Gamut Volume and DMax

[Robert Boire]

Hi,


I have been following the printer discussion with interest but am still struggling with the definition of
"gamut volume" and "DMax".

I was wondering is anybody has a link to a definition or explanation readily at their fingertips. It would help a lot. I have searched this site and come up with lots of hits but I have not found it (yet).

Thanks for the help.

[Ken]

"gamut volume" and "DMax"

"Gamut" is the number of colors, increasing by multiple millions as the gamut gets "wider" or "larger" (most commonly used terms). A small gamut example would be sRGB, which is a basic monitor profile. A "very large" gamut example would be Kodak's ProPhotoRGB, which is the profile used by most high-end medium format digital backs and Adobe's recommended Photoshop working space. It encompasses colors that printers can not reproduce (yet) and of course can't be displayed on any monitor (yet).

Michael Reichmann's explanation is the best I've read. http://luminous-landscape.com/tutorials/prophoto-rgb.shtml

[digitaldog]

"Gamut" is the number of colors, increasing by multiple millions as the gamut gets "wider" or "larger" (most commonly used terms). A small gamut example would be sRGB, which is a basic monitor profile. A "very large" gamut example would be Kodak's ProPhotoRGB, which is the profile used by most high-end medium format digital backs and Adobe's recommended Photoshop working space. It encompasses colors that printers can not reproduce (yet) and of course can't be displayed on any monitor (yet).

Michael Reichmann's explanation is the best I've read. http://luminous-landscape.com/tutorials/prophoto-rgb.shtml

Sorry but gamut is not the number of colors. A document with 24 bits in sRGB and ProPhoto RGB have the same number of colors as defined by the encoding. The gamut is the size (scale) of the color space in relationship to human vision, plotted on that horseshoe shape plot you often see called the CIE chromaticity diagram. Michael’s plots are in 3d but you get the point.

This may help:http://www.adobe.com/digitalimag/pdfs/phscs2ip_colspace.pdf

Dmax is a term for maximum density as (usually) measured and defined (Status T as an example).

[Ken]

Sorry but gamut is not the number of colors.

Right. For simplification to commonly understood terms, "number of colors" gets the message across without delving into "color" (that word alone has spawned volumes by Munsell's and competing theories), bit depth, etc. It's like the word "focus" which is a generally understood concept without regard to all that it actually entails.

In the same vein, "D-max" could be described simply as "how deep the black is" in a print, without getting into densitometry, ink chemistry and mechanical application, perception, etc.

[probep]

Hi,


I have been following the printer discussion with interest but am still struggling with the definition of
"gamut volume" and "DMax".

I was wondering is anybody has a link to a definition or explanation readily at their fingertips. It would help a lot. I have searched this site and come up with lots of hits but I have not found it (yet).

Thanks for the help.

Some documents:
about gamut volume - http://www.t2f.nu/t2f_rapp_f_81.pdf
about optical density Dmax - http://www.jsa-inc.com/files/Understand_Dens_en.pdf

[digitaldog]

Right. For simplification to commonly understood terms, "number of colors" gets the message across without delving into "color"

Ah, OK but again, it has zero to do with the number of colors so its just confusing (especially to someone asking what color gamut means). I guess you could say a larger gamut has more saturated colors instead of more colors.

[abiggs]

what I don't like is some manufacturers of paper like to use a DMAX figure to show some sort of superiority of product, but the problem is that it doesn't take into account how good that profile would be if you print a color image out. Some colors could gloss over and have so many issues if the deepest blacks are the only desired result.

[bjanes]

Hi,


I have been following the printer discussion with interest but am still struggling with the definition of
"gamut volume" and "DMax".

I was wondering is anybody has a link to a definition or explanation readily at their fingertips. It would help a lot. I have searched this site and come up with lots of hits but I have not found it (yet).

Thanks for the help.

Gamut volume is usually expressed in DeltaE units, where DeltaE is the just noticeable difference in color. See Bruce Lindbloom's site for details.

Here is a 3D plot showing the gamut volumes of sRGB and ProPhotoRGB

[attachment=21126:Gamuts.png]

Regards,

Bill

[Robert Boire]

I was wondering is anybody has a link to a definition or explanation readily at their fingertips.

Whoops, I forgot to put the word "simple" before "link"

Well thanks everybody. I actually understand gamut, its gamut volume and DMax I was struggling with.

