ImagePrint—it's about the print quality, too!

[Richh]

Thanks for an informative article. I've used ImagePrint for over a decade and still learned from the article. A couple points worth emphasizing...

Prints made using ImagePrint vs. through the Epson driver (in my case) are often significantly better. Specifically, where there is a lot of density, prints made using the Epson driver can "block up" and look, well, splotchy. Detail is lost. Yes, some of the lost detail can be recovered through further adjustments in PS and more test prints but in my experience, prints made via ImagePrint will still look better—and never worse.

If considering a RIP printing solution, it's worth trying ImagePrint in demo mode. Definitely compare images with rich full gamut colors and density and the differences should be apparent. The ease of use will be apparent.

Lastly, ImagePrint has EXCELLENT technical support. Colorbyte's calls are taken by real professionals with years of experience at Colorbyte vs. the offshore script readers used by most tech companies. Colorbyte's support, if needed, is as good as it gets.

[John Hollenberg]

Have you compared prints between driver and imageprint with current printers or was your comparison made a decade ago?  What printer do you have?  Thanks.

[ramd41@gmail.com]

I originally printed through ImagePrint with an Epson 4900.  I recently went to a Canon Imagegraf Pro 4100, with which I'm very pleased and purchased an ImagePrint Black license for that (gulp!).  Fortunately, the quality is excellent.  I cannot compare it to custom made profiles, but I do appreciate the ease of picking virtually any paper profile you could possibly need and produce an extremely high quality fine-art print.  When going to the Canon printer I considered purchasing the i1Pro 3 Plus for Print which runs about $2,200.  So only a $200 difference from Imageprint Black.  From all I've read from various respected sources, including Alan Briot and Mark Dubovoy (see their articles on this site), the ImagePrint profiles are superior.  So I plunked down the cash and have never looked back.  In my opinion, Dan Wells captures the experience with ImpagePrint Black extremely well.  It is all that he says it is. It produces prints superior to any service I've ever used.  I commend Dan's article.  Well done!

[John Hollenberg]

Thanks.  I may get a demo of Imageprint if I end up purchasing a printer they support.  If so I would compare their profile to a custom profile before making the purchase of Imageprint.  Unfortunately, they don't support the Canon ipf 6300 (my current printer). I am considering the latest 24 inch Epson printer (7570) due to the poor archival life of the latest Canon inks.  The ipf 6300 inkset has very good archival properties on papers without optical brighteners.

[Richh]

Have you compared prints between driver and imageprint with current printers or was your comparison made a decade ago?  What printer do you have?  Thanks.

John, my first hands-on experience using ImagePrint was with an Epson 4900, which was probably a decade back. The qualitative difference at that time was immediately obvious. More recently, I've been printing with a P7000 using the ImagePrint Black RIP. Occasionally, I'll print an A&B comparison to the latest Epson driver. The differences range from subtle to noticeable, depending on the type of image file being printed. Not the night-and-day difference I was seeing with the 4900 but the more the image is saturated with rich colors and detail in the darker areas of the print, the more obvious the differences are to me.

[John Hollenberg]

More recently, I've been printing with a P7000 using the ImagePrint Black RIP. Occasionally, I'll print an A&B comparison to the latest Epson driver. The differences range from subtle to noticeable, depending on the type of image file being printed. Not the night-and-day difference I was seeing with the 4900 but the more the image is saturated with rich colors and detail in the darker areas of the print, the more obvious the differences are to me.

Very helpful feedback.  Thanks.

[digitaldog]

John, my first hands-on experience using ImagePrint was with an Epson 4900, which was probably a decade back. The qualitative difference at that time was immediately obvious. More recently, I've been printing with a P7000 using the ImagePrint Black RIP. Occasionally, I'll print an A&B comparison to the latest Epson driver. The differences range from subtle to noticeable, depending on the type of image file being printed. Not the night-and-day difference I was seeing with the 4900 but the more the image is saturated with rich colors and detail in the darker areas of the print, the more obvious the differences are to me.

