either of those spaces 'assumed' as the source data. V4 was supposed to clear that up to some respect.

My V4 profile has tables for both Perceptual and Colorimetric, four tables in total. RawTherapee allows the output intent to be set as Perceptual. However, as you have helped me realize, setting it doesn't necessarily mean you get it 

IF I understand you (I don't have RT), it's doing what Photoshop has done since day one; provide RI options when only one table is present. IOW, you can ask to convert to Adobe RGB (1998) and pick Perceptual in the UI, you'll get RelCol. Kind of bugs me.

In answer to your questions - RT_sRGB looks 'better' than V4 out of the box on my screen. I don't print anything.

Then use that.

[Jim Kasson]

IOW, you can ask to convert to Adobe RGB (1998) and pick Perceptual in the UI, you'll get RelCol. Kind of bugs me.

And, last time I looked, when you convert between two RGB spaces with different white points and picked absolute colorimetric, Ps ignores your choice and changes the white point but the chromaticity of colors on the gray axis of the old space is now the chromaticity of the white point of the new space. That always kind of bugged me, but I see why they did it; to cut down on support calls.

Jim

[digitaldog]

And, last time I looked, when you convert between two RGB spaces with different white points and picked absolute colorimetric, Ps ignores your choice and changes the white point without affecting the chromaticity of the colors on the gray axis. That always kind of bugged me, but I see why they did it; to cut down on support calls.

Could be, I'd never use that RI in that context (maybe the Adobe folks are protecting us?).

[Jim Kasson]

Could be, I'd never use that RI in that context (maybe the Adobe folks are protecting us?).

Do what I mean?

[Lundberg02]

Raw Therapee, like Photoshop, allows you to set a rendering intent. RT allows setting it globally. As apparent from this discussion, that does not mean you actually get anything but Relative.
 I can print to my six color Epson from a Pro Photo RAW conversion through Photoshop manages colors and I actually get different results for each intent. I believe that means that the Epson profiles have all four intents. It makes sense that a printer would need to have them.
But I reiterate that intents are a farce.  Color management is still not finished. You can't do 10 bit on a Mac. I can't set the LUT in my new Dell without buying an i1 Display Pro. ColorSync adjusts all gammas to God knows what, so it doesn't make any difference what the image profile has.
I do what any sane person would do, I get the image the way I want it, and beat on things until the inkjet looks the same. There is no magic button.

[digitaldog]

I believe that means that the Epson profiles have all four intents.

Correct.

[mouse]

RawTherapee comes bundled with a number of custom-made high quality
output profiles:

I think the meaning of "output profile" has been well clarified by dd.

I use RT and here is my concept of what selection of an output profile does in that software.
Its effect is the same as using "convert to profile" in Photoshop and then saving the file as tagged with that profile.  In which case the profile serves as an input profile for whatever (color managed) software opens that file.

Please correct me if I have got this wrong.

[Lundberg02]

But can you get a rendering intent other than relative colorimetric? I doubt it.

[mouse]

But can you get a rendering intent other than relative colorimetric? I doubt it.

If you are asking me:  As you noted above RT allows one to set a rendering intent.  But I cannot see, on my monitor, any difference between the options.  But perhaps in a print a difference can be seen.

[xpatUSA]

But can you get a rendering intent other than relative colorimetric? I doubt it.

Your doubt is unfounded. You can 'get' a rendering intent other than relative colorimetric if you have an appropriate output profile installed. I have one such, of type 'mluc', installed in RT. Here is a comparison of two output gamuts:



Should you wish to play, you can download the profile from color.org - I used sRGB_v4_ICC_preference_displayclass.icc

cheers

[Lundberg02]

Yes. "If you have one installed". But no one does except you.  The printer will probably have the intents, so you can do it that way if you need to.

[xpatUSA]

Yes. "If you have one installed". But no one does except you.  The printer will probably have the intents, so you can do it that way if you need to.

