Ad
Ad
Ad
Pages: [1] 2 »   Bottom of Page
Print
Author Topic: matte papers, gamut and color range question  (Read 10586 times)
madmanchan
Sr. Member
****
Offline Offline

Posts: 2101


« on: January 16, 2006, 03:36:59 PM »
ReplyReply

Hi everyone, I'm printing some autumn color landscape images on matte papers like Entrada Fine Art Natural and Epson Velvet Fine Art using my Epson 2200.  Now, my understanding is that the color gamut of these papers is more limited to that of, say, RC papers like Epson Premium Luster and Ilford Smooth Pearl.  This seems to be especially true of reds and yellows.  Of course, these colors feature quite prominently in many of my autumn color images -- lots of deep reds and yellows, and I'd like to try to reproduce them as faithfully as I can.

I am having trouble getting these deep saturated reds and yellows in my prints on these papers.  Any tips for how to improve the color range on these papers?  Are there some color editing tricks one can use in PS?  Or is it just a limitation of these papers and I'd be better off with RC and glossy papers?

FYI, my LCD is well-calibrated and I print using ImagePrint 6.1 Lite using the Colorbyte-supplied profiles.  

Thanks,
Eric
Logged

Mark D Segal
Contributor
Sr. Member
*
Offline Offline

Posts: 6818


WWW
« Reply #1 on: January 16, 2006, 04:50:09 PM »
ReplyReply

The paper may not be the critical part of the problem. It could simply be that your colours are out of the printer's gamut whatever the paper. What rendering intent are you using? What colour space are you using? Do you select black point compensation?

Try switching rendering intent back and forth between Perceptual and Relative Colorimetric and see whether that helps.

If you are using a wide gamut colour space such as ProPhoto, try switching back to ARGB98 - sometimes that helps.

One is looking for the combination of settings that achieves the most satisfactory gamut compression for the needs of the particular image.

If none of that does the trick, which I have experienced quite often, I "custom-bake" the tones as close to what I want as I can by adding a Selective Colour Adjustment Layer. Then I go into RED, and reduce cyan, increase yellow and either increase or decrease black depending on what the result looks like. Then I go into YELLOW, and increase yellow, with small adjustments to C, M, K as needed.

Alternatively, or in combination, you may add a Hue-Saturation Adjustment Layer, select the Red and then the Yellow channel and adjust H, S, B in each to taste.

I make all of these adjustments with my soft-proof activated, so I am simulating on the monitor what should come out in the print. These manoeuvres are often a partially successful way of - in effect - manually creating an in-gamut (i.e. printable) range of colour in the troublesome hues that comes close enough to what is in your mind's eye to be satisfactory.
Logged

Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....." http://www.luminous-landscape.com/reviews/film/scanning_workflows_with_silverfast_8.shtml
madmanchan
Sr. Member
****
Offline Offline

Posts: 2101


« Reply #2 on: January 16, 2006, 06:42:36 PM »
ReplyReply

Hi Mark,

Yes, the issue I'm running into is that my colors are out of gamut (the PS soft-proof shows this), and this tends to happen more on the matte papers I'm using because their color range is more limited than that of, say, the RC papers.  (The plots on drycreekphoto.com confirm this.)

So I guess I should have rephrased my question this way: are there ways to massage the gamut compression / gamut mapping so that deep reds and yellows can be obtained when the original colors are out of gamut, or am I out of luck?

To answer your questions ... I am using ProPhoto RGB as my working RGB space.  I have experimented with both Perceptual and Rel. Col. intents.  I have not tried converting to Adobe RGB first before trying the gamut mapping -- I did not know that this might help, but I'll give it a try.

Are selective color adjustment layers and hue-saturation adjustment layers the best way to control these wild reds and yellows?

Thanks,
Eric

Quote
The paper may not be the critical part of the problem. It could simply be that your colours are out of the printer's gamut whatever the paper. What rendering intent are you using? What colour space are you using? Do you select black point compensation?

Try switching rendering intent back and forth between Perceptual and Relative Colorimetric and see whether that helps.

If you are using a wide gamut colour space such as ProPhoto, try switching back to ARGB98 - sometimes that helps.

One is looking for the combination of settings that achieves the most satisfactory gamut compression for the needs of the particular image.

If none of that does the trick, which I have experienced quite often, I "custom-bake" the tones as close to what I want as I can by adding a Selective Colour Adjustment Layer. Then I go into RED, and reduce cyan, increase yellow and either increase or decrease black depending on what the result looks like. Then I go into YELLOW, and increase yellow, with small adjustments to C, M, K as needed.

