Luminous Landscape Forum
Raw & Post Processing, Printing => Printing: Printers, Papers and Inks => Topic started by: artobest on May 03, 2010, 01:00:40 pm
-
Hi
I'm getting posterization printing black and white on Hahnemuhle Photo Rag Baryta with my Z3200ps. Initially I was using the canned Hahnemuhle profile, but the posterization was so atrocious I made a custom profile (using the HP APS). This gave improved results, but still obvious banding in linear gradients, and subtle but definite posterization in real-world images.
Can anyone suggest a solution? Or point me in the direction of a better profile? Maybe I should make a whole new preset using a different paper type?
Thanks for your help.
Peter
-
Hi
I'm getting posterization printing black and white on Hahnemuhle Photo Rag Baryta with my Z3200ps. Initially I was using the canned Hahnemuhle profile, but the posterization was so atrocious I made a custom profile (using the HP APS). This gave improved results, but still obvious banding in linear gradients, and subtle but definite posterization in real-world images.
Hi Peter,
Just asking the obvious, is the image free from posterization? What does the histogram look like near the troublesome brightnesses? What kind of Profile conversion is taking place (working space=?, Bit depth=?, rendering intent=?)?
Cheers,
Bart
-
double post deleted
-
Hi Peter,
Just asking the obvious, is the image free from posterization? What does the histogram look like near the troublesome brightnesses? What kind of Profile conversion is taking place (working space=?, Bit depth=?, rendering intent=?)?
Cheers,
Bart
Hi Bart
To answer your questions in order: No, the image is not posterized. I am printing both a photo and a test image with gradients, soft transitions etc.
The histograms for both image are dense and relatively smooth.
The space of the test image is Gray Gamma 2.2. The b&w photo is printing (somewhat redundantly) from my working space, which is ProPhoto. Bit depth of the photo is 16-bit, the test image is 8-bit. Rendering intent for both is relative colorimetric, but I would expect that changing that wouldn't make any significant difference, it being a b&w image and all ...
By the way, I've tried printing from both Qimage and Photoshop, with the same results.
P
-
Hi Bart
To answer your questions in order: No, the image is not posterized. I am printing both a photo and a test image with gradients, soft transitions etc.
The histograms for both image are dense and relatively smooth.
The space of the test image is Gray Gamma 2.2. The b&w photo is printing (somewhat redundantly) from my working space, which is ProPhoto. Bit depth of the photo is 16-bit, the test image is 8-bit. Rendering intent for both is relative colorimetric, but I would expect that changing that wouldn't make any significant difference, it being a b&w image and all ...
By the way, I've tried printing from both Qimage and Photoshop, with the same results.
Hi Peter,
1. Could you try assigning AdobeRGB to your test image after changing it's mode to RGB, and print again?
2. What happens when you convert your B&W image from Pro Photo RGB to Adobe RGB, do you see the posterization?
3. What happens if your use AdobeRGB for your B&W from the start, instead of Pro Photo RGB?
When it is caused by a destructive profile conversion, it might be less with Adobe RGB as a starting point, hence the suggestions.
Cheers,
Bart
-
I don't think you mentioned which paper preset you used, but the Baryta papers don't need and can't tolerate a very high ink load. I get good results with Fine Art Pearl (less ink), that only has an ink limit setting of 32, which is very little ink.
-
Hi Bart
I will try some of these options, although being black and white, the images aren't undergoing a terribly destructive conversion, especially since I'm using Relative Colorimetric intent. As an experiment, I tried sending an image to the printer using printer colour management and selecting sRGB as the profile in the printer, but it didn't really improve matters.
Colorwave, I am using the canned PR Baryta profile - the one that appears on both Hahnemuhle's and HP's websites. After that failed, I reprofiled the paper and got significantly improved results. I also tried adjusting the ink limit downwards, which did open up the shadows a little but did nothing for the linearity (I didn't reprofile for the lower limit, just recalibrated -should I reprofile?).
FWIW, soft proofing suggests the prints should come out fine - it shows perfect linearity with the canned profile, and only minor stepping on my custom one.
Perhaps I should make a linearising curve. What's the best way to do that manually?
Cheers, and thanks for the suggestions.
-
Well, two custom profiles, multiple ink limit adjustments and recalibrations and dozens of test prints (and half a roll of paper!) later, still no luck. Has anyone out there achieved linearity with black and white prints on Photo Rag Baryta? With any printer?
I'm not nit-picking here - my output is truly horrendous - really bad posterization that just gets worse at larger sizes. I've tried printing from PS and Qimage, with and without printer driver CM, switching Qimage's CM off altogether as recommended by Ernst Dinkla, changing embedded profiles on the images, trying RGB and Greyscale (gamma 2.2). Soft-proofing in PS doesn't help - it indicates great results from Hahnemuhle's canned profile, which is actually the worst.
I haven't seen this issue mentioned before ... and I'm starting to get really quite fed up.
