Problem assigning a Color Profile of P3 to raw image files?

The OS doesn't assign anything to RAW images. You choose what color space you want RAW images rendered to in Camera Raw. As shown next, I prefer ProPhoto RGB. Yours is mostly likely set to Display P3, which is one of the choices.

Whatever you choose is then the assigned color profile once you open the images from Camera RAW to Photoshop's interface. It's also of course then the profile you'll see when saving the file.

If you want it to be something else, you've never been able to change that in the save or save as dialogue. You have to do that either when opening from Camera RAW, or afterwards with Convert to Profile. Either way, it needs to be done before trying to save your original RAW image to a new file type.

I am curious though why you would want to use such a small profile space as sRGB. You're throwing out an awful lot of color gamut and range that way.

I thought I explained this pretty well the first time. But, apparently not. Let's try again.

The OS is NOT - repeat - NOT assigning/attaching any profiles to your RAW images. Period.

It's using Display P3 as a default color space to translate the RAW image to RGB on the fly so your monitor can display it. Neither your video hardware or monitor can directly display a RAW image. It first has to be translated to RGB values so your RGB device (your monitor) can do something with it. The RAW image itself never changes.

Photoshop is no different. Literally everything you view in Photoshop is, in the end, translated to your monitor profile. That's the color you're looking at. Which only makes sense since that's what you're basing all of your color decisions on. Preview does the same thing. So does every other app in existence. It doesn't matter what color space an image has attached to it, if any, or what an app internally uses for color rendering. Your monitor profile is how you view everything.

And no, Adobe's Camera RAW is not using Display P3 as if that is somehow an already attached profile. That's impossible on a RAW image. It's just a default RGB rendering space so the RAW image can be displayed. But you can change that setting to any RGB profile you want.

So, again, ignore what the OS or Preview says about Display P3 with RAW images. It's only being used to convert the RAW data into something will can be viewed on your monitor.

P3 isn't a great color space. That's why I use the large RGB space, ProPhoto RGB (or Wide Gamut RGB), in Camera Raw. No color gets clipped until it converts down to my monitor profile, which is larger than Adobe RGB.

Here's how Adobe RGB and Display P3 compare.

The red through magenta range is a bit greater in P3 than Adobe.

But Adobe is much larger in the greens.

Not quite as much in the cyan range, but still a lot more than P3. And almost the same in the deep blues.

What does all of that mean? It means if you're using P3 in Camera RAW, you're compressing a lot of color down into a range that is already smaller than Adobe RGB, and that's what you'll be stuck with when you open the image from Camera RAW into Photoshop's main interface. You must start with a larger color space in Camera RAW in order to preserve as much of the original Lab color in the RAW image as you can. Use Wide Gamut RGB or ProPhoto RGB. Then let that convert down to your working color space in Photoshop (I presume you're using Adobe RGB).

more files in our reprographics department that just look wrong (dull washed out) and are tagged as P3

Then someone in your reprographics department needs to stop using Display P3 in Camera RAW. Once they save the file out of PS to a TIFF (or whatever), then the final color translation from RAW to RGB will be based on Display P3's range, and Display P3 will be attached as the image profile. That's where it will be stuck with no easy way to get back to something closer to the original color.

I suspect colour settings in Camera Raw and in CS are set to P3 as defaults and people are accepting the assignment of P3 to RAW files which is just not good.

It may or may not be human error but this particular error is now far more prevalent so something has changed and we would like it changed back!

There's nothing to change. At least, not by Apple or Adobe. Neither of them can possibly know how millions of users manage their color needs. All they can do is pick a default. And that's all it is - a generic default. It's up to the user to understand color management, and then choose the color spaces they need for their workflow.

but 'assigning' Adobe RGB instead nearly always brings them back to a more natural looking state

Assigning should only be used when the image has no profile. And even then, you have to have a good idea what space it needs to in so the color ends up looking like something natural.

You're not bringing anything back to a natural state when you assign a profile. You're only forcing the color in an image to stretch out to a larger RGB space. Including color that may have already been correct. Basically, it's a sledgehammer approach that has no real use.