Alternatively, or in combination, you may add a Hue-Saturation Adjustment Layer, select the Red and then the Yellow channel and adjust H, S, B in each to taste.

I make all of these adjustments with my soft-proof activated, so I am simulating on the monitor what should come out in the print. These manoeuvres are often a partially successful way of - in effect - manually creating an in-gamut (i.e. printable) range of colour in the troublesome hues that comes close enough to what is in your mind's eye to be satisfactory.
[a href=\"index.php?act=findpost&pid=56084\"][{POST_SNAPBACK}][/a]
Logged

Mark D Segal
Contributor
Sr. Member
*
Offline Offline

Posts: 6818


WWW
« Reply #3 on: January 16, 2006, 07:16:06 PM »
ReplyReply

Eric,

If the whole solution to your problem consisted of changing papers, end of story - just use a wider gamut paper; but I suspect (a) it won't solve the problem, and ( like me you want to eat your cake and have it (i.e. matte with great gamut)! So what to do?

I've raised this in Forums before and gotten back answers about selection of rendering intent and colour space. When neither solved it, I fished around for other solutions and hit on the Adjustment Layer approach, because it is easy, finely targeted, and to a considerable extent produces more acceptable results.

It is impossible to say whether this is the best approach, because as you know, there are always a gazillion ways of doing things in Photoshop and unless one knows all the ways and has tried them, one cannot say with any certainty what is best. All one knows is what is workable from what one has tried. If there are others out there reading this stuff who have a better bag of tricks for solving this gamut compression issue, I'd be just as keen as you to read what they advise.
Logged

Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....." http://www.luminous-landscape.com/reviews/film/scanning_workflows_with_silverfast_8.shtml
Ray
Sr. Member
****
Offline Offline

Posts: 8849


« Reply #4 on: January 16, 2006, 10:34:49 PM »
ReplyReply

Quote
To answer your questions ... I am using ProPhoto RGB as my working RGB space.  I have experimented with both Perceptual and Rel. Col. intents.  I have not tried converting to Adobe RGB first before trying the gamut mapping -- I did not know that this might help, but I'll give it a try.
[a href=\"index.php?act=findpost&pid=56085\"][{POST_SNAPBACK}][/a]

I'm surprised you haven't experimented with 'saturation' intent. There's no doubt this rendering intent delivers more saturated colors. I believe the 2200 is a pigment based printer with a similar gamut to the 7600 which I use. Some time ago I was trying to get a 'wow!' effect in prints I was making of a friend's painted copies of some Van Gogh works (which I'd photographed using my D60). The painter had emphasised the saturation of the primary colors that Van Gogh was fond of and which often look pale and faded in many reproductions of Van Gogh in art books. I was really surprised at just how rich and solid I could get the reds, yellows and blues using saturation intent, especially the blues and yellows, (using the ProPhoto space of course).

Another consideration is the printer profile. I've actually done the following experiment with my 7600, although it can get a bit tedious messing around in this way, and sometimes I seriously wonder if I should be doing something else like getting a life   .

I created a pure yellow square in a new document in Photoshop, in the ProPhoto space, ie. 255,255,0. In proof colors, this yellow was way out of gamut. It required a reduction of 11 pts in hue/sat control to get it back in gamut which, interesting or not, resulted in the numbers changing to 241,241,14.

But I ignored the out-of-gamut warning and printed the yellow square on Premium Lustre, once using the Bill Atkinson profile for Prem Lustre and again using the standard Epson profile that ships with the printer.

The difference was staggering. The Atkinson profile produced a rich, solid yellow as yellow as the ink in the cartridge. The Epson profile produced a weak yellow that was almost beige. In fact it didn't even look as yellow as the same 255,255,0 yellow in sRGB which I also printed with the epson profile.
Logged
madmanchan
Sr. Member
****
Offline Offline

Posts: 2101


« Reply #5 on: January 17, 2006, 08:31:14 AM »
ReplyReply

Thanks for your suggestions and experiences, Ray.  I'll experiment some more along these lines.  I don't have many options for profiles (don't have the equipment to build my own), but I am using the ImagePrint profiles which are supposed to be about as good as "canned" profiles get, so I'm hopeful I can optimize the reds/yellows using the techniques you and Mark suggested.

Eric
Logged

Mark D Segal
Contributor
Sr. Member
*
Offline Offline

Posts: 6818


WWW
« Reply #6 on: January 17, 2006, 12:59:02 PM »
ReplyReply

I'm interested in the observation about Bill Atkinson profiles that Ray mentioned - and there is nothing like hands-on visual research to see things as they are.

