Trial & Error Process of Custom ICC Profiling (Photo Rag Satin)

[John Caldwell]

Recently I engaged Booksmart Studio to prepare Hahnemüle Photo Rag Satin custom ICC profiles for three Epson printers (3800, 4900, 9900). Booksmart advised that we profile the paper for MK and PK instances. The tech expressed that he felt that a proper PK profile would offer greater Dmax and better results on the whole than an MK profile would. It was suggested that the PK target be printed using default Epson driver settings for Ultra Premium Luster media type, but with a wide platen setting. The MK target was printed with Velvet Fine Art default settings.

The resulting MK and PK profiles are quite different for all three Epson printer models. Soft proofing does predict the results quite accurately, and in all cases, the PK blacks lack density, and quarter tone contrast is low when compared to the MK result. The result is, in other words, opposite what I was encouraged to expect from the PK selection. The MK results are much snappier, and to my eye the Dmax is much greater.

Booksmart explains that the poor Dmax resulting from the custom PK profile probably reflects insufficient ink being laid down and as dictated by the driver settings during the target image printing. When I asked what they would do next, they suggested a few more attempts at printing the PK targets, but using a different media type selection in the Epson driver. It was stated that since we are looking for more ink, rather than less, that I use media types that are more glossy than the type I used when printing the 1st attempt's PK target. Glossy or semi-gloss seem like ether might be good choices. Booksmart advised this method of preparing new targets over keeping the media type constant, but adjusting instead the ink limits upwards in the driver for a given media type.

So I apologize for the long question, but I am seeking your advice - either in general or in specific- about how you would handle this problem of poor PK custom profiles. I'm attracted to making PK work for the Photo Rag Satin because it would would allow me to make PK-MK ink changes less often and because at least one source reports that this paper is at its best when PK is used.

Many thanks,

John Caldwell

[elolaugesen]

I always use MK for the Han   Photo Rag as does my supplier who buys this papers and renames them.
I do my own printer profiles using SpyderPrnt
I do not use the satin paper any more - just tested it    but I do use the PHOTO RAG
cheers  

[John Caldwell]

Photo Rag Satin, which differs from Photo Rag, is ambiguous as to its being Mk or PK. Traditional Photo Rag is my idea of a clearly MK paper. In the past, we had always used Hahnemüle's canned profiles for the satin paper, but even Hahnemüle isn't consistent with their canned profiles - they specify Mk in some instances (Epson 3800 & 9900), but PK for the Epson 4900. In any case, we've always used the canned MK profile for our 3800 with really good results, but the prospect of achieving the same or better result with Pk is attractive.

Thanks for your interest.

John Caldwell

[Alan Goldhammer]

See Scott Martin's writeup on media selection HERE.  You can print out the image using different settings and see which one works best.  The key here will be getting the best Dmax.  I remember looking at this paper a year or so ago but wasn't all that impressed with its performance compared to the others that I was printing on at the time.  I'm not so sure the recommendation to use either PK or MK ink is valid.

[John Caldwell]

Alan, Thank you. I knew peripherally of Scott's work in this area but have not tuned in as I should. This looks helpful.

PR Satin is not for everyone or every image. To my knowledge, it's a very unusual surface. I've never heard anyone reference and paper that's like it, but I've heard a number say it's not for them.

Thanks for this lead, Alan.

John

[Czornyj]

MK is no-brainer for PRS. It has some misleading, shiny wax-like addition in paper base, but there's no coating on the surface, so there's no chance to achieve higher d-max with PK in this case.

[Ernst Dinkla]

PR Satin is not for everyone or every image. To my knowledge, it's a very unusual surface. I've never heard anyone reference and paper that's like it, but I've heard a number say it's not for them.

John

HM White Etching Satin 270gsm should be identical but is no longer produced. This was discussed in March 2012.


--
Met vriendelijke groet, Ernst

340+ paper white spectral plots:
http://www.pigment-print.com/spectralplots/spectrumviz_1.htm
update april 2012: Harman by Hahnemühle, Innova IFA45 and more

[tastar]

In my opinion, you discovered the short comings of RGB profiling without a RIP and without an instrument that can measure color. The RIP's that we use (EFI, Colorburst and Wasatch) provide the ability to linearize the printer before building a profile. The linearization process can be used to set maximum ink densities for C,M,Y,K (and Orange and Green for the 900 series printers), and overall ink limits as well. And, we always tweak our linearizations to get target dMax values that match the environment in which the profile will be used (which is generally offset printing proofing, but the same principles apply with profiling for photography). And, when we are building a profile, the part of the process that takes the longest is getting a good linearization to use as the basis for building the profile - which gives consistency to the output from printer to printer.

The target from Scott Martin mentioned below is a great idea - it would somewhat approximate what the RIP profiling process does with linearization. We've done a couple of RGB profiles in the past, and have picked a media, printed a standard linearization target, and tweaked the maximum ink density using the sliders available in the printer driver (don't ask which sliders we used). The difference in the final RGB profiles with and without this tweaking was enormous.