(more)

It took me a while to find what you're referring to. Get Info doesn't show any color space information. But I finally figured out you're talking about Preview.

Yes, Preview uses Display P3 as a default color space for displaying an untagged image, which all RAW images are. Nothing is actually assigned to the image. In other words, Preview is just using a default to apply some kind of color rendering control for display on the monitor.

If I open an image in Preview that already has an embedded profile, then Preview correctly shows that profile name. Assuming you're doing all of your color work in Photoshop, then simply ignore what Preview says. It's meaningless.

Minor note, on a Mac, you're in macOS. iOS is the operating system for iPhones and iPads.

There is definitely something new going on. In the last few months we have been receiving more and more files in our reprographics department that just look wrong (dull washed out) and are tagged as P3 .

Have yet to fully test and identify where this is happening but 'assigning' Adobe RGB instead nearly always brings them back to a more natural looking state. Note assigning and not converting. I suspect colour settings in Camera Raw and in CS are set to P3 as defaults and people are accepting the assignment of P3 to RAW files which is just not good.

It may or may not be human error but this particular error is now far more prevalent so something has changed and we would like it changed back!

In fact thinking about this, it makes no sense for the P3 monitor space to be assigned as the profile for a Camera Raw image. What camera exists that shoots in a monitor space? Display space should be used to interpret the image for correct rendering to the display, but not assigned as the source space for the photograph. ...Maybe it makes sense for digital art where effectively the monitor is the source device...

Thanks for taking the time to reply in such detail Kurt. I have some understanding of colour but appreciate your depth and clarity of thought :)

I agree with what you say, but I am really here reporting on a new issue which we did not encounter before, rather than questioning why it looks wrong. Regardless of the rights and wrongs we are seeing an increase in images being presented by 'non-colour-savvy' users with P3 as the profile, and poor colour as a result (as demonstrated by the colour-sync visuals you present - colour has been 'thrown away'). Agreed, assigning a larger profile once the info has gone is just stretching the colour back out - nicely explained and a good point! What we really want is to go back to the original RAW condition and then assign the larger profile (or camera profile). Unfortunately we are not in a position to do that because we receive files already 'damaged' by the compression.

What is different is that this was not happening in the past - non-colour-savvy users were not falling into this trap and we were more likely to get images in AdobeRGB, or Srgb presumably having followed a RAW --> AdobeRGB/ Pro-photo --> conversion to Srgb, pathway. ...Whatever is causing P3 to prevail, it is messing up more images than before!

in my reply to self from my earlier message I noted...

"In fact thinking about this, it makes no sense for the P3 monitor space to be assigned as the [default] profile for a Camera Raw image. What camera exists that shoots in a monitor space? Display space should be used to interpret the image for correct rendering to the display, but not assigned as the source space for the photograph. ...Maybe it makes sense for digital art where effectively the monitor is [could be considered as] the source device..."

I suspect that phones also come into this, as so many images are either shot or stored on phones. Also the mix of end-use has shifted from print to screen, so I can see the difficulty in deciding where to set defaults. We encourage 'prepress' colour settings but so many designers now work across media, they need to know more, however - in my experience at least - as the tech has got easier to use, and the technicalities have been abstracted away, the end-users have become less aware of image quality issues and pitfalls - they expect it to 'just work'! ...in that sense this P3 issue is a side-effect of something changing in the 'just make it work' pipeline.

Ugh! You don't even know what the photo vendor is doing on their end, then.

I use a professional EIZO monitor for image and video work. It's calibrated and profiled to D50. Which, quite bluntly, is what everyone should be using. Not the horrible 6500K default. D50 is the only color setting that comes as close as possible to the color we see on a sunny day at noon. It's what the entire printing industry uses. If you purchase a color viewing booth from a place like GTI Graphics, the bulbs and gray paint they use on the units is D50.

Yes, there's a purpose for explaining that first. 🙂

Here's what may be happening. The photo vendor may not have their monitors calibrated at all. And worse, they may be using an RGB profile for their monitor that has nothing to do with a monitor. Worst choice of all would be using sRGB as a monitor profile.