I have not noticed any stunning difference in colour rendition on Epson Enhanced Matte paper between the IP profile and the Epson driver/Epson profile, be it with a 4800 or a 9800. I have also had a custom profile made for my 4800, and eventhough it is a good profile, it doesn't really distinguish itself from the other two. So what does BIll Atkinson do that neither Epson nor Colorbyte do? I wonder whether Bill Atkinson will produce commercially available profiles for the 4800/7800/9800 series of Epson printers and papers?
Logged

Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....." http://www.luminous-landscape.com/reviews/film/scanning_workflows_with_silverfast_8.shtml
Brian Gilkes
Sr. Member
****
Offline Offline

Posts: 431


WWW
« Reply #7 on: January 17, 2006, 04:26:26 PM »
ReplyReply

Mark,
Bill does a lot. Check out his site. He also has a new set of profiles for Epson papers and the U K3 inkset.
Eric,
The ARGP(1998) space is insufficient for the wide gamut of this inkset . Try Ektaspace PS5 from Joseph Holmes. He does Autumn colours too.
HTH
Brian
www.pharoseditions.com.au
Logged
Mark D Segal
Contributor
Sr. Member
*
Offline Offline

Posts: 6818


WWW
« Reply #8 on: January 17, 2006, 04:54:14 PM »
ReplyReply

Thanks Brian, I'll take a look.

Now, while it is true that ARGB98 is in some respects a smaller gamut working space than Ultrachrome K3 can reproduce, at the same time alot of gamut compression can be needed to reduce from ProPhoto to K3. When confronted with the problem we are discussing, by changing the colour space from ProPhoto to ARGB98, in effect one is making Photoshop squeeze the offending colours back into gamut, and it may well over-shoot the objective a bit, depending on the image. That is why I often prefer making my own targeted colour adjustments; but sometimes narrowing the workspace does wonders in just one simple move.
Logged

Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....." http://www.luminous-landscape.com/reviews/film/scanning_workflows_with_silverfast_8.shtml
Ray
Sr. Member
****
Offline Offline

Posts: 8849


« Reply #9 on: January 17, 2006, 11:00:58 PM »
ReplyReply

Quote
I have not noticed any stunning difference in colour rendition on Epson Enhanced Matte paper between the IP profile and the Epson driver/Epson profile, be it with a 4800 or a 9800. I have also had a custom profile made for my 4800, and eventhough it is a good profile, it doesn't really distinguish itself from the other two. So what does BIll Atkinson do that neither Epson nor Colorbyte do? I wonder whether Bill Atkinson will produce commercially available profiles for the 4800/7800/9800 series of Epson printers and papers?
[a href=\"index.php?act=findpost&pid=56129\"][{POST_SNAPBACK}][/a]

That's an interesting comment, Mark. I've often wondered if I should splash out  on Image Print. I'm getting along fine with Qimage. However, if IP's profiles for the 7600 are significantly better than Bill's profiles, the exhorbitant price of IP (for wide format printers anyway and from an amateur's perspective) just might be justified.

My understanding is, Bill Atkinson was so impressed with the potential of the 7600, but disappointed with the Epson profiles that shipped with it, he went to a lot of expense and trouble creating his own profiles, but presumably in collaboration with Epson. I've been told by the local Epson agent that this collaboration has extended to the new series of printers, the 4800 and wider, so these printers are now shipping with Bill Atkinson calibre profiles as standard, but with the Epson name. I therefore wouldn't expect you'd see much difference using a profile for the 9800 downloaded from his website, but I'd be interested to hear if you do try his profiles and see a difference.
Logged
Mark D Segal
Contributor
Sr. Member
*
Offline Offline

Posts: 6818


WWW
« Reply #10 on: January 18, 2006, 07:54:20 AM »
ReplyReply

Brian, I have scrolled through all of Bill Atkinson's (impressive) offerings, and I observe that he has not published profiles for enhanced matte paper, which is what I use. While luster and glossy finish may exhibit better luminosity in the quarter-tones, I don't know whether this also means it has noticeably better colour gamut in red-yellow. I intend to look into this more carefully when an opportunity arises to process the same file through two Epson 4800s, one equipped with matte black and the other with photo black. It is too time-consuming and expensive to perform this test by changing inks on the same printer.

