Ad
Ad
Ad
Pages: [1]   Bottom of Page
Print
Author Topic: Does the Epson 4000/7600/9600 see red?  (Read 7584 times)
Mark D Segal
Contributor
Sr. Member
*
Offline Offline

Posts: 7121


WWW
« on: September 14, 2005, 09:32:23 PM »
ReplyReply

I started this thread just before the Board crash, and it seems to have gotten lost, so apologies to those of you who may have seen it already. I think I have finally run up against a gamut constraint on my Epson 4000 with some shots from the Chinatown Festival in Bangkok Thailand. When I soft-proofed an RGB 170-24-29 it turned from in-your-face red to kind of purplish, and of course printed likewise. This was on Enhanced Matte paper, but behaved the same for all the coated Epson papers as well. I then went to compare the test print stuff I did between the 4000 and the 2400 and noticed that the K3 inks would probably have been a bit more successful printing closer to this hue. Have any of you run up against serious gamut constraints with Ultrachrome inks? Have you found any work-around - a way to trick the printer somehow? (I tried stuff in LAB, HSB, etc. no avail - it seems the gamut is the gamut.)
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
Jonathan Wienke
Sr. Member
****
Offline Offline

Posts: 5759



WWW
« Reply #1 on: September 14, 2005, 11:51:28 PM »
ReplyReply

Can you post an example image? That would make it a lot easier to tell what's going on, as RGB values are meaningless outside the context of a color space.
Logged

Mark D Segal
Contributor
Sr. Member
*
Offline Offline

Posts: 7121


WWW
« Reply #2 on: September 15, 2005, 08:45:46 AM »
ReplyReply

Hi Jonathan,

Thanks for responding to this, and yes you are absolutely right - I should have mentioned the color space - it is ProPhoto. Perhaps also helpful would be the Lab and HSB values: Here goes: L 55, a 116, b 52; H 358, S 88, B 86; from that you can see how "far-end" these colours are.

Unfortunately, there is some kind of hyper-paranoid security policy with my ISP (one of the foremost ISPs in Canada) that disallows me from linking a download to a web-forum of images I would post on a website which they service. I may solve this problem by opening a pBase or some such account , but not done it yet - so bottom line: no, unfortunately I can't provide an image - but I can send you a downsized JPG by private email if you'd like to have a look. Please let me know.
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 #3 on: September 15, 2005, 04:17:41 PM »
ReplyReply

Posting the colour may not be of much use as unless one is using a NEC Specra View os the new Eizo (the 220?) with a viewable colourspace around AdobeRGB (1998) .
I did notice that at a trade show a few montha ago that Epson had selected images with saturated reds to show how the new K3 inkset was superior to the old Ultrachrome.Flesh tones and some blue-greens look better too.
Despite this, better reds may be obtainable with an edited profile.
HTH
Brian
Pharos Editions
Logged
Mark D Segal
Contributor
Sr. Member
*
Offline Offline

Posts: 7121


WWW
« Reply #4 on: September 15, 2005, 04:28:09 PM »
ReplyReply

Interesting idea Brian; I don't have the kit needed to make and edit printer profiles, but perhaps I should pull some profiles from my "library" and try several for size - see how they cope with red. Otherwise I suppose I could hook-up with a profiler with whom I could work to give the reds a boost.
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
Jonathan Wienke
Sr. Member
****
Offline Offline

Posts: 5759



WWW
« Reply #5 on: September 15, 2005, 07:34:05 PM »
ReplyReply

You can mail me a reasonably-sized image to jonathan@visual-vacations.com.
Logged

drew
Sr. Member
****
Offline Offline

Posts: 477



WWW
« Reply #6 on: September 16, 2005, 09:04:47 AM »
ReplyReply

I am probably missing something here, but if the colour change occurred on soft-proofing, then I fail to see how the printer can be at fault.
Logged

