What's the point of using color spaces exceeding the visible spectrum?

[Coloreason]

 What's  the point or benefit of using color spaces like ProPhoto RGB which vastly exceeds the visible spectrum?

[digitaldog]

http://www.adobe.com/digitalimag/pdfs/phscs2ip_colspace.pdf

[Peter_DL]

What's  the point or benefit of using color spaces like ProPhoto RGB which vastly exceeds the visible spectrum?

To hold colors, such as e.g. intensive (ly saturated) yellow hues as found in nature
by means of a simple construct with just three R/G/B primary colors.

Peter

--

[Coloreason]

Thank you for the input

[Waeshael]

Technically there is no color space that can exceed the visible spectrum, because colors are made by the brain. What the camera collects is light of various wavelengths from long to short. The sensor just captures the photons from these various wavelengths and counts them up and says there were 2 million counts for the long wavelength,  3 million counts for the middle wavelength, and zero counts for the shortest wavelength, and if the human visual system experienced this mix of wavelengths it would create the sensation of something it calls yellow. The camera software attempts to emulate what the brain would produce by sending various signals to the monitor which has a matrix of red, green, and blue LEDs. It stimulates the right combination of LEDs to produce a sensation in the brain of "yellow." These signals are nothing like the original wavelengths of light coming from the scene - but they have the same effect on us as if we were at the scene - well almost. There is actually no "color" in nature.
So the purpose of color management is to create something from a monitor display, or from a film, or a print, that makes us believe we are seeing the wavelengths of light that were coming at us in the original scene. Ha-ha. So don't sweat it; the technology isn't available to do this right.

It is true that the cameras can detect wavelengths of light that we can't recreate in our brains - very long wavelengths and very short. In addition cameras can create combinations of signals that are beyond what the monitor can display because the RGB color system that is in most electronic display systems is very limited in what it can produce - no bright yellows and no deep reds for instance. Printers on the other hand can produce bright yellow, and a whole lot of other colors that the RGB system can't make.
 So, do you want the final image to display colors you can't see on your monitor -If you have see "kodachrome" -like colors in the original scene, and these colors are important to you, you can set up your color management system to save those "color" conversions and move them to the printer, even though you won't be able to see them on the display monitor. The printer paper/ink selection can then print those "kodachrome" colors.

If you are shooting for the web, then you should start with sRGB color space and continue with it to the final JPEG for the web. Older web browsers don't color manage and anything other than sRGB leads to flat pictures. If your end viewer is looking at your pictures on a laptop or an LED display, don't worry about working in a color space bigger than the monitors can display - which is sRGB. (there are monitors with a bigger color space but they have more imaging issues that you would probably want to deal with. Barco sells them for $10,000 + for medical use.)

I use a big color space to work in and I do convert my captured data (either in camera or in the RAW converter) to at least "adobe 1998" color space, because I do a lot of PP on my images and I need "elbow room" to move colors around without them getting "scrunched" by a limited color space. In the end I have to convert to sRGB to show them on the web - though for my friends they get images profiled for something larger in case the end result is to be a print.

People that make a living producing heirloom quality prints spend a great deal of time and money dealing with color management issues, and I would defer to their teaching if you decide to head in that direction - much useful teaching is right here at TLL.

[ErikKaffehr]

Hi,

The short answer is that to be able to describe as many visible colors as possible you need a color space that contains colors that are outside our color vision.

A somewhat longer answer is that RGB color is described by three primaries, name Red, Green and Blue. These primaries bound a triangle which defines the colors that the RGB can describe. To enclose all colors that can be resolved by human vision this bounding triangle needs to be larger than the area resolved by the human eye.

The "horseshoe-diagram" is bounded by spectral colors, that are fully saturated. The largest color spaces like "Prophoto RGB" have primaries outside the "horseshoe-diagram" and I don't think that those colors are physically possible as they would be more saturated than spectral colors. They are mathematical constructs needed to be able to represent all colors using the primaries.

Best regards
Erik

What's  the point or benefit of using color spaces like ProPhoto RGB which vastly exceeds the visible spectrum?

[Tim Lookingbill]

Large gamut color spaces, or working spaces as they should be referred to, allow mapping of colors more easily when editing. All color viewing devices like display's and printers have unique 3D color gamut models that allow this mapping to occur using apps like Photoshop.

One's display may be capable of displaying one particular hue of cyan over another and a larger working space will make it easier to get there.

This was the argument against color management (especially with regards to calibrating and profiling displays) from the beginning by those who insisted on using their monitor as their working space because they saw they could get/retain colors achieved in legacy files edited in that space.

They didn't understand that their computer was a dumb machine and must be told everything even when a new standard of encoding data for ICC based color management was introduced that seemed to prevent them from doing things the old way when what was really happening was getting everyone on the same page (map) requiring everyone have the same directions for getting there.

Large working spaces allow mapping through edits of all possible colors produced by the device (display) more easily. You can see this just by assigning a large working space to a new Photoshop document and examining colors in the Color Picker compared to assigning like say sRGB to a new PS doc. They will be noticeably different especially when viewed on a wide gamut display.