-
Has anyone out there achieved linearity with black and white prints on Photo Rag Baryta? With any printer?
I hadn't commented before since I don't have an HP printer, but I've gotten excellent results with this paper on Canon iPF printers, both for color and B/W. It's my favorite paper. I've printed monochrome using the RGB printing path with a custom profile, as well as using the monochrome mode in the Canon print plug-in, which I've profiled with Quadtone RIP.
I wonder if it's possible you just got a bad roll? Maybe try to get a small box of sheets from another source to see if you still have the problem?
-
That's a good suggestion. I don't know whether a bad roll could create this problem, but I'll try anything. I have had brilliant results in colour from this paper (including one from this roll that looked stunning) but my b&w work is generally on matte papers or other semi-gloss stock, such as Museo Silver Rag.
-
I've used all of these gloss fiber papers on my 44" Z for years and they all perform well, except for the Ilford that has too much bronzing.
I've never experienced the problems you are having. I have used both the HP Baryta media setting and the Fine Art Pearl less ink setting both with good results. I usually use the Fine Art Pearl setting these days. I have gone through a lot of rolls of the Photorag Baryta and consider it my favorite of them all, with the Crane Silver Rag and Cone Type 5 right behind it. Also great results with the Innova Semi-Matte that curls so badly in rolls.
I also send this file over from a gg 2.2 tagged file to an HP driver grayscale workflow with no ICC profile for the majority of monochrome. However, I also have done a lot of toned RGB printing using both the Hahnemuhle PRB canned profile but primarily my own Z created profile and both worked well, though the custom one having a more accurate gamma.
The first thing I would do if I saw ANY banding is to do a nozzle cleaning. I know it is very rare to have to do one on a Z but I've done one twice in 2.5 years when I saw banding as a result of a lot of cotton rag media going through the printer and shedding lint. You also never know about major changes in humidity that happen this time of year that could be a factor.
The second thing I would do before even making another test is to do the auto head alignment. Either of these situation could cause micro banding.
I haven't run across any bad batches Photorag Baryta, though that is certainly a remote possibility, but unlikely.
John
That's a good suggestion. I don't know whether a bad roll could create this problem, but I'll try anything. I have had brilliant results in colour from this paper (including one from this roll that looked stunning) but my b&w work is generally on matte papers or other semi-gloss stock, such as Museo Silver Rag.
-
That's a good suggestion. I don't know whether a bad roll could create this problem, but I'll try anything. I have had brilliant results in colour from this paper (including one from this roll that looked stunning) but my b&w work is generally on matte papers or other semi-gloss stock, such as Museo Silver Rag.
Hmm, if color is fine I doubt it could be a bad roll (if you had previously mentioned that I must have missed it, sorry). It could be a nozzle like Dean mentioned, that might cause problems with B/W prints that aren't apparent in color images if the problem is one of the black/gray channels.
-
The problem isn't banding, it's posterization. I've done a nozzle check recently, and all cylinders are firing.
I think it's most probably a profiling issue - the problem is evident on-screen when soft-proofing with my own custom profiles, but not visible at all when soft-proofing with the canned PRB profile - yet that one is the worst offender.
John: so you recommend the Fine Art Pearl media setting as the basis of a new preset?
-
According to HP's documentation, both paper presets have the same ink limit setting. IMO, I don't think there is a difference in the two other than name.
-
Ron, you may be thinking of the Photo Pearl (more ink) setting. According to the documentation I have here, the Fine Art Pearl (less ink) setting has a substantially lower ink limit than the HP Baryta setting, as well as being slower (more passes). I would be interested in trying a light ink profile, but I'm afraid of losing dMax.
-
Hmmm. According to the (rather old) HP document I have, both HP's Fine Art Barite preset and the Pearl (less ink) are identical: Gloss enhancer, high carriage setting, photo black, ink limit of 32, rendering resolution of 600, printing resolution of 1200x1200, and 16 pass bidirectional printing mode. My APS profiles for the HP and Hahnemuhle papers were both made when the papers were released and this information was current. I know that HP did some tinkering to some of the paper presets a year ago and they may have changed, but I don't have any more recent documentation than the document released in August of 2007. I'd be delighted to have something more recent, if you know of something. I would be rather surprised if HP raised the ink limit with the Satin Barite preset, though, as they have issues with coclking on their paper (whereas Hahnemuhle does not). What information do you have about these? Also, if you are having banding problems, why would you not want to be printing with the 16 pass mode, just to be safe?
-
Hi Ron,
As I said earlier, I am not having banding problems. Just posterization - most likely a profiling problem that I don't have the technical expertise to crack.
Anyway, the latest HP technical paper on 'Working with Other Commercially Available Papers' is here (http://h10088.www1.hp.com/gap/Data/Working_with_Other_Commercially-Available_Papers_on_your_HP_Designjet_Z3200_Photo_Printer_Series.pdf) (pdf).