Now Ray, I just this minute produced a three square-inch print filled with solid yellow in ProPhoto 16-bit on my Epson 4800 using the Epson profile and driver and Epson Enhanced Matte paper. The result is as yellow as it gets - just what my monitor shows me and just what I would expect from RGB 255, 255, 0 (LAB with B=128). So I begin to wonder how much better I can do with another paper and profile - that remains to be seen.

But Eric's first post seems to indicate the red-yellow gamut problem persists regardless of the paper, which is what I would expect. Eric, could you download an Atkinson profile if you are using luster paper, try it, and let us know whether you confirm Ray's observation that it does a much better job with these colours compared with the Epson profile?
Logged

Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....." http://www.luminous-landscape.com/reviews/film/scanning_workflows_with_silverfast_8.shtml
Chris_T
Sr. Member
****
Offline Offline

Posts: 541


« Reply #11 on: January 18, 2006, 12:45:11 PM »
ReplyReply

Quote
Yes, the issue I'm running into is that my colors are out of gamut (the PS soft-proof shows this), and this tends to happen more on the matte papers I'm using because their color range is more limited than that of, say, the RC papers.  (The plots on drycreekphoto.com confirm this.)

So I guess I should have rephrased my question this way: are there ways to massage the gamut compression / gamut mapping so that deep reds and yellows can be obtained when the original colors are out of gamut, or am I out of luck?
[a href=\"index.php?act=findpost&pid=56085\"][{POST_SNAPBACK}][/a]

I also notice that the problem colors will show up as OOG when softproofing with the EEM profile but not with a RC paper profile. However, I am not sure how to interprete this. OOG is straight forward when comparing two known and well defined color spaces, say between Adobergb1998 and srgb.

But I think OOG from a printer profile is less straight forward. My understanding is that a printer profile is basically a look up table mapping one color space to another, and is NOT a color space by itself. Let's assume that your work space is Adobergb1998, and you get OOG when softproofing with the EEM profile. What color space is PS comparing against Adobergb1998 to detect OOG?

Even if a printer profile contains sufficient info for PS to construct a complete color space, OOG detection does not necessarily mean that the problem colors are outside the paper's gamut. All we know is that that particular profile cannot map the problem colors from Adobergb1998 to a color space created by the profile.

Along this line of thought, I too wonder if a profile can be created to "massage the gamut compression / gamut mapping so that deep reds and yellows can be obtained when the original colors are out of gamut".
Logged
Chris_T
Sr. Member
****
Offline Offline

Posts: 541


« Reply #12 on: January 18, 2006, 12:54:41 PM »
ReplyReply

There are plots like the following that let you see how a profile is mapping color spaces:

http://www.curvemeister.com/downloads/profileplotter/

After plotting several canned profiles from different vendors on similar papers (say all matte), I notice that the plots look very similar. The implication is that there won't be much noticeable differences between these profiles on any one of the similar papers.

However, plots of Bill Atkinson's (matte) profiles do look quite different, especially along the shadow ends.
Logged
Mark D Segal
Contributor
Sr. Member
*
Offline Offline

Posts: 6818


WWW
« Reply #13 on: January 18, 2006, 04:01:26 PM »
ReplyReply

There would appear to be a certain amount of confusion here. Vary one thing at a time and see what happens. Start by selecting a colour space and stick with it. Then soft proof EEM versus RC paper by loading the correct printing profile for the paper you are soft-proofing. A printer profile is paper-specific and corrects for differences between the numbers in the file and what the printer would produce from those numbers on its own without the profile. If your monitor calibration and profile are good and your soft proofing is properly set-up, this comparison should show you on the monitor what differences in colour rendition to expect between one type of paper and another, using the same colour working space. Now change the colour working space to something else, let the same file be interpreted according to that new colour working space, and then repeat the same soft-proofing for EEM versus RC paper. Now you have a consistent set of four soft-proof conditions comparing one paper versus another for each of the two colour working spaces. Having done all that systematically you'll have a quite clear impression of which combination of paper and working space best solves the gamut problem for the same image.
Logged

Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....." http://www.luminous-landscape.com/reviews/film/scanning_workflows_with_silverfast_8.shtml
Micheal_Kelly
Newbie
*
Offline Offline

Posts: 3


« Reply #14 on: January 19, 2006, 01:40:16 AM »
ReplyReply

Eric,
I use a free program from http://colorlab.hig.no/ICC3D/ called ICC3D. It is a really great University project that has little documentation. I found it worth the effort to figure it out. It will allow you to load an image with it's color space profile embedded or specified separately. Then you can load the profile of the paper you want to compare your image to. It will then plot a 3D graph of your image with the limits of the color gamut of the paper overlaid and you can rotate the image around in three space.