[digitaldog]

Large gamut color spaces, or working spaces as they should be referred to...

Only if they are working spaces. There are wide gamut color spaces that are not wide gamut working spaces. This is not simply semantics! When I convert my images into an output color space and edit it, its not in a working space and can be wide gamut.

One's display may be capable of displaying one particular hue of cyan over another and a larger working space will make it easier to get there.

Yes. And the color that is out of display gamut may be reproducible on another device so its important not to throw the baby out with the bath water just because one device in the change is of a lower, limited gamut.

This was the argument against color management (especially with regards to calibrating and profiling displays) from the beginning by those who insisted on using their monitor as their working space because they saw they could get/retain colors achieved in legacy files edited in that space.

Prior to Photoshop 5, its the only way we could work. Until that product, the masses had to edit using their display color space as a working space (the term working space was not yet generally even understood by users)

They didn't understand that their computer was a dumb machine and must be told everything even when a new standard of encoding data for ICC based color management was introduced that seemed to prevent them from doing things the old way when what was really happening was getting everyone on the same page (map) requiring everyone have the same directions for getting there.

We had ICC color management long before Photoshop 5. We just had to use our display color space as our working space. We had a working space called ColorMatch RGB years before Photoshop 5 and the concept of a working space (or differing working spaces).

[Peter_DL]

The short answer is that to be able to describe as many visible colors as possible you need a color space that contains colors that are outside our color vision.

A somewhat longer answer is that RGB color is described by three primaries, name Red, Green and Blue. These primaries bound a triangle which defines the colors that the RGB can describe. To enclose all colors that can be resolved by human vision this bounding triangle needs to be larger than the area resolved by the human eye.

The key moment probably was when the red lamp had to be moved to the other side of the target color to be matched by three R/G/B light sources, thus delivering a negative r matching function, in the original color matching experiment.
I think it was before 1976 - though I was not there 

Peter

--

[Tim Lookingbill]

Thanks for the clarifications, Andrew.

That ColorMatchRGB PressView monitor must've been one gorgeous display back then. Wonder how it rendered and matched CMYK cyan viewed under a D50 light source? Under the Solux 4700K desk lamp it's insanely intense and lovely to look at, but I swear that color plays tricks on my eyes.

My fairly new Dell 2209WA sRGB-ish LCD can't even come close to rendering that cyan properly, but my old G5 iMac that finally died could.

But I'm not understanding your mentioning the differences of working spaces and color spaces with regard to wide gamut. If we don't use these mathematically synthetic spaces as working and archiving color spaces then what other use are they for?

[digitaldog]

But I'm not understanding your mentioning the differences of working spaces and color spaces with regard to wide gamut. If we don't use these mathematically synthetic spaces as working and archiving color spaces then what other use are they for?

A working space is a synthetically created theoretical color space built with simple values, white point, chromaticity values for RGB and a TRC Gamma. They are based on a Quasi-Device Independent, non real world emissive like device. They are well behaved meaning that R=G=B is always neutral. They can be low or high (wide) gamut. We can have a wide gamut color space (convert ProPhoto to Epson 9800 Luster) that is not a working space. Its still a color space. Its still wide gamut (assuming everyone agrees what is wide). Its not a working space. So:

Large gamut color spaces, or working spaces as they should be referred to...

In some cases, they should not necessarily be referred to as working spaces.

[Tim Lookingbill]

Oh, not the in-depth and enlightening explanation I was expecting, Andrew.

Next time I'll pick my words more carefully and be more precise. I think I did make a valid point about working spaces which is the what the topic was about since it reads...

What's the point of using color spaces exceeding the visible spectrum? I assumed the OP was referring to ProPhotoRGB and other wider synthetic color spaces. Never implied or thought a printer color space (Epson 9800 Luster) exceeded the visible spectrum.

[digitaldog]

What's the point of using color spaces exceeding the visible spectrum?

They allow us to encode colors that fall outside other such color spaces where the limits of the primaries fall within visible spectrum.

If you want to fit a round peg in a square hole, there has to be overlap. Even if that overlap wastes some space (the shape has to be round).

[Coloreason]

Thanks for the answers everyone, and sorry for the later reply
These explanations make perfect sense:

primaries bound a triangle which defines the colors that the RGB can describe. To enclose all colors that can be resolved by human vision this bounding triangle needs to be larger than the area resolved by the human eye.

If you want to fit a round peg in a square hole, there has to be overlap. Even if that overlap wastes some space (the shape has to be round).

by the way, talking about working spaces, some explanations here made me think that there may be two different meaning of this. AFAIK working spaces is a feature (tool) in color managed programs like Photoshop, Illustrator, etc. to assign color profiles to new documents or placed/ copied documents depending on user's policy and also to display untagged documents in a certain (space) profile. Also the choice of working spaces in the color settings affect the color values in the color palettes with converting the values when switching between palettes with different color models like RGB and CMYK, - the values represent the color in the space (profile) selected for a working space. In other words, I thought all a working space (profile) is simply an option in the color settings of these programs to choose for a special propose one of the color spaces(profiles) available.

