Ad
Ad
Ad
Pages: [1]   Bottom of Page
Print
Author Topic: Please help with rich blacks in Epson 3800  (Read 4930 times)
reactivesource
Newbie
*
Offline Offline

Posts: 6


« on: July 03, 2009, 01:40:36 AM »
ReplyReply

Please Help! I am trying to achieve rich blacks on enhanced matte paper off of my 3800. I cant seem to get the shadow detail I use to get with my 2200 using the matte driver, matte ink and matte PK profile. Obviously the old profile was operating on the assumption I was using photo ink but, I was using matte ink and my blacks were really popping and the detail in the shadows was awesome. I have been working with Chromix for a profile but no luck in achieving this. If anyone has any advice or can offer a profile, I will gladly pay. Attached is an image of the 3800 Chormix deep color profile and the 2200 matte pk profile.



Brett Munoz
Logged
xilvar
Newbie
*
Offline Offline

Posts: 12


« Reply #1 on: July 03, 2009, 02:58:51 AM »
ReplyReply

Hm. a few things you might try:

1) print a non-color managed black square of 0,0,0 to see how dark the 3800 prints under whatever paper setting you're using. Maybe try a few paper settings.
2) assuming its dark enough have a profile made. I've only had my isis for about a week but I could give it a try.
3) I've found that both my colormunki and my isis read glossy super dark regions better than matte. It seems that they can't tell the difference between ultra-dense matte colors that my naked eye can. My solution is that for some of my profiles I coat the test chart prints even if I dont intend to ever coat final prints. This seems to solve the problem of blacks blocking up for me.

Presumably the reason #3 works is because light scatter is reduced allowing the instrument to properly measure the difference between blacks which would normally be obscured (at least to the instrument) by light scatter.

I complained loudly to x-rite for a while about this problem and they claim to have fixed it in the colormunki software released last week. Interestingly it does in fact appear to be working better now even without the coating. My suspicion is that they 'fixed' it by throwing away the ultra-dark readings in some cases and just using a heuristic to guess how they should approach 0,0,0. I bet that previously the generated profile never reached 0 because it measured say 16 to be just as dark as 0.

george


Quote from: reactivesource
Please Help! I am trying to achieve rich blacks on enhanced matte paper off of my 3800. I cant seem to get the shadow detail I use to get with my 2200 using the matte driver, matte ink and matte PK profile. Obviously the old profile was operating on the assumption I was using photo ink but, I was using matte ink and my blacks were really popping and the detail in the shadows was awesome. I have been working with Chromix for a profile but no luck in achieving this. If anyone has any advice or can offer a profile, I will gladly pay. Attached is an image of the 3800 Chormix deep color profile and the 2200 matte pk profile.



Brett Munoz
Logged
Mark D Segal
Contributor
Sr. Member
*
Offline Offline

Posts: 6945


WWW
« Reply #2 on: July 03, 2009, 07:17:35 AM »
ReplyReply

Quote from: reactivesource
Please Help! I am trying to achieve rich blacks on enhanced matte paper off of my 3800. I cant seem to get the shadow detail I use to get with my 2200 using the matte driver, matte ink and matte PK profile. Obviously the old profile was operating on the assumption I was using photo ink but, I was using matte ink and my blacks were really popping and the detail in the shadows was awesome. I have been working with Chromix for a profile but no luck in achieving this. If anyone has any advice or can offer a profile, I will gladly pay. Attached is an image of the 3800 Chormix deep color profile and the 2200 matte pk profile.



Brett Munoz

All of this is covered expertly and extensively in Eric Chan's 3800 webpage, which you can access here: Eric Chan's 3800 Printer Noptes and Resources. Also, if you do a search on this website (Luminous-Landscape Forums) you find much related material.
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
reactivesource
Newbie
*
Offline Offline

Posts: 6


« Reply #3 on: July 03, 2009, 01:21:41 PM »
ReplyReply

Quote from: MarkDS
All of this is covered expertly and extensively in Eric Chan's 3800 webpage, which you can access here: Eric Chan's 3800 Printer Noptes and Resources. Also, if you do a search on this website (Luminous-Landscape Forums) you find much related material.