This lets me actually see where the printer gamut is falling down. Next it has 9 algorithms to "massage" the colors of your image into the color gamut of the paper/ink combo. You can then again see by rotating in 3d space how well the particular algorithm did at mapping the colors into the paper/ink gamut. I find it takes me out of "blind guessing" into understanding where I need to go to get an image to print as I want it to or to just give up trying and not waste time. Often I have to try all 9 to find one that works well and that is tedious but if you really want to print that out-of-gamut image it can be worth the effort.

Good Luck
Logged
PeterLange
Guest
« Reply #15 on: January 20, 2006, 11:29:03 AM »
ReplyReply

Quote
I use a free program from http://colorlab.hig.no/ICC3D/ called ICC3D. It is a really great University project that has little documentation. I found it worth the effort to figure it out. It will allow you to load an image with it's color space profile embedded or specified separately. Then you can load the profile of the paper you want to compare your image to. It will then plot a 3D graph of your image with the limits of the color gamut of the paper overlaid and you can rotate the image around in three space.

This lets me actually see where the printer gamut is falling down. Next it has 9 algorithms to "massage" the colors of your image into the color gamut of the paper/ink combo. You can then again see by rotating in 3d space how well the particular algorithm did at mapping the colors into the paper/ink gamut. ...


Wow!

This weekend is for testing.

I always had the hope that solutions will come up or become known, if just said problem would be more often expressed and discussed.

Thanks so much for posting this link.

Peter

--
Logged
madmanchan
Sr. Member
****
Offline Offline

Posts: 2101


« Reply #16 on: January 20, 2006, 02:15:56 PM »
ReplyReply

I just noticed something odd.

I'm printing with ImagePrint 6 Lite for my Epson 2200 and for the most part have been using the "version 6" profiles that Colorbyte provides.  But in some cases the v6 profiles aren't available yet, so Colorbyte says to use the older 5.x profiles, still available on their site (color only, not gray).  

After noticing the red/yellow issue with many of my images (see my original posts), I downloaded both v5 and v6 profiles for my frequently used papers, Ilford Smooth Pearl, Epson Premium Luster, and Moab Entrada.  I performed a 3D plot of both gamuts and was very surprised to find that v5 gamut is significantly larger (and completely encloses) the v6 profile's gamut.  I repeated these steps for several other paper profiles and found the same.

Anybody else notice this?  I'm quite puzzled by this ...

Eric
Logged

Brian Gilkes
Sr. Member
****
Offline Offline

Posts: 431


WWW
« Reply #17 on: January 20, 2006, 02:39:09 PM »
ReplyReply

Thanks from me too Micheal, I concur with Peter, this could be another Great Leap Forward.
I'll certainly carefully check it out.
Mark, you could see if Bill has a profile for Archival Matte ,and try that when using Enhanced Matte.. Although I haven't tested it,  as I don't use those papers, I have heard they are very similar.
None of these canned profiles will match a custom built one made by an expert-not just someone with a spectro. and some software. The uniformity between the 800 series printers does mean that canned profiles, especially Bills are pretty good.
Cheers
Brian
www.pharoseditions.com.au
Logged
Mark D Segal
Contributor
Sr. Member
*
Offline Offline

Posts: 6818


WWW
« Reply #18 on: January 20, 2006, 02:49:42 PM »
ReplyReply

Brian, I am using an Epson 4800. Bill Atkinson does not have a matte profile for this printer or any of the other Epson K3 inkset printers yet. And yes, you are right, they are calibrated out of the factory to a common standard very carefully, so the difference a custom profile makes is extremely slight for this series of printers. I've seen conflicting evidence about the difference IP makes to the separation of deep tones at the lower end of the luminosity grey scale.
Logged

Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....." http://www.luminous-landscape.com/reviews/film/scanning_workflows_with_silverfast_8.shtml
colourperfect
Jr. Member
**
Offline Offline

Posts: 75


WWW
« Reply #19 on: January 27, 2006, 01:24:08 PM »
ReplyReply

I always turn softproof on, turn gamut warning on and then use a hue / sat adj layer, choosing the appropriate colour, to reduce saturation down until this gamut warning grey has mostly disappeared.

You can use the + and - droppers to tune colours which are on the border.

seems to work well and gets the maximum use from your profile gamut

Ian

http://www.colourperfect.co.uk
Logged
Pages: [1] 2 »   Top of Page
Print
Jump to:  

Ad
Ad
Ad