So, to appropriately use this paper, and to get consistent results from printer to printer, I would recommend developing a baseline starting point on each printer that matches the other printers. And, if you just have one printer, so that you get consistency from one paper to another, you should still work to develop baseline values for each paper. Once that baseline is defined, you can then use a service like Booksmart Studio to build the profile. From what I know about this paper, I would guess that it would be able to produce dMax values that are close to photo papers. When we head back to the shop on Monday, assuming that we have some test sheets, we'll run a CMYK linearization and post the dMax values that we get from the paper.

Tony

[Czornyj]

So, to appropriately use this paper, and to get consistent results from printer to printer, I would recommend developing a baseline starting point on each printer that matches the other printers.

Why not just use Hahnemühle recommended driver settings?

From what I know about this paper, I would guess that it would be able to produce dMax values that are close to photo papers.

No, it's not - it's uncoated.

[Pete Berry]

PR Satin is not for everyone or every image. To my knowledge, it's a very unusual surface. I've never heard anyone reference and paper that's like it, but I've heard a number say it's not for them.
John

Yes, a very strange paper. Several years ago I printed a test image with PRS as well as other baryta papers (Hahne. PRP, FAP; Ilford GFS; Harmon FbAl Gloss) and a couple of standard RR and Ilford satin/pearls - all with the same PK driver settings (iPF5000, highest ink load "Special #5" media selection) for identical inking to see what differences I could see. Except for the strikingly flat and very poorly saturated PRS, there was very little difference in the papers when laid out together and viewed from a couple of feet, aside from that due to the base warmth or coolness.

Pete

[Ernst Dinkla]

The Hahnemühle handling sheet for the Photorag Satin mentions MK ink for the HP B9180. I do not think one should see this paper as a Fiber-Baryta quality but more in the Photorag matte paper range. The recommended media preset is also the one normally used for Photorag (though in its HP disguise).


--
Met vriendelijke groet, Ernst

340+ paper white spectral plots:
http://www.pigment-print.com/spectralplots/spectrumviz_1.htm
update april 2012: Harman by Hahnemühle, Innova IFA45 and more

[teddillard]

The tech expressed that he felt that a proper PK profile would offer greater Dmax and better results on the whole than an MK profile would.

This statement makes no sense to me.  MK ink produces more DMAX to compensate for the more porous matt-type papers.  PK inks perform just as you've seen- less dense blacks, ("the PK blacks lack density, and quarter tone contrast is low when compared to the MK result") but on a photo-surface paper it's not a problem, and gives you a correct black.  There's no way a "proper PK profile" will give you more DMAX on anything, to my knowledge.  The followup explanation from them seems like so much backpedaling.  

MK inks don't dry properly or adhere, for lack of a better word, to the surface of a photo-type paper.  They give you a deeper black, but smear, on an incompatible paper.  If you print with MK inks to whatever stock you're using, and it dries properly, then that will give you the most density.  

With due respect to the Booksmart tech, I think you got some bad advice from a tech who doesn't understand Epson ink systems.

...seems simple.  Run a test print on the paper with some good black areas with MK inks, if you think they are compatible.  If they don't smear and dry OK, then build the profile with them.  Game over. 

[Ernst Dinkla]

I have not used that paper type but one thing is obvious from the samples I have; the satin expression of the paper surface happens where the ink is laid down, the remaining white paper keeps a more matte surface. Like a satin spot varnish does on an image in offset printing. That is a nice effect in my opinion. The question is whether that effect is kept with MK too, the black ink Hahnemühle recommends for the HP B9180. The sample I have is printed with the HP Z3100 by Hahnemühle, so I would think they sticked to their recommnendation in using the Vivera MK black and the surface does what I described above. However I read the instruction for the Z3100 at the HM site and in that case they recommend PK. I do not think HM is sure itself about what black to use. There are several Z3100 owners that praise this paper in combination with that printer, they could tell what to use.


--
Met vriendelijke groet, Ernst

340+ paper white spectral plots:
http://www.pigment-print.com/spectralplots/spectrumviz_1.htm
update april 2012: Harman by Hahnemühle, Innova IFA45 and more

[John Caldwell]

This has been an educational discussion and I thank everyone for their participation.

Tony: Although I don't understand how printer/paper linearization with the CMYK environment can be later translated to the RBG world, I believe I understand your point of establishing just what amount of ink a paper can optimally "receive" as a starting point, and building your profiles from this point. I'd welcome any consultation you can provide in this area.

Ernst, as you point out Hahnemüle doesn't seem sure about what black in to pair with the Photo Rag Satin. When we recently added an Epson 4900, I was quite surprised to see that Hahnemüle's canned profile for PR Satin called for PK, rather than the MK we'd used in the past with the 3800 series. Then when Booksmart suggested profiling PK, stating better results would be had, I got interested but less sure of what to do.

I should be clear that the result we are getting with the MK custom profile on the 4900 is very good, and I'm tempted to quit there expect thatI'd learn nothing further about controlling the variables involved.

I've yet to try the PR Satin PK profiles on the Epson 9900 and 3800 machines, but soft proofing predicts the same poor Dmax we saw with the 4900.