I looked and could not find anything on rich blacks where the matte black is mixed with cyan, magenta and yellow to produce a deep black.
Logged
reactivesource
Newbie
*
Offline Offline

Posts: 6


« Reply #4 on: July 03, 2009, 01:23:46 PM »
ReplyReply

Quote from: xilvar
Hm. a few things you might try:

1) print a non-color managed black square of 0,0,0 to see how dark the 3800 prints under whatever paper setting you're using. Maybe try a few paper settings.
2) assuming its dark enough have a profile made. I've only had my isis for about a week but I could give it a try.
3) I've found that both my colormunki and my isis read glossy super dark regions better than matte. It seems that they can't tell the difference between ultra-dense matte colors that my naked eye can. My solution is that for some of my profiles I coat the test chart prints even if I dont intend to ever coat final prints. This seems to solve the problem of blacks blocking up for me.

Presumably the reason #3 works is because light scatter is reduced allowing the instrument to properly measure the difference between blacks which would normally be obscured (at least to the instrument) by light scatter.

I complained loudly to x-rite for a while about this problem and they claim to have fixed it in the colormunki software released last week. Interestingly it does in fact appear to be working better now even without the coating. My suspicion is that they 'fixed' it by throwing away the ultra-dark readings in some cases and just using a heuristic to guess how they should approach 0,0,0. I bet that previously the generated profile never reached 0 because it measured say 16 to be just as dark as 0.

george

I can produce good blacks, I am just having issues producing rich blacks where MK is mixed with CMY to produce a deep black.  
Logged
xilvar
Newbie
*
Offline Offline

Posts: 12


« Reply #5 on: July 06, 2009, 12:55:21 AM »
ReplyReply

ah. my apologies. i didnt recognize rich black as a term

the 2200 will print additional cmy based upon profile even beyond 100% limit on black channel using the standard driver?

somehow i would have expected that the absolute ink limit would be hit just at 100% k. i'll have to try that myself

two ideas for the 3800:
1) increase the ink limit to maximum in driver and then use the profile to throttle back. (incorporate the gloss coating before chart read suggestion i made earlier)
2) use a rip so you can just directly set your per channel ink limits and base luts then again profile (presumably in cmyk mode)

xilvar
Logged
jjlphoto
Sr. Member
****
Offline Offline

Posts: 467


« Reply #6 on: July 06, 2009, 01:18:46 PM »
ReplyReply

Quote from: xilvar
1) increase the ink limit to maximum in driver and then use the profile to throttle back.

I hadn't considered that. I too am bummed out by how dead my B&W prints look on PK ink papers. I even tested Museo Max, which is supposed to be have the greatest D-Max of the PK ink papers. A little better than the Hahnemuhles, but still weak.

Where do I find the ink limit function? Closest thing I could find is in Paper Config, there is a Color Density, -50% to +50% with the default in the middle.
Logged

Thanks, John Luke

Member-ASMP
xilvar
Newbie
*
Offline Offline

Posts: 12


« Reply #7 on: July 06, 2009, 05:14:10 PM »
ReplyReply

Yeah, that's the right setting. Epson gives it a slightly different name every now and then. It isn't an absolute ink limit of course, but a relative limit to whatever the paper type is set to. I vaguely recall that Glossy has the highest default ink limit then Lustre, then Semigloss. I could be wrong though...

On costco glossy for example I can actually set my 2200 to +15% and still have the black I'm using fully dry before reaching the rollers. There's a significant increase in gamut and black depth when I profile. Now I just need to find the time to print the 4096 target again at +15%, gloss coat it, and run it through the isis.

Incidentally, I get my black a bit differently than most everyone else. I just always use matte black. Even on glossy. I like it better, and a light spray of clearjet over the top fixes up the gloss differential when I want to get rid of it.

xilvar

Logged
jjlphoto
Sr. Member
****
Offline Offline

Posts: 467


« Reply #8 on: July 06, 2009, 06:59:46 PM »
ReplyReply

Quote from: xilvar
I vaguely recall that Glossy has the highest default ink limit then Lustre, then Semigloss. I could be wrong though...