What happens then is everything looks incredibly vibrant. Almost day-glow. You'd think they'd notice that and pick some other canned profile. But if they choose anything with a short color range, like the ancient ColorMatch RGB, their monitor color is going to be WAY off.

So, they think they're sending you images with rich, bright colors, because that's what their monitor looks like. But you see more of what it really looks like. Dull and washed out.

I could be entirely wrong here since I don't know, and can't see their setup. But it sure sounds like they don't know how to manage color.

There's only a couple of ways P3 could end up being the tagged profile. One, they have Camera Raw set to use P3 as the viewing color space. Once the RAW image is opened into Photoshop and saved, that's the profile it gets. Or two, they're using a wide gamut space in Camera Raw, but once it's opened into Photoshop, they may be doing a Convert to Profile and converting the color to P3. Hard to say since it also depends on what the working RGB color space is set to. Is that set to P3? Lots of questions to ask your vendor.

Per your self answer (I missed your added response), RAW images don't have a profile. They can't. It's not an RGB image so cannot be assigned an RGB profile. So any thought that a monitor profile comes into play somehow can't happen. What you view in Camera Raw is determined by the working RGB space you want CR to handle the RAW color translation as. Pick sRGB, and colors will be dulled down. Change it to WideGamut RGB, open the same RAW image, and it will be much more saturated.

But yes, phones and tablets are a nuisance. Though I have to say with my iPhone 14 Pro, with the pro color option and RAW turned on, it takes photos with amazingly accurate color. Almost as good as the Nikon D800 I used to have.

And then there's the device itself. As I look at my previous postings with the images on my 6th gen iPad, it isn't very accurate. While the sRGB and monitor comparison are obviously different on the EIZO, they look identical on the iPad. The red point of the 3D profile within the tristimulus? I can't even see it against the red to the right of where it should be.

But that's also not unexpected. The iPad is using a small RGB space for the display. In the opposite effect of stretching color out, the iPad can't display the richest red beyond a certain point in Lab, so everything past the iPad's 255,0,0 point gets squished down to the same/nearest color it can display.

Basically, color management can drive you nuts. Even if you understand it very well and have things set up properly on your end, you have no control over the millions of monitors out there that aren't even close to accurate. What other people see is very unlikely to look like what you see.

(continued)

This is easy to understand once you get it, though hard to explain, but I'll give it a shot.

Lab is the only fixed color space. It's a computer model of all color we as humans can perceive. All RGB color spaces are a hunk of Lab. Each RGB space is different model - or hunk - of Lab.

ColorSync Utility shows Lab as a huge rectangle. But most models show it as a sphere.

Now, drop an RGB profile into Lab to get idea of how a chunk of Lab works. This isn't a greatly accurate view, but gives you the idea. In any RGB profile, the gamut points are the tips of the most vibrant colors available in that profile. No matter which RGB profile you're talking about, 255,0,0 is the richest red possible - in that profile. So the red point of the profile seen below is 255,0,0, which leave a lot of room between that and the richest red we can see at the limit of Lab's sphere.

Now compare sRGB to my monitor's RGB values. Both are 255,0,0 , but sRGB on the right is much duller than the one on the left. There's nothing wrong here. It's just an indication of how much smaller of a color space sRGB is than my monitor. Also notice the difference in the Lab values. sRGB is closer to identical values, which means it's closer to neutral gray, which is the vertical center of the tristimulus above.

A better visual of these 255,0,0 points is below. Point A is the richest possible red in sRGB. Point B is the richest red possible for my monitor. Quite a bit further out into the brighter, more vibrant reds of Lab.

So, what is the point of all of this? It's to explain why your images looked "better" when you assigned Adobe RGB to your image.

When you did that, you told Photoshop to completely ignore all color related to Lab in the current RGB space and simply move everything. This happens everywhere, including to colors that were already accurate to the subject. As a for instance, pin one end of a rubber binder to neutral gray at the center of Lab. Pin the other end to point A above. Now take point A and stretch it to point B, making that 255,0,0 instead of where it was. Every RGB value along the length of that binder also moves, even if it shouldn't.

Everything got richer when you assigned Adobe RGB because you forced the color values to change, not because it was correct.