Advantage to using a color space larger than capture/print?

[Tim_Lookingbill]

<blockquote>But its the price we pay for a theoretical working space that is this size.</blockquote>

<p>The only difference I see converting from ProPhotoRGB to a much smaller color space such as sRGB (which BTW CS3 Bridge previews my ProPhotoRGB Raw files through) is that bright, intense and dark, rich greens, yellows, cyans and oranges viewed in ACR will slightly shift in hue and saturation toward a noticeably duller version even when converting in Photoshop and ACR to sRGB. Converting to AdobeRGB shows a much more subtle shift.</p>

<p>Now I thought this was a display gamut inducing behavior since my old iMac is closer to sRGB, but a while back Patrick proved it happens even on his wider gamut display. Is this a preview bug? Or is this gamut clipping within a matrix to matrix based working space profile conversion? Or is the math behind the theory just not that perfect at this time?</p>

<p> </p>

[gooseberry]

<blockquote>

<p><em>I thought this was a display gamut inducing behavior since my old iMac is closer to sRGB, but a while back Patrick proved it happens even on his wider gamut display.</em></p>

</blockquote>

<p>I was under the impression that even the new LED-backlit 27" iMac monitors aren't much larger than sRGB, but anyway, since everything has to be matched to the display's gamut, I'd sooner expect any color shifts occuring due to conversion to be visually noticeable in high-end monitors (as they actually can display at least some of those colors), but that's only my logic as opposed to factual knowledge.</p>

<p>What <a href="http://www.cambridgeincolour.com/tutorials/color-space-conversion.htm">rendering intend</a> are you using? -- sounds like you may be using <em>Perceptual</em> (it "squashes" everything even if there are no out-of-gamut colors).</p>

[digitaldog]

<blockquote>

<p>I was under the impression that even the new LED-backlit 27" iMac monitors aren't much larger than sRGB.</p>

</blockquote>

<p>I believe that is correct. </p>

<blockquote>

<p>What <a rel="nofollow" href="http://www.cambridgeincolour.com/tutorials/color-space-conversion.htm" target="_blank">rendering intend</a> are you using? -- sounds like you may be using <em>Perceptual</em> (it "squashes" everything even if there are no out-of-gamut colors).</p>

</blockquote>

<p>RGB working space to working space conversions use Relative Colorimetric as these are simple matrix profiles and don’t have the Perceptual or Saturation tables.</p>

[Tim_Lookingbill]

<p>Patrick does have the 27" iMac but I was referring to his older NEC he had at the time of the discussion on color shifts converting from ProPhotoRGB to sRGB.</p>

<p>This color shift isn't a deal breaker in the commercial world. You only see it after converting it. It's subtle enough that if you walked away and came back to the display after your eyes adjusted you wouldn't notice the difference.</p>

<p>When I enhance Raw shots I've taken of sunlit intensely colored flowers and similarly colored subjects in 16 bit ProPhotoRGB in ACR, I tend to punch it up a bit while keeping the ProPhotoRGB output histogram from clipping with no posterization (saturation blooming) showing up in the flower. I then do a test by switching to sRGB within ACR and notice the shift. Cadmium yellow, a noticeably intense yellow with a bit of red (unlike lemon yellow which has a bit of cyan), will slightly take on a kind of rust brown tint converting to sRGB.</p>

<p>If I convert to my own custom iMac profile this won't happen which may point to a preview bug.</p>

<p>Otherwise it's still a kind of let down when I see this happen, but I just walk away from the computer and come back to the Bridge preview and I don't notice a thing.</p>

[patricklavoie]

FYI I still have a NEC 2690wuxi connect to iMac 27... Just change the computer from a macpro to a more powerful iMac :

)

[gooseberry]

<blockquote>

<p><em>What <a rel="nofollow" href="http://www.cambridgeincolour.com/tutorials/color-space-conversion.htm" target="_blank">rendering intend</a> are you using? -- sounds like you may be using Perceptual (it "squashes" everything even if there are no out-of-gamut colors).</em></p>

<p>RGB working space to working space conversions use Relative Colorimetric as these are simple matrix profiles and don’t have the Perceptual or Saturation tables.</p>

</blockquote>

<p>What do you mean, Andrew? -- under 'Convert to Profile' in PS one is given four options to choose from in the 'Intent' drop-down menu, and <em>Perceptual</em> is one of them. Not that I run any comparative tests, but they all seem to be active when I convert from ProPhoto to sRGB.</p><div></div>

[jacopo_brembati]

<p>Tomek,<br>

ProPhoto and sRGB are matrix profiles.<br>

Perceptual intent does'nt exist.<br>

To execute the transform, CMM modify perceptual to relative.</p>

<p>Jacopo</p>

<p> </p>

[digitaldog]

<blockquote>

<p>What do you mean, Andrew? -- under 'Convert to Profile' in PS one is given four options to choose from in the 'Intent' drop-down menu, and <em>Perceptual</em> is one of them. </p>

