Ad
Ad
Ad
Pages: « 1 ... 3 4 [5]   Bottom of Page
Print
Author Topic: Please help!! Very Urgent! Epson 9880 colours blocking up  (Read 13726 times)
Farmer
Sr. Member
****
Offline Offline

Posts: 1629


WWW
« Reply #80 on: September 16, 2011, 12:32:58 AM »
ReplyReply

Might be relevant:

http://theonlinephotographer.typepad.com/the_online_photographer/2011/09/kinda-interesting.html

http://theonlinephotographer.typepad.com/the_online_photographer/2011/09/when-good-profiles-turn-bad.html
Logged

jc1
Jr. Member
**
Offline Offline

Posts: 90



WWW
« Reply #81 on: September 16, 2011, 12:40:08 AM »
ReplyReply


Thanks for the links.



Clarification:

My previous test is very general and  may not reflect the color rendering capability of the profiles.
Need to perform color space conversion test.

Any suggestion?

Regards,
jc
Logged
Ernst Dinkla
Sr. Member
****
Offline Offline

Posts: 2806


« Reply #82 on: September 16, 2011, 03:00:31 AM »
ReplyReply


Which is also discussed at the LL CM forum:

http://www.luminous-landscape.com/forum/index.php?topic=57588.20

Remains the question in Ctein's case why the original smaller image/file did get the right colors with Photoshop's CM on + the normal profile instead of using the printer's CM. The difference between the two files must have been in the image editing for the larger size, a step that was sketched as a routine procedure. Interesting to see that Mike Chaney came to the rescue, there was another issue last week with Jpeg export changes in the Photoshop 12-04 upgrade that he analysed and at least solved for Qimage Ultimate users. As written before it would have been nice if this thread had been transferred to the Qimage forum too.


met vriendelijke groeten, Ernst

Try: http://groups.yahoo.com/group/Wide_Inkjet_Printers/





Logged
jc1
Jr. Member
**
Offline Offline

Posts: 90



WWW
« Reply #83 on: September 16, 2011, 03:44:50 AM »
ReplyReply

Gamutvision Hue Contour Test

The result is presented graphically as follow.
 

 
where
Profile1: custom profile by nick
Profile2: profile generated with PM5, with measured data supplied by nick
Profile3: profile generated with another profiler, with measured data supplied by nick
Profile4. profile generated with  the profiler used to create Profile3, with measured data
           derived from AtoB of Profiler1.

Imperfect Profile
From the above result and my other observations (not shown), I have good reasons to suggest that Profile1 is an imperfect profile. 
 
jc
 
Logged
jc1
Jr. Member
**
Offline Offline

Posts: 90



WWW
« Reply #84 on: September 16, 2011, 10:02:03 AM »
ReplyReply

Color Reproduction with Profile1

The original test image was cropped and sampled digitally, total of 60x80 test patches.



The digital sample (source image, in sRGB color space ) was read into CS, converted to printer space with Profile4 and Profile1 respectively. The converted images were displayed in Lab space as shown below.



Observations:
The top grapher plot (Green color points) shows the converted color points with Profile4, similar result with Profile3 can be obtained.

The bottom grapher plot (Magenta color points) shows the converted color points with Profile1 and with Perceptual Intent, similar result could be obtained if conversion was carried out with RelCol.

With Profile1, the conversion from sRGB to printer color space is "over saturated", as the converted color points were expanded to the printer gamut boundary. By right, the in-gamut color with RelCol by definition, should stay put or should have little effect due to the conversion.

The "over-saturated" color coupled with the imperfect hue gradient near the gamut boundary, give rise to the color reproduction issue.

jc
Logged
NickCroken
Full Member
***
Offline Offline

Posts: 112


« Reply #85 on: September 16, 2011, 10:44:58 AM »
ReplyReply

Based on this it appears that it is indeed the profile making software that I'm using.  I have printed the image before with perfect colour/gradient although that is now appearing to be a fluke.  Most of my images don't have blues like that in them so I can see why it would take so long for this issue to show up.  I'm anxiously awaiting a response from David Miller at Datacolor.
Logged
NickCroken
Full Member
***
Offline Offline

Posts: 112


« Reply #86 on: September 16, 2011, 11:01:51 AM »
ReplyReply

David's response just came in, it is below in blue. Have any of you had any issues using photo black for printing on epson premium matte canvas before?  The friend I bought the printer off of never had issues with it and from what I've read online, it shouldn't be an issue.