[RazorTM]

http://www.adobe.com/digitalimag/pdfs/phscs2ip_colspace.pdf

That's an excellent explanation of the basics.  Thanks for the link!!!

[Tim Lookingbill]

AFAIK working spaces is a feature (tool) in color managed programs like Photoshop, Illustrator, etc. to assign color profiles to new documents or placed/ copied documents depending on user's policy and also to display untagged documents in a certain (space) profile. Also the choice of working spaces in the color settings affect the color values in the color palettes with converting the values when switching between palettes with different color models like RGB and CMYK, - the values represent the color in the space (profile) selected for a working space. In other words, I thought all a working space (profile) is simply an option in the color settings of these programs to choose for a special propose one of the color spaces(profiles) available.

Some of what you said is right and some of it's wrong but it's way too complicated to sort out and explain.

I can assure you what I said about Working Spaces is correct with regard to Editing Spaces and how it works with the color gamut of your monitor. The bigger the Working/Editing/Color Space the easier it is to utilize the monitor's full color gamut editing color in images more so for Raw over jpegs from a digital camera which "Encode/Write" their jpeg RGB data to a Working Space, Editing Space, Color Gamut, Color Space or Profile (usually sRGB or AdobeRGB). The terms are all interchangeable in representing that hole to fit the peg into.

[Coloreason]

Some of what you said is right and some of it's wrong but it's way too complicated to sort out and explain.

I can assure you what I said about Working Spaces is correct with regard to Editing Spaces and how it works with the color gamut of your monitor. The bigger the Working/Editing/Color Space the easier it is to utilize the monitor's full color gamut editing color in images more so for Raw over jpegs from a digital camera which "Encode/Write" their jpeg RGB data to a Working Space, Editing Space, Color Gamut, Color Space or Profile (usually sRGB or AdobeRGB). The terms are all interchangeable in representing that hole to fit the peg into.

Hi, thanks for your reply. May be my understanding about working spaces is incomplete but I'm pretty sure what I said is perfectly correct. I think what you and some others mean when they say "Working spaces" is another term for "Color spaces" described in color profile files. And often people use "Color profile" and "Color space" meaning the same thing which make sense. I guess some people also use "Working space" meaning the same thing too because any color profile can be assigned and used as a working space. However, I think this adds to the confusion and believe that it will be more clear if using the term "Working space" only means the currently assigned/used working space in a color managed program.
Just my 2 cents

[Tim Lookingbill]

You don't assign any profile, color space, working space to an image that doesn't already have one embedded. You have to KNOW what color space/profile that image's RGB data was ENCODED/WRITTEN to for images that don't have an embedded profile/color space/working space (like some jpegs from certain brands/models of digital cameras). The intended preview will be messed up/not as intended if assigning the wrong color space to an image.

Your use of the term assign is what's confusing the issue because it makes people think it's OK to assign (as it's worded in Photoshop) any color space to an image. Maybe you're using the term loosely and even then I wouldn't know how it would apply in this case but regardless, I'ld caution against using this term for the reasons lined out above.

[Coloreason]

You don't assign any profile, color space, working space to an image that doesn't already have one embedded. You have to KNOW what color space/profile that image's RGB data was ENCODED/WRITTEN to for images that don't have an embedded profile/color space/working space (like some jpegs from certain brands/models of digital cameras). The intended preview will be messed up/not as intended if assigning the wrong color space to an image.

Your use of the term assign is what's confusing the issue because it makes people think it's OK to assign (as it's worded in Photoshop) any color space to an image. Maybe you're using the term loosely and even then I wouldn't know how it would apply in this case but regardless, I'ld caution against using this term for the reasons lined out above.

When I say "assign" I mean in color managed programs like Photoshop, you open the Color Settings window and in the Working Spaces section, choose (assign) a color profile for RGB, CMYK, and etc. My point was that because any color profile can be selected (assigned) for a working space, some people like you and even the author of this Adobe tutorial use the term "Working Space" instead of  the "Color Space" of a color profile. I believe this is confusing because this setting does nothing to a document with a color profile (as you said, you don't assign to an image with a profile) but only affects (assigns) the working space profile for display (altering the video card color values only) of untagged images (images without color profiles). The selected working space in the Color Settings also appears for the default color space (to be assigned) when you choose File > New to create a new document but you can select any other color profile from the menu.

[MarkM]

some people like you and even the author of this Adobe tutorial use the term "Working Space" instead of  the "Color Space" of a color profile.

People use the term to refer to color spaces that are designed to be used as working space such as AdobeRGB(1998). Just because you can assign your printer profile as the working space in photoshop's color settings doesn't make it a working space any more than eating soup with a fork makes a fork a spoon.

I think Andrew answered the question perfectly well up above in response #10.