</blockquote>

<p>Indeed, the option is there, but with the current V2 ICC working space profiles, there is no perceptual table. Try converting using both options, subtract the two and you’ll see, they are identical. </p>

[john_a5]

<p>For those of you on the mac that want to compare spaces, open your colorsync utility(in Utilities folder in the Applications folder). You can select any profile and view it in a 3D space. If you right click on the selected profile, you can "hold it for comparison". Then, when you select any other profile, you will see how the two compare, even your custom monitor profile. The base profile turns white and the one you are comparing to stays full color, you can see where they mismatch or are contained by each other this way.</p>

[Tim_Lookingbill]

<p>Tomek, below are two sites that deal with the new sRGB v4 color space profile that does have a Perceptual rendering intent table as well as a Black Point Compensation option where both change the preview of the image and output histogram with regard to clipping. The second link is an Adobe forums discussion I started discussing some of its advantages. Just a warning, lots of color geek talk.</p>

<p>http://www.color.org/srgbprofiles.xalter</p>

<p>http://forums.adobe.com/thread/311580</p>

<p> </p>

[digitaldog]

<p>Even with V4 profiles, its not going to do much for you unless its the destination profile. </p>

[chris_ciotti]

<blockquote>

<p>" . . . on the web and if you print with a external lab the result will be worst that anything else."</p>

</blockquote>

<p>This is not entirely true Patrick. Labs like West Coast Imaging (http://www.westcoastimaging.com) can handle ProRGB and the difference is obvious in the final print.</p>

[patricklavoie]

<p>well i was talking about 98% of them let say.. some high end lab (not your typical local one) can handle Adobe RGB or else because they know what they are doing.. but saddly most of them out there still require or use srgb as there color space.. and there old printer cant get something better anyway.</p>

<p>As for obvious in the final print.. a sRGB well done and well saturated to the max of the printer capability could yield to similar result than a lab who use Pro Photo.. not saying its the same thing, just saying that if you saturated enough your sRGB before printing it could look like one made from Pro Photo.. i have made many test to comfirmed that ; )</p>

[jack_hogan]

<p>Since I print mainly to a photo inkjet printer, I have been using Bruce Lindbloom's BetaRGB color space: big enough to cover virtually all output papers (including some not covered by AdobeRGB), but not as large as humongous and virtual ProPhoto.</p>

<p>As I understand it, when a color space is converted into another, the bigger the difference between the two, the more the possibility of introducing significant shifts. That's the rationale for using the smallest working space that will render all of the colors of your desired output medium. And nothing bigger. Any counter indications to BetaRGB?</p>

[gooseberry]

<blockquote>

<p>Tim wrote:<br /><em>[...] the new sRGB v4 color space profile that does have a Perceptual rendering intent table as well as a Black Point Compensation option where both change the preview of the image and output histogram with regard to clipping [..] lots of color geek talk.</em></p>

Andrew wrote:<br /><em>Even with V4 profiles, its not going to do much for you unless its the destination profile.</em></blockquote>

<p>Cheers guys! I have had a read and indeed, it did my head in. I wouldn't even know where to save this profile :S <br />What makes me think it may not be a good idea for me to use it is the advice not to mix v4 with v2; the profiles that I did manage to find on my drive don't specify which version they are (at least it's not stated in the file names themselves).</p>

<p>Anyway, I read numerous articles about rendering intents and BPC when I was trying to understand what all these options mean and how I should use them, yet, surprisingly, no-one mentioned that none of these applies when converting to the most common sRGB, so I'm really glad this insight surfaced here!</p>

[Tim_Lookingbill]

<p>Tomek, you don't have to use the v4 sRGB profile off that site as specified. I downloaded and assigned and/or converted to and from v2 sRGB on my six year old computer that doesn't have not one v4 profile installed except this one.</p>

<p>There is a noticeable shift to the preview mainly in the shadows, but then I don't want that to happen anyway so I don't see the use of this special sRGB. I only offered it as a learning tool and tinker toy.</p>

[gooseberry]

<p>Howdy Tim, I see on your profile picture that you're on an iMac as well. I'm on the aluminum one, but suspect that won't matter for the question I want to ask you: where should I save this (or any other color profiles, for that matter) on my computer? I suspect that would be the folder: Macintosh HD >> Library >> Application Support >> Adobe >> Color >> Profiles >> Recommended -- is that the correct access path?<br>

Thanks!</p>

[Tim_Lookingbill]

<p>HD>Library>Colorsync>Profiles folder. Putting it there allows the profile to be seen by all apps on my system which is OS 10.4.11 Tiger.</p>

<p>Don't know what OS X version you're using. It's very easy to find this out online, but I doubt the directory for profile placement has changed in Leopard and Snow Leopard.</p>

<p> </p>

[gooseberry]

<p>Makes sense (that was the other option that crossed my mind).<br>

I'm on 10.5.8, which has the same folder structure; sRGB v4 is already sitting there...<br>

Ta!</p>

[gooseberry]