Again: sorry about the delay.  There are possibly two different things going on here.

First, there's the target print itself on canvas, which has an unusually light black. You're getting L in the 26 range, which means there just isn't as much density in the shadows as there would normally be. For comparison (I have an Epson 7900 and I haven't done any work on canvas, myself, but I have other measurement files from customers that have used canvas in the past) As with fine art matte papers, canvas typically has measured L values for black in the upper teens, not the mid-to-upper 20's. (Two different canvas measurement files that I have for Epson printers; and one of this is an Epson 4000 with the much older Epson inkset; has L = 17 for the black patch) The result is a much more limited gamut; the difference in gamut volume going from an L in the high teens to L around 26 is typically a 30% decrease in overall gamut volume, which is not a great thing.

To improve this: there's nothing you can do about the canvas itself, and you're using Epson inks, so you can't do better than that. The only question is: are you -sure- that you're using a media settings that's going to give you matte black ink? (The measurement file is noted with "Radiant Paper White" as the media setting, which would give you the correct black ink). IF you were using a media setting for a glossy/luster paper by mistake, it would explain the weak black and shadows. (Typically this is what happens when people try to avoid switching between photo and matte black in older Epsons, like the 4800/80 and probably also the 9800/80, and they end up printing with photo black ink on matte papers - which will always lead to poor results/weak blacks, no matter what Epson would have them believe) If you -are- using matte black ink, then there may be other media settings that will give you darker blacks, more density, and the gamut would improve.

So with all that said, I did some testing with your sample of the image in Photoshop, to see if I'd see the same banding in the softproof that you're getting in the physical print. If I leave black ink simulation turned off: no, I'm not seeing it. If I turn it on, then interestingly enough, I'm seeing stratification in those areas that's similar to what you're getting in the prints, although still not quite as bad. So that still tends to point to the limited gamut you're getting in your target print as potentially being the cause for this.

The second possibility (by itself, or in combination with the above) is the number of patches. You're using the 225 patch target, and if the response of the printer on canvas isn't sufficiently modeled by that, (if it behaves unpredictably enough "between" the samples provided by those patches) then the profile won't give you accurate results for the colors that fall "between the cracks" and the softproof also won't be as indicative of what would be going on when you actually print through the profile. In general, I'd probably have recommended using more patches when profiling canvas, in any case. So I think the other thing you should do is print and measure the 729 patch target (use the single page target, and print it large enough so that you'll get decent sized patches to measure) and start using that instead of the 225.

So for the next round of tests, my recommendation: (and then we'll see what happens):

- Make sure you're using matte black (if for whatever you're using photo black, switch to matte black)

- You "should", I hope, be able to make a target print where the black measurement has a lower L value, hopefully, much lower. I'd be looking to see 20 or less.

- Print and measure the 729 patch target; attach a new measurement file to the ticket; and also do a test print.

One other thing (I'm sure you've got Photoshop set correctly for printing, but just in case):

- What's your Photoshop RGB Working Space set to?

- In the Photoshop Print dialog, you're using Document (not Proof) when printing, correct?
Logged
davidh202
Sr. Member
****
Offline Offline

Posts: 546


« Reply #87 on: September 16, 2011, 04:13:27 PM »
ReplyReply

I recently used up my PCM and am currrently using Ex Canvas Matte...
Nick,
I do have the PCM spec sheet and this is what it says...

"Matte Black Or Photo Black may be used on this canvas.Matte Black will yield a higher D-Max but will easily scuff. Use Matte Black only if you are going to using a protective spray on the canvas. Use Photo Black if you are going to put the print behind glass without the use of a protective spay"

Ambiguous instructions I'd say ;-)
 I never had a problem with Photo Black.
David

 All said and done I think Mike at Qimage was right though, and deserves a follow up on your post in the Qimage forum ;-)
David
Logged
NickCroken
Full Member
***
Offline Offline

Posts: 112


« Reply #88 on: September 19, 2011, 09:36:47 AM »
ReplyReply

Fear not David, I'll be creating a summary post as soon as this is completely fixed and publishing it online and in the forums.  I feel that if I have had this issue than there are other folks out there experiencing the same thing with equal frustration.
Logged
Pages: « 1 ... 3 4 [5]   Top of Page
Print
Jump to:  

Ad
Ad
Ad