-
Thanks for the link, Peter. What is interesting is that there are actually two presets on the chart. One is HP Baryte Photo Paper, with an ink limit of 46, and the other is just called Baryte Photo Paper, with an ink limit of 32, the same as Fine Art Pearl (less ink). Aside: Has HP changed the spelling of their paper to reflect the spelling the rest of the world uses (it was Barite)?
As I said, I had serious cockling issues using the preset with ink set to a limit of 32 on HP's paper, so I have no clue how they could get by with that much more ink. Posterization is sometimes a sign of over inking, so I would not rule it out as the root of your problem until you test it. As for my comment about the additional passes being a conservative approach, I realize that it is not the obvious cause, and probably totally unrelated to large scale posterization issues. I haven't seen what your problems look like, but thought it might fall under "all of the above" if you were having smaller scale transition problems. Photo Rag Baryta prints beautifully for me on a Z3100ps with the settings I mentioned, output through Photoshop, with no posterization whatsoever and I don't feel like I have any dMax issues. YMMV
-
Thanks for that, Ron. I'll try reprofiling using the 32 ink limit and see where that gets me.
P
-
Thanks for that, Ron. I'll try reprofiling using the 32 ink limit and see where that gets me.
P
It could be related to what I experienced in printing B&W in B&W mode on the Z3100 and the Z3200PS.
I still use the Z3100 + PCL3 driver for critical B&W printing. I get better linear out put there.
http://www.pigment-print.com/review/Z3200FirstPage_2.htm (http://www.pigment-print.com/review/Z3200FirstPage_2.htm)
met vriendelijke groeten, Ernst Dinkla
Try: http://groups.yahoo.com/group/Wide_Inkjet_Printers/ (http://groups.yahoo.com/group/Wide_Inkjet_Printers/)
-
Ernst, do you see any improvement on the Z3200ps printing with the PCL3 driver rather than the Postscript one?
-
Ernst, do you see any improvement on the Z3200ps printing with the PCL3 driver rather than the Postscript one?
The BW tests I did were made with the PCL3 driver. The curves linked to at that page show what the problems were.
I went back to using the Z3100 and the PCL3 driver for black and white in black and white mode.
The positive answers you got were from Z3100 users if I´m correct.
Keith Cooper who also wrote a review of the Z3200 mentions this issue as well
http://www.northlight-images.co.uk/reviews...hp_z3200ps.html (http://www.northlight-images.co.uk/reviews/printer/hp_z3200ps.html)
If I had to solve this issue I would check whether it has something to do with a 1.8 Gamma in the Printer´s CM control where it should be a 2.2 Gamma or something like that. A correction curve in a custom printer filter of Qimage might do it. The clipping of Dmax as reported has to be tackled as well. It could have to do with installing the PS3 driver first and then the PCL3 driver. the first has extra 1.8 Gamma color spaces in driver´s CM choices. The last not. I reported the issue to HP and expected an answer. It is a long time ago and as written I like the Z3100 output so didn´t check whether it ever improved since.
met vriendelijke groeten, Ernst Dinkla
Try: http://groups.yahoo.com/group/Wide_Inkjet_Printers/ (http://groups.yahoo.com/group/Wide_Inkjet_Printers/)
-
What OS are you using?
If it's a mac try turning off Black Point Compensation and see if that makes a difference.
-
It's Windows. But just out of curiosity, why would turning off BPC make a difference with Macs?
-
It's Windows. But just out of curiosity, why would turning off BPC make a difference with Macs?
There is an issue on Mac 10.6 that affects what's printed. Doesn't happen to all printers but turning it off can help.
-
What OS are you using?
If it's a mac try turning off Black Point Compensation and see if that makes a difference.
The curves I have on my pages are measured on 21 step (QTR like) greyscale wedges, untagged targets printed in B&W mode of the HP Z3200 PCL3 driver that is set on driver's CM as recommended by HP. Qimage as the application and its CM off. The same route delivers linear output with the Z3100 and doesn't with the Z3200. So there's a difference. Both printers calibrated before the job. has nothing to do with the OS or BPC.
met vriendelijke groeten, Ernst Dinkla
Try: http://groups.yahoo.com/group/Wide_Inkjet_Printers/ (http://groups.yahoo.com/group/Wide_Inkjet_Printers/)
-
So Ernst, is there any point going down the QTR curves route with the Z3200? Is it not possible to make a curve that will fix the problem?
-
So Ernst, is there any point going down the QTR curves route with the Z3200? Is it not possible to make a curve that will fix the problem?
It will fix the linearity but not the Dmax loss that I see between the best Dmax on the calibration patches and the Dmax on the 100% patch of the 21`wedge. Anyway I have explained it quite well on my pages.
I will dig into this once again but right now the paper white spectral plots on my pages have my attention.
met vriendelijke groeten, Ernst Dinkla
Try: http://groups.yahoo.com/group/Wide_Inkjet_Printers/ (http://groups.yahoo.com/group/Wide_Inkjet_Printers/)