I fold.

Good luck.

[mouse]

Your doubt is unfounded. You can 'get' a rendering intent other than relative colorimetric if you have an appropriate output profile installed. I have one such, of type 'mluc', installed in RT. Here is a comparison of two output gamuts:



Should you wish to play, you can download the profile from color.org - I used sRGB_v4_ICC_preference_displayclass.icc

cheers

Forgive my ignorance, but on this lovely comparison I see only ONE (output) gamut presented; presumably that of sRGB.  The dots, I assume, represent the colors in your sunflower image when saved with two different profiles.  With one (green) all colors are fully contained within the sRGB gamut; and with the other (red) barely or not completely contained within that gamut.

Can you unfold long enough to clarify?   Thanks.

[xpatUSA]

Forgive my ignorance, but on this lovely comparison I see only ONE (output) gamut presented; presumably that of sRGB.  The dots, I assume, represent the colors in your sunflower image when saved with two different profiles.  With one (green) all colors are fully contained within the sRGB gamut; and with the other (red) barely or not completely contained within that gamut.

Can you unfold long enough to clarify?   Thanks.

It's just poor terminology, not that unusual in the oh-so-precise world of photography. I'm sure mine is incorrect by LL standards. By 'output gamut', I meant that of two images rendered using two different profiles.

The dot colors are as you assumed and are also as shown by the plot list to the right of a*b* axes.

In the 3D L*a*b* view on my screen it is seen that red dots are hard up against the sRGB gamut but, as you have observed, the green dots are not.

Thank you for your forbearance 

[mouse]

xpat,  Thanks for your reply.  So now the question that begs to be asked:  When you print the two images saved with these "output"  profiles, is there a discernible difference?

[xpatUSA]

xpat,  Thanks for your reply.  So now the question that begs to be asked:  When you print the two images saved with these "output"  profiles, is there a discernible difference?

Sorry, can't answer that - I don't print. On-screen, for that particular image, there is a discernible difference.

cheers,

[digitaldog]

Sorry, can't answer that - I don't print. On-screen, for that particular image, there is a discernible difference.

Considering the limitations of the display gamut, it's only pertinent for those who's output is to screen. However, print output devices differ a lot, so one output test certainly would not answer all the questions but it's a useful start. Especially if it is a fairly wide gamut device like a modern ink jet.

If your output is only to screen, one has to wonder why anyone would even spend the time here as everyone who views that image will likely see something different. Are they using an sRGB or extended gamut display? Is the mode of viewing color managed? Is their display calibrated and how? Are they vewing on an iPhone, another phone, a tablet, a high end display? It's a mess out there in the on-screen only output world.

[Lundberg02]

Amen. The deacons will assist anyone who needs help to their cars.

[Lundberg02]

Ok. left to right, absolute, relative, perceptual, and saturation. Printed using Photoshop manages colors to my six color Epson on copy paper with the correct settings, saved as a jpg. The tiff and pdf versions were slightly less differentiated. Original image was clipped from the Printer Evaluation Image_V002_ProPhoto.tif available online.
I prefer the Absolute and the Saturated myself, but I always use Perceptual for prints just because the theory is more satisfying.

[xpatUSA]

If your output is only to screen, one has to wonder why anyone would even spend the time here . . .

Andrew, if you would prefer that I don't "spend the time here" then I will be more than happy to accommodate you.

It does seem difficult for a certain type of person to understand that an interest in the technicalities of photography does not necessarily require the practice professional-style printing and does not necessarily require publication on-line of stunning images that magically look the same on all output devices.

. . .  as everyone who views that image will likely see something different. Are they using an sRGB or extended gamut display? Is the mode of viewing color managed? Is their display calibrated and how? Are they vewing on an iPhone, another phone, a tablet, a high end display? It's a mess out there in the on-screen only output world.

So, goodbye to you and goodbye to this benighted topic.