kirk_thompson Posted November 9, 2004 Share Posted November 9, 2004 My subjects - & prints - tend to have few out-of-gamut colors, but today I'm making a print with some bright reds & am experimenting with different ways of dealing with clipped/out-of-gamut colors. I'm surprised that my library of PS books says little about this, & I've also been surprised at what different results I got in 4 different print-experiments. (1) The simplest option seemed to be to use Perceptual Rendering Intent, & let the algorithms take care of the problem. But this resulted in very poor modulation of the red tones - they looked 'squished' onto the paper. (2) I also tried Relative colorimetric, & was surprised that the red tones were more varied & better-distributed than with Perceptual. My books had suggested the opposite - that RC is best for images with all the colors in-gamut, & Perceptual is generally better for images with out-of-gamut colors. (3-4) The next thing I did was to Color-Select the reds & desaturate them until the grey gamut warnings went away. This produced better results with both Perceptual & Relative Colorimetric Intents. Relative Colorimetric offered much better modulation of the red tones. What surprised me most was that the sharpness - not just the tonality - in the red areas varied so much in the four different prints. The ones that I'd desaturated were noticeably more detailed in the reds, & the desaturated Relative Colorimetric one looked sharpest of all. I wonder if this is always the case - is it always or generally best to deal with out-of-gamut colors by desaturating the specific color(s) & using RC Intent? Are there other ways of dealing with clipped colors that I should have tried? Another strand of thought: I also wonder if definition within color-areas also varies with the printer profile one uses. I've noticed that Nikon/Epson greens, even when in-gamut, sometimes go all mushy with Perceptual intent, & look better with Relative Colorimetric; but this seems to vary somewhat with what profile I use. In particular, I've tried some relatively inexpensive custom printer profiles that seem to mess up in this area. (I noticed somewhere on the web that Ethan Hansen had advice about this, but don't remember it well - so maybe I'd just better get some of his profiles!) Does anyone have experience/advice to compare on these issues? (If it matters: Epson 4000, matte papers - EM for workprints, HPR & Entrada for final ones.) Link to comment Share on other sites More sharing options...
sean de merchant httpw Posted November 9, 2004 Share Posted November 9, 2004 Take a look at <b>Professional Photoshop: The Classic Guide To Color Correction</b> by <a href="http://www.ledet.com/margulis/articles.html">Dan Margulis</a>. It is a bit more CMYK oriented than most texts, but it is the only book that covers how to deal with these issues that I have read (or seen). <p> That said, the fact that desaturating the reds means brightening the greens and blues where the reds are saturated. This means creating more detail in the unwanted color (cyan or green+blue). Hence it is very reasonable to have the images where you got the sharpest details in the red to be the ones where you desaturated the reds. Manually stepping in and using <a href="http://www.ledet.com/margulis/Poetry.pdf">plate blending</a> (individually manipulating the color channels) is a powerful approach to this type of problem. <p> As an aside, many disparage (i.e., are confused by or simply cannot comprehend) Margulis' writing and favor just learning to do it in RGB. Margulis' text teaches basic color color primarily in CMYK, but it is not just a CMYK text and he covers the concepts very effectively. And just so you know, there are times when it is easy to fix things in CMYK that are very hard in RGB. In short, CMYK is not just for prepress.<p> hope this helps,<p> Sean Link to comment Share on other sites More sharing options...
bruce watson Posted November 9, 2004 Share Posted November 9, 2004 Try this also: Desaturate everything. I often finds this is enough, and it has the added benefit of maintaining the relationships between the colors. If that doesn't do it for you, consider a combination of partly desaturating the color that's giving you a problem and partly desaturating all colors at once. Finally, note that you don't have to print with absolutely everything in gamut. A few spots that are out of gamut probably won't be noticed, and you can generally sacrifice a little bit out of gamut to get a little better saturation. Finally, you could switch to dye inks, which just solves the problem without your having to do anything in Photoshop at all ;-) Link to comment Share on other sites More sharing options...
kirk_thompson Posted November 9, 2004 Author Share Posted November 9, 2004 Hi, Sean - I'll read the article carefully; I'd take Emily Dickinson's advice on anything petaining to perception! So when a specific color is desaturated in a masked area, the relative strength of the complementary color increases (in that area); & this creates greater sharpness - or at least the illusion thereof? (I should admit that I'm suffering from terrible Red-Envy this week. I went to SF MOMA for another look at the Eggleston exhibit. The primary colors of his dye transfer prints are a lost art! (A museum docent told me that dye transfer printing halted not from disuse in the Type C era, but from dangerous toxicity in manufacturing the chemistry.) I've greatly enjoyed, even sort of idolized, his work in books, but CMYK reproduction hasn't done much justice to the original colors.) Link to comment Share on other sites More sharing options...
sean de merchant httpw Posted November 9, 2004 Share Posted November 9, 2004 <i>So when a specific color is desaturated in a masked area, the relative strength of the complementary color increases (in that area); & this creates greater sharpness - or at least the illusion thereof? </i><p> Pretty much. To desaturate a color is to "muddy" it with its complementary color. So when your reds go out of gamut and get clipped, the detail in the complementary color does not get clipped.<p> Whereas if you compress the reds (i.e. perceptual rendering intent) you will also compress the complementary color. So you get less of everything. <p> enjoy, <p> Sean Link to comment Share on other sites More sharing options...
imaginator Posted November 16, 2004 Share Posted November 16, 2004 What is the source of the image? Digital camera, scanned film? Some films have trouble with saturated subjects, especially certain colors (and with subjects like flowers, I think UV contributes) I know it sounds extreme, but I've considered using several different films, each matched to a certain color subject, and doing mid-roll rewinds if needed (or better yet, have two camera bodies, each with a different film) I started thinking about this when I had so much trouble dealing with images of flowers, and also because my printer uses pigment inks (it's just an Epson C82 with DuraBrite inks, and the supplied profile is useless... better to use no profile at all) I haven't really figured out which film to use... it's more of an issue of which films not to use (for each specific color issue with saturated subjects) One film might be fine for most colors, but should be avoided for something like saturated reds. With slide films, you might even try switching labs (Kodak/Fuji) for some control over color saturation. OK, I'm gettting way off topic, and still learning all this myself, but wanted to mention the film apsect (if you use digital, then ignore this ramble) It sort of bothered me that there was no mention of the source image... as if it didn't matter since everything can be "fixed" in PS. Link to comment Share on other sites More sharing options...
Recommended Posts
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In Now