<blockquote>

<p>Justine wrote:<br /><em>I've been searching and reading a lot about color spaces, especially ProPhoto. My question is: Is there a real noticeable difference in the final print switching into and out of ProPhotoRGB during editing if my camera captures, and my printer prints, AdobeRGB?</em></p>

</blockquote>

<p>Up till now, I had my workflow set-and-forget to ProPhoto and have never even considered using any other profiles than these that come bundled with PS, but I might be warming up to deviating from this simple-yet-blunt approach. Late Bruce Fraser's <em><a href="http://www.creativepro.com/article/out-of-gamut-finessing-photoshop-color">Finessing Photoshop Color</a></em> article on CreativePro explains some trade-offs between input- and output-centric philosophies in a typical for him easy-to-understand way, which is nicely supplemented by Bruce Lindbloom's <a href="http://www.brucelindbloom.com/index.html?WorkingSpaceInfo.html"><em>RGB Working Space Information</em></a> considerations of some more technical aspects. And finally, a quote from Joseph Holmes's of Natural Light Photography <a href="http://www.josephholmes.com/propages/AboutRGBSpaces.html"><em>All About RGB Working Spaces</em></a> article:<br /> <br />The most important consideration by far when selecting and using a space is to avoid clipping when the image is converted into it. [...] The way that out-of-gamut colors are treated when they are mapped into working spaces is worse and more damaging, on average, than the way that out-of-gamut colors are treated when they are mapped into a printer profile. [...]</p>

<p>To restate the above, when a space is designed, the gamut of the colors being mapped into it is more important to take into account than the gamut of the space into which colors will later be mapped for output. Also, it is important that even colors which will ultimately prove to be outside of your printer's gamut are not clipped first (by having entered a too-small working space), because those clipped colors will print worse than they would if merely mapped inward by the printer profile. They will tend to be more lacking in detail and shifted in hue.</p>

<p>It's also the case that the working space has to give you room to work — to let you edit your images without clipping them avoidably, although giving yourself too much room to work can also backfire because printer profiles have a hard time moving colors a long ways to bring them into gamut. Care should always be taken to avoid both clipping <em>and</em> pushing colors way too far out of your printers' gamuts.</p>

[jack_hogan]

<p>I would add to the list of references the excellent http://graphics.stanford.edu/courses/cs178-10/applets/locus.html which, in addition to the theory, gives you an in-depth hands-on approach.</p>

<p>My understanding is that -ideally- your camera and printer's color spaces should coincide so that you would just stay in that as your working color space. Since this is never the case, second best would be a working color space that could contain all of the colors that your camera AND printer (assuming that's your final output medium) can produce plus a bit of headroom to deal with rounding errors. ProPhoto is, according to some of the references mentioned in the post above, too big - therefore sometimes creating virtual colors during processing that are so far off the real colors that the output medium is able to produce that the conversion of some ProPhoto colors to the printer's color space is the mathematical equivalent of guesswork.</p>

<p>Short of paying for Joseph Holmes' custom made profiles (which I am nevertheless considering), I have bought into Bruce Lindbloom's BetaRGB as a workable compromise. I'd be interested to hear what other forum members think about this approach, or if there are newer, better compromises available today.</p>

[gooseberry]

<p>I've got one, hopefully quick, general question regarding color profiles I'd like to ask: what's in the .icm/.icc files? -- do they contain only a handful of numbers describing White Point, Gamma, and xy "coordinates" for the three primaries, as suggested by <a href="http://forums.adobe.com/thread/375792?tstart=2">Ian Lyons's post on Adobe Forum</a> re defining BruceRGB?</p>

[jack_hogan]

<p>*Edit my earlier post: the beginning of the second paragraph should read as follows*<br /> My understanding is that -ideally- your camera and printer's <em>gamuts </em>should coincide so that you would just <em>use the smallest color space that would contain that gamut</em> as your working color space.</p>

[digitaldog]

<p>Digital cameras technically don’t have a color gamut, certainly one that is easily defined like that of a printer gamut. </p>

[Tim_Lookingbill]

<p>Jack, interesting color science link. I would need a ton of coffee in order to wrap my head around what and how to apply it to photography. I have no thoughts on using BetaRGB and I did try it out a while back. I don't use it because I didn't see any advantage in the limited amount of printing I do over using any other color space.</p>

<p>As Andrew indicated about digital camera's not having a color gamut, I wonder how the color scientists that created that site measured the gamut of the Nikon D200 and Canon 30D in producing their RGB 3D gamut diagram included in the applet. From examining the applet they seem to show a digital camera having a bigger gamut than human vision. I may be reading it incorrectly, so not sure.</p>

<p>The only way of possibly measuring the gamut of a digital camera's capabilities of capturing <em>scene gamut </em>is to measure the response of their RGB (Bayer) filters in front of the sensor, but then you'ld have to figure out what gets passed from the sensor electronics and through the camera's A/D converter to know for sure and all that's proprietary information involving voltage measuring. Also once you have to rely on software to reconstruct what's captured off the sensor then it becomes anyone's guess. Wonder what scene was used to establish gamut capture capability if that's how those color scientists measured it on that applet site.</p>