How about when running PK ink?

I'm assuming you print your profile targets with the ink limit cranked up to +15%?
« Last Edit: July 06, 2009, 07:02:08 PM by jjlphoto » Logged

Thanks, John Luke

Member-ASMP
xilvar
Newbie
*
Offline Offline

Posts: 12


« Reply #9 on: July 06, 2009, 11:02:20 PM »
ReplyReply

I believe my driver is using the mk ink as if it were pk. The 2200 windows drivers weren't very picky about which ink was in the printer, so you could print anything on anything. The newer printers you have to manually tell them they have an ink that they don't think they have, but I believe its still possible.

Anyway, what I said about ink limit being highest on glossy is what the driver is programmed to, not necessarily what is true in reality.

I'm not sure how far PK can actually be pushed on these printers and papers, because I stopped using it entirely. I hated switching, and I had more problems with drying time using PK than MK.

Yep, for the profiles I've been playing with this on I've printed the targets at +15. I originally decided on +15 by using this image: http://stabilize.net/users/gpang/limit_check.png

Basically I cranked up to +20%, printed that image with no color management, opened up the top of the printer and watched the squares going by. I noticed that at +20% the full black square was still slightly wet even after exiting the printer. So I reduced to +15 and tried again. The squares are wide like they are so that they will all be run over by any exit rollers/pizza wheels in increasing density order. Thus if an ink is REALLY wet the rollers will leave a trail in the white bordering areas.

xilvar
Logged
jjlphoto
Sr. Member
****
Offline Offline

Posts: 467


« Reply #10 on: July 07, 2009, 08:27:47 AM »
ReplyReply

Quote from: xilvar
I believe my driver is using the mk ink as if it were pk. The 2200 windows drivers weren't very picky about which ink was in the printer, so you could print anything on anything. The newer printers you have to manually tell them they have an ink that they don't think they have, but I believe its still possible......

xilvar

Okay, I see. On the 3800, both PK and MK are always installed. When you select a glossy media type in the driver, the printer uses PK. When you select one of the matte media types, the printer uses PK. No way to change that.


Logged

Thanks, John Luke

Member-ASMP
Mark D Segal
Contributor
Sr. Member
*
Offline Offline

Posts: 6945


WWW
« Reply #11 on: July 07, 2009, 08:59:15 AM »
ReplyReply

Quote from: jjlphoto
Okay, I see. On the 3800, both PK and MK are always installed. When you select a glossy media type in the driver, the printer uses PK. When you select one of the matte media types, the printer uses PK. No way to change that.

That's right - but there's an added factor: the print head is still an eight channel head; there is no ninth channel for the additional black. Therefore the printer does an ink-switching routine when you swith media type between those supported by MK and those supported by PK. For the 3800, this switching routine is pretty seamless and the round-trip only consumes about 4 or 5 ml of ink - very cheap compared with its predecessot - the 4800.

The printer does not normally allow, and Epson does not recommend (based on its extensive testing) using MK ink on gloss surfaces or PK ink on matte surfaces, though the latter is likely to be less troublesome than the former (expect smearing).

As for DMax, however you slice it, in general matte surfaces cannot achieve the DMax of gloss surface. For example DMax on Epson Enhanced Matte (or whatever they call it now) is about 1.8. DMax on Ilford Gold Fibre Silk is around 2.4. This makes a huge difference in the achievavle depth of black and therefore tonal gradation within the blacks one can achieve with any printer or profile.

The 3800 should be able to deliver about the highest quality performance available for rendition of the lowest quadrant of the tonal scale on any pape without fiddling with ink limits (which can have other unintended side-effects). The post-processing work you do UNDER SOFTPROOFING to bring out shadow detail and the quality of your profile are most likely the key factors to achieve top notch results.
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
jjlphoto
Sr. Member
****
Offline Offline

Posts: 467


« Reply #12 on: July 07, 2009, 09:51:50 AM »
ReplyReply

Quote from: MarkDS
That's right - but there's an added factor: the print head is still an eight channel head; there is no ninth channel for the additional black. Therefore the printer does an ink-switching routine when you swith media type between those supported by MK and those supported by PK. For the 3800, this switching routine is pretty seamless and the round-trip only consumes about 4 or 5 ml of ink - very cheap compared with its predecessot - the 4800.