Andrew Richards My Webpage
Mark D Segal
Contributor
Sr. Member
*
Offline Offline

Posts: 7121


WWW
« Reply #7 on: September 16, 2005, 09:21:08 AM »
ReplyReply

Drew, one or both of us may be missing something, but soft-proofing, as far as I understand it, is intended to show on the monitor how colours and tonality will reproduce in print, once you give it the appropriate printer profile. I am working in ProPhoto which is a huge colour space. So if there is a big disconnect between what you see on the monitor without soft proofing versus what you see with soft proofing, it tells you that the printing process is imposing gamut constraints on certain colours, which in this particular case happened to be unusually severe. Hence the purpose of my post is to see whether other people have had similar experience, or can replicate the problem on their Ultrachrome-Epsons, and pehaps have devised some kind of work-around to compensate for the limitation somehow.
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
Jonathan Wienke
Sr. Member
****
Offline Offline

Posts: 5759



WWW
« Reply #8 on: September 16, 2005, 09:54:50 AM »
ReplyReply

Can you send me the RAW? The crop you sent me already exhibits magenta clipping before any soft-proofing.
Logged

Mark D Segal
Contributor
Sr. Member
*
Offline Offline

Posts: 7121


WWW
« Reply #9 on: September 16, 2005, 10:36:12 AM »
ReplyReply

Jonathan,

This is not a digital image. It was scanned from a colour negative in a Minolta Dimage ScanElite 5400 film scanner using ProPhoto colour space. The original scan was then processed in PSCS2. There is an embedded curve adjustment to bring the overall image contrast to what I wanted with the soft-proofing activated. So it is quite possible that the image has come clipping in it. For diagnostic purposes I would be quite happy to delete the curve and send you what would then be a "RAW" version to the extent any such thing exists from this scanning process. Please let me know.

Thanks for the interest and regards,

Mark
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
Mark D Segal
Contributor
Sr. Member
*
Offline Offline

Posts: 7121


WWW
« Reply #10 on: September 16, 2005, 01:48:34 PM »
ReplyReply

Jonathan,

Following up on my last post, I went back into that photograph, turned off the Curves adjustment layer, and indeed that wiped out the clipping you mention. In fact, I should not have clipped it at all in the first place - it was a bit of poor curving on my part. I can achieve the desired contrast under soft proof without clipping. That improvement, however, has no impact on the basic problem.

So now the interesting part: I once again decided to have another "GO" at those reds in Photoshop, with something pedestrian - a "Selective Color" adjustment layer. Leaving soft proof activated, I went into RED, decreased Magenta by 37%, increased yellow by 28% and increased Black by 9%. These three tweaks didn't produce a soft-proofed red that truly resembles the hue of Chinatown red, but it dramatically reduced the purple to an extent that the overall effect is somewhat more realistic. I'd still like to be able to do better.
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
Jonathan Wienke
Sr. Member
****
Offline Offline

Posts: 5759



WWW
« Reply #11 on: September 16, 2005, 08:52:01 PM »
ReplyReply

Perhaps it's being clipped in the scanner then. Try sending a similar crop straight from the scanner with no tweaking done.
Logged

neil snape
Guest
« Reply #12 on: September 17, 2005, 04:00:05 AM »
ReplyReply

The numbers of the red you sent are quite saturated. What happens in relative rendering will be the red of the larger space (Kodak ROM) to the nearest max gamut boundary in the printer space. Sometimes the shift can go in a more or less straight line along the hue angle, sometimes not. When it doesn't it can have quite severe clipping to colors that doe  not look much like the larger gamut. On perceptual the profile maker will have tried to map the colors visually to points of well adapted shifts, usually in the case of a red to a less saturated but equal L value. What is not usually a problem may be the reference gamut of the profile creator package. In this case it may be as the ROM space is unweildly with certain colors.
I've yet to find a monitor that correctly displays say a Coke can red with tonal separation. This is yet another element on soft proofing a printer red on screen , in particular to Epsons, yes even the K3's that do not produce reds that well in the saturated and three quarter down shades.
Logged
Mark D Segal
Contributor
Sr. Member
*
Offline Offline

Posts: 7121


WWW
« Reply #13 on: September 17, 2005, 05:20:28 AM »
ReplyReply

Jonathan - thanks - I'll do that over this weekend and send it on to you.

Niel - I'm using a Dell (Sony Trinitron) CRT - does a pretty good job displaying red, and the soft proof is really a reliable indicator of what will come out of the printer. What you are saying about gamut mapping sounds plausible to me - I think the question may be to what extent one can improve it or implement work-arounds.
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
Pages: [1]   Top of Page
Print
Jump to:  

Ad
Ad
Ad