Anyways its actually fascinating reading.Guess I will have to brush up on my math.

if anybody wants another link here it is:

http://www.gamutvision.com/docs/gamutvision_equations.html



Thanks again

Robert

[PeterAit]

Sorry but gamut is not the number of colors. A document with 24 bits in sRGB and ProPhoto RGB have the same number of colors as defined by the encoding.

Isn't it more accurate to say that your two 24-bit documents have the same potential number of colors, but more of those colors will be in gamut, and fewer out of gamut, when rendered in ProPhoto as opposed to sRGB?

[digitaldog]

Isn't it more accurate to say that your two 24-bit documents have the same potential number of colors, but more of those colors will be in gamut, and fewer out of gamut, when rendered in ProPhoto as opposed to sRGB?

Two 24 bit color files of differing gamut have the potential to define 16.7 million colors based on the encoding. The gamut has nothing to do with those numbers however.

[Deepsouth]

In reading this, a light bulb (7500K) came on. In looking at the pretty gamut plots, we are not seeing what colors a device has within its gamut. We are seeing those colors re-plotted (translated?) to the color space of the monitor. So we can't look at a gamut plot and say, "Oh, my device can reproduce that particular color that I see on the plot". We can say, a particular gamut is wider in one vector than another, and therefore the gamut plot is giving us a relative indication. Pardon me if this is MOTO, but I've not seen this assertion in any discussion of gamut I've read (and I've seen dozens).

A new (to me) question. How do RGB gamut plots address non-spectral colors, like brown and orange? Do we/should we care, since these colors are mixtures of spectral colors?

[digitaldog]

In looking at the pretty gamut plots, we are not seeing what colors a device has within its gamut.

You should be. The overall gamut of human vision is shown (in 2D, the horseshoe shaped plot) and within it, the shape of the color space’s gamut. Inside that plot are colors in its gamut. 2D plots only tell part of the story of course.

So we can't look at a gamut plot and say, "Oh, my device can reproduce that particular color that I see on the plot".

You can. And depending on the utility, you can load an actual image and see how it plots (in its color space) to the gamut of some other device or image.

[Deepsouth]

You should be. The overall gamut of human vision is shown (in 2D, the horseshoe shaped plot) and within it, the shape of the color space’s gamut. Inside that plot are colors in its gamut. 2D plots only tell part of the story of course.



You can. And depending on the utility, you can load an actual image and see how it plots (in its color space) to the gamut of some other device or image.

I need to clarify my contention. I know what the gamut plot is supposed to represent, but the as-seen colors are not necessarily what a given device will produce. A perfect example is a gamut plot of a printer. It is representing what the printer can reproduce via its CMYK (etc) mode, not what the monitor I am using can represent in RGB. In other words, say I look at a gamut plot of my printer  in a JPG and put my cursor on particular color in the JPG of the plot. The color read out is some value of RGB as three numbers. This color is being mapped to RGB but is supposed to represent a CMYK (etc) device. Can my printer actually reproduce that RGB combo? Yes, -if- the printer and monitor gamut coincide for that RGB value. A quick way to demonstrate this is to enable the Gamut Alarm. I can see the color on an RGB monitor, but Gamut Alarm tells me values that my printer won't reproduce.

I'm not trying to pettifogg or be obtuse, but I deal with relative (dimensionless)  and "absolute" units of measure every day and this seem to be a case in point.

[digitaldog]

A perfect example is a gamut plot of a printer. It is representing what the printer can reproduce via its CMYK (etc) mode, not what the monitor I am using can represent in RGB.

You just plot them on top of each other and that information is visually provided, again, far more useful in 3D.

A quick way to demonstrate this is to enable the Gamut Alram. I can see the color on an RGB monitor, but Gamut Alarm tells me values that my printer won't reproduce.

Its a far cry from useful compared to plotting the two in something like ColorThink where you can see exactly where the gamut mismatch is, by how much (you can even plot the detla’s). The gamut warning in Photoshop treats an out of gamut of 1 deltaE and an out of gamut of 20 deltaE the same with an ugly colored overlay. Not real useful. And if its out of gamut, its out of gamut. You can decide not to print the image, but at some point, you’re going to have to and you either clip to the edge of the printer gamut or compress that and other colors hoping to maintain the color relationships.