Perceptual intent printer profiles - source gamut baked in?

I read somewhere that the Perceptual intent (in printer profiles) is created under the assumption that images with a certain SOURCE gamut will be used. I.e - the source gamut is "baked in" to the perceptual intent of the printer profile. Is this correct? If so, can the source gamut be determined by inspecting the profile, or does one have to actually ask the creator of the printer profile what assumption was made?

I'm just an amateur, but seeing how beautiful my wide gamut laptop screen is, I'm interested to explore the whole issue of color gamut and printing.

Thanks in advance for any info.

I'm mostly assuming what I am going to say. The printer color gamut is limited by the ink system which is often the limiting factor in a print. Presumably, perceptual intent scales the color in a print (if necessary) to fit within the color gamut of the printer system. There also may be some dependence on the paper used.

Leon Wittwer said:

I'm mostly assuming what I am going to say. The printer color gamut is limited by the ink system which is often the limiting factor in a print. Presumably, perceptual intent scales the color in a print (if necessary) to fit within the color gamut of the printer system. There also may be some dependence on the paper used.

Click to expand...

The question is whether it scales the colors actually in the print, or if it scales all the colors in source colorspace.

Consider a situation where you shoot two photos of a model in a studio setting. There are no dramatic colors in the model, or in her clothing. However, between the two shots, a small neon green ball rolled into the background of the image.

Suppose we want to map both of these images into the sRGB colorspace, and suppose that everything except the neon green ball is within the sRGB color gamut.

If the mapping only looks at the colors actually appearing in the image, then the image without the ball doesn't need any colors changed. Only the image with the ball will require a color shift. The result will be two images with different colors.

If the mapping is based on the source colorspace, then the mapping will be the same for both images. The colors in both resulting images will be the same. However, the colors will be shifted, and shifted more than strictly necessary. The shift will be enough to accommodate not only the neon green ball, but enough to accommodate a neon red or neon pink (even those aren't in these photos).

Leon Wittwer said:

I'm mostly assuming what I am going to say. The printer color gamut is limited by the ink system which is often the limiting factor in a print. Presumably, perceptual intent scales the color in a print (if necessary) to fit within the color gamut of the printer system. There also may be some dependence on the paper used.

Click to expand...

Yes, that much I do know - the perceptual intent scales, rather than simply hard clipping like Relative or Absolute does. It's my understanding that the scaling can involve EXPANDING the source gamut to fill out the printer gamut, though - for example - printers typically have a wider gamut than sRGB in some hue areas.

Btw, I'm not interested in high accuracy at the moment. All I want is to feel confident that my image and printing gamuts are being matched/mapped well. I'm not even sure yet whether Perceptual will always be the most sensible intent.

Good example - thanks Michael.

I suspect that whilst printer profiles can have a fixed Perceptual gamut map baked in, professionals don't use that method to print - they use external software, where they have full control, and where of course the source gamut is known, and can be handled optimally.

I have been printing Epson for many years, now using the 2880. I do not think that "printer color management" is smart enough to scale based on the actual colors in a print. It scales the entire print to ensure the source gamut fits in the printer gamut. I normally let Photoshop manage my printing and use relative colorimetric for most of my prints. I seldom see colors that I can't print. I have never heard of a system that expands a source gamut to fit a print gamut. I would think that that might significantly change the look of a print.

I contacted Graeme Gill (the author of the Argyll CMS), and you're right - the Perceptual intent does not normally perform gamut expansion. (that's what the Saturation intent is for, but that's more for business graphics - not photos).

When you say you have never encountered a color that you can't print, do you mean in the formal sense (the colors in your images are technically within the printer gamut), or are you just talking subjectively?

Leon Wittwer said:

I have been printing Epson for many years, now using the 2880. I do not think that "printer color management" is smart enough to scale based on the actual colors in a print. It scales the entire print to ensure the source gamut fits in the printer gamut. I normally let Photoshop manage my printing and use relative colorimetric for most of my prints. I seldom see colors that I can't print. I have never heard of a system that expands a source gamut to fit a print gamut. I would think that that might significantly change the look of a print.

Click to expand...

Leon Wittwer said:

I have never heard of a system that expands a source gamut to fit a print gamut. I would think that that might significantly change the look of a print.

Click to expand...

I've been thinking more about this, and I'm not so sure it's such a silly idea.. For example, my phone has a vivid AMOLED display, and would be displaying many photos more vividly than they probably should be, yet I find the appearance to be very nice. (same goes for my wide gamut laptop display, although it appears that color management is probably being used more often now, even on the web, than in the past). I've actually asked a printing house about this - will report back. I asked them whether they have any advice on fully exploiting the gamut of their photo printers. Their standard process accepts images in either sRGB or Adobe RGB.

GregS2 said:

GregS2 wrote:.

When you say you have never encountered a color that you can't print, do you mean in the formal sense (the colors in your images are technically within the printer gamut), or are you just talking subjectively?

Click to expand...

I use Photoshop and it tells me when something is out of gamut for whatever color space I'm using for my working space (usually Adobe RGB). I do landscapes most of the time (see site below) and, with the exception of an occasional sunrise or sunset, seldom see any out of gamut colors.

GregS2 said:

Yes, that much I do know - the perceptual intent scales, rather than simply hard clipping like Relative or Absolute does. It's my understanding that the scaling can involve EXPANDING the source gamut to fill out the printer gamut, though - for example - printers typically have a wider gamut than sRGB in some hue areas.

Click to expand...

I think the gamut expansion is uniform in all directions so there could be printer hues that you can't use. A non-uniform expansion would likely change the appearance of a print in odd ways.