That's understood. The only reason MK and PK came up was that one of xilvar's reply indicated he was doing something unconventional with his 2200 by selecting a glossy stock for a customized ink delivery amount, but having MK ink installed instead of PK. Maybe I misunderstood him. I do question his suggestion that Glossy media setting has the highest ink default limit. I would think the opposite would be true as too much ink on glossy stock will instantly puddle up, where-as matte stocks absorb so much, you have to have a lot of ink laid down to get good tonality.
Logged

Thanks, John Luke

Member-ASMP
Mark D Segal
Contributor
Sr. Member
*
Offline Offline

Posts: 6945


WWW
« Reply #13 on: July 07, 2009, 10:23:02 AM »
ReplyReply

Quote from: jjlphoto
That's understood. The only reason MK and PK came up was that one of xilvar's reply indicated he was doing something unconventional with his 2200 by selecting a glossy stock for a customized ink delivery amount, but having MK ink installed instead of PK. Maybe I misunderstood him. I do question his suggestion that Glossy media setting has the highest ink default limit. I would think the opposite would be true as too much ink on glossy stock will instantly puddle up, where-as matte stocks absorb so much, you have to have a lot of ink laid down to get good tonality.

I don't think you misunderstood him - or if you did, that makes two of us. Again, IMHO. fiddling around with ink is not necessarily the optimal approach for achieving the OP's objectives.
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: 6945


WWW
« Reply #14 on: July 07, 2009, 11:08:57 AM »
ReplyReply

Another factor I would mention here is to check whether the target black value is correct for the printer and paper combination in order to achieve the fullest tonal range available from the printer. This procedure is described in detail in Tim Grey's "Color Confidence" pages 198 and 199.
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
xilvar
Newbie
*
Offline Offline

Posts: 12


« Reply #15 on: July 07, 2009, 08:53:12 PM »
ReplyReply

Just for the record I think some people missed the same terminology reference the OP made which I missed.

For the purposes of this discussion:
'Rich Black' == 'Using 100% black PLUS additional percentages of C, M, Y, etc. to create a black which looks deeper than usual'

On lasers for example it could be 400% black where 100% is used of all 4 colors.

On inkjet total physical ink limits prevent that of course.

I would imagine to achieve more ink on page than the usual 0,0,0 you would pretty much have to alter the ink limit
OR
use a RIP where you can just tell it exactly what to do.

Yes, I'm unsure what order the media settings are in, but based on the relatively scanty references I can find on google it seems like glossy is actually highest on at least some printers.

xilvar
Logged
neil snape
Sr. Member
****
Offline Offline

Posts: 1432


WWW
« Reply #16 on: July 08, 2009, 01:48:11 AM »
ReplyReply

Thing is you're sending an rgb data stream to the printer where the ASIC divides it up into it's bit map as needed. You are not the one controlling the ink color maps ( separations) nor are you controlling the black at the printing level.

If you increase the inking at the driver level your are fudging the ink limiting somewhat in certain regions. You are not able to however control the black in any significant way for creating a rich black. Some high end rips let you do this, but it is complicated as the balance will quickly destroy the linearity and gamut of the set up.
The most recent printers like the 9900 and HP Z printers use a very high GCR truly GCR as the grey component is replaced by large amounts of grey ink. The only way I can think of fudging a rich black is to print it twice but you need to carefully control registration....
Logged
xilvar
Newbie
*
Offline Offline

Posts: 12


« Reply #17 on: July 09, 2009, 05:02:09 PM »
ReplyReply

Incidentally, another suggestion is to play around with quadtone rip. It's a free/shareware rip and you can control all the channels directly. It only prints greyscale images, but you can use whatever channels you want.
Logged
Pages: [1]   Top of Page
Print
Jump to:  

Ad
Ad
Ad