Many thanks to all.

John Caldwell

[tastar]

I know that this is pretty simplistic - I'm sure that the Color Management forum people would be on my case for this reply. But, when we profile for offset proofing, we shoot for GRACoL ink densities of C: 1.45, M: 1.45, Y: 1.0, K: 1.7. If you want more pop in images and don't care about matching printing press sheets, these values can be increased as long as you aren't putting more ink on the sheet than it can handle. We read maximum ink densities for the PRS this morning, and with Photo Black ink, we got: C: 1.67, M: 1.72, Y: 1.29, and K: 1.48. With Matte Black, we got  a K of 1.91. So, this sheet has a much better maximum Photo Black density than matte sheets, which would generally fall in the 1.2 - 1.3 black density range using Photo Black. But, black heavy images would have a much more dense and richer look with Matte Black ink (but the matte black ink would dull down the satin finish, too, I would guess).

Re. using a CMYK profile for RGB photography, the CMYK profile that you would build with a RIP is the output profile, so it applies standardization to the printer output. If you are sending the RIP an RGB image, then the RIP can assign an input profile, like Adobe RGB, or sRGB or ProPhoto RGB to the image. I remember talking to a printer on the West Coast a few years back who was using the ColorBurst RIP, building CMYK profiles for their printers to get predictable output, and using ProPhoto RGB as an input profile to get knock out good photographic prints.

Tony

[Ernst Dinkla]

I know that this is pretty simplistic - I'm sure that the Color Management forum people would be on my case for this reply. But, when we profile for offset proofing, we shoot for GRACoL ink densities of C: 1.45, M: 1.45, Y: 1.0, K: 1.7. If you want more pop in images and don't care about matching printing press sheets, these values can be increased as long as you aren't putting more ink on the sheet than it can handle. We read maximum ink densities for the PRS this morning, and with Photo Black ink, we got: C: 1.67, M: 1.72, Y: 1.29, and K: 1.48. With Matte Black, we got  a K of 1.91. So, this sheet has a much better maximum Photo Black density than matte sheets, which would generally fall in the 1.2 - 1.3 black density range using Photo Black. But, black heavy images would have a much more dense and richer look with Matte Black ink (but the matte black ink would dull down the satin finish, too, I would guess).

Re. using a CMYK profile for RGB photography, the CMYK profile that you would build with a RIP is the output profile, so it applies standardization to the printer output. If you are sending the RIP an RGB image, then the RIP can assign an input profile, like Adobe RGB, or sRGB or ProPhoto RGB to the image. I remember talking to a printer on the West Coast a few years back who was using the ColorBurst RIP, building CMYK profiles for their printers to get predictable output, and using ProPhoto RGB as an input profile to get knock out good photographic prints.

Tony

For an inkjet printer with 6 to 12 inks you would aim for the highest chroma on the 100% CMYRGB patches, the highest Dmax for black and create a CMYK profile or an N-Color profile. And depending on the printer's gamut, use either AdobeRGB or one larger than AdobeRGB like ProPhoto as the latest 10-12 ink printers have a gamut wider than AdobeRGB.

Given a rich variety of media presets, documents describing the media presets and media presets made by third parties for third party papers, there are good reasons to stay with the normal printer driver and RGB-device ICC profiles. The RIPs are no longer superior in weaving or dithering patterns, the creation of N-Color profiles for RIPs is in many cases not delivering a better print. My experience with the Wasatch SoftRip for the HP Z3100 showed that even Wasatch made three lousy profiles for that printer, the normal HP driver had much better UCR, less metamerism, good neutrals, good gamut, efficient ink use. HP improved it for the Z3200. There will be exceptions like the GMG RIP + superior profile creation software, all at a price, but HP did a splendid job in my case.

On the D 1.91 for MK on PRS, that is already a high number for MK, on matte papers a  D 1.8 is normally a high score. It would not surprise me if the satin effect comes through the MK patch and creates the D 1.91. I find your D 1.2 to 1.3 for PK on matte papers a low estimation, I measure D 1.482 for the Z3200 Vivera PK 100% patch on Photorag 310 grams, so equal to your PK on the PRS. And D 1.776 for Z3200 Vivera MK on Photorag. Both taken from a Z3200 calibration target about a week old. The Z3100 scores higher on Photorag in my experience. I might find the time to make a calibration on the single (less than A4) piece of virgin PRS I have, wonder whether the satin effect is lost with MK in the deepest blacks.


--
Met vriendelijke groet, Ernst

340+ paper white spectral plots:
http://www.pigment-print.com/spectralplots/spectrumviz_1.htm
update april 2012: Harman by Hahnemühle, Innova IFA45 and more

[Czornyj]

But, when we profile for offset proofing, we shoot for GRACoL ink densities of C: 1.45, M: 1.45, Y: 1.0, K: 1.7.

that's so 90's...

It would not surprise me if the satin effect comes through the MK patch and creates the D 1.91.

In case of my LUCIA EX and K3 VM ink sets the satin effect was still there, even in deepest MKs. As far as I remember Viviera has higher dmax for MK, but so or so I don't expect it would be different.