My overall experience as well. And as I pointed out in another thread, I had been a long time ImagePrint user, actually dating back to pre OS X days. I recall spending like 2 days trying to get it installed and running a the time, a nightmare. But once installed, it was the ONLY game in town for B&W printing prior to Epson's Advanced B&W. Several of photographer Greg Gorman's shows (all B&W) were printed using IP at the time (once we pulled out enough hair to get it installed). 

[Dan Wells]

Drivers have gotten better, but ImagePrint (especially IP Black) still makes a noticeable difference. ColorByte has REALLY good color science! In the early days, inkjets were unusable or close to it without ImagePrint - they were the ones who solved the green metamerism problem on some of the early Epsons.

Now, you can pull a usable print with a printer manufacturer driver and a paper manufacturer profile (those, too have gotten quite a bit better in recent years, helped by the fact that there is much less unit to unit variation among printers than there was).

ImagePrint still does better - their profiles are a cut above anything else I've used, and they are using custom dithering algorithms that significantly improve on anything the printer manufacturers are doing.

One thing to remember is that our printers have a great deal of graphic arts heritage. The drivers the manufacturers supply are modifications of drivers written for printing signs. No printer manufacturer is going to write a driver just for photography, when most large-format printers are going into the sign industry. ColorByte has done exactly that - ImagePrint Black is driving the printer optimally for photography.

[digitaldog]

Green ”metamerism problem”, seriously? No, they did no such thing.

"Get your facts first, then you can distort them as you please." --Mark Twain 

Metamerism is a simple attribute: two (2) samples with different spectra compared to each other with a given set of viewing conditions, produce a match. Metamerism only applies to two color patches when they are compared to each other. It is incorrect to refer to one color from a given ink or paper and say that it suffers from metamerism. A "meterameric pair of color patches" means that they appear to match under a given illuminant. However, they may not appear to match under another illuminant. Metamers ("metameric stimuli") by definition are two different spectrums that appear to be the same color. If they don't look the same color, they are not metamers.

If you are viewing a print (lots of colors) and within differing viewing conditions, and there is a mismatch, this could be called a metameric mismatch or metameric failure. But its not metamerism. One sample (the print) compared to itself in differing viewing conditions, the proper term would be when they appear to match is color constancy and when they don't, color inconstancy. The lack of a defined term for the metameric mismatch is the problem.  "Metameric failure" is the best so far because it is unambiguous.

[digitaldog]

Your understanding of ink jet photographic heritage is kinda spotty too:

https://americanhistory.si.edu/press/releases/national-museum-american-history-adds-nash-editions-digital-photography-equipment

http://www.wilhelm-research.com/nash/Nash_Editions_Book_Wilhelm_Print_Permanence_Essay.pdf

[Dan Wells]

What would you consider the proper term for the green/greenish-bronze color cast the early Epsons had under certain lighting (which ColorByte figured out how to suppress in most circumstances)? It may not quite properly be metamerism, but what, exactly is it?

While I was not printing at the time, I am aware of the VERY early Nash Editions prints on Iris Graphics proofing printers. Most of us never had an opportunity to use them, and they were a completely different type of machine from the Epsons, Canons and HPs that took over from them. A drum-based printer using continuous flow inkjet is a very different technology from one that holds the media flat and uses drop-on-demand. They, too had graphic arts heritage, but it was high-end proofing rather than the sign industry.

[digitaldog]

What would you consider the proper term for the green/greenish-bronze color cast the early Epsons had under certain lighting (which ColorByte figured out how to suppress in most circumstances)? It may not quite properly be metamerism, but what, exactly is it?

I already provided an answer. And, there is NO green cast depending on the illuminant. Colorbyte and others figured out how to suppress it; it's why for no money at all, Epson users can invoke the Advanced B&W mode.

If it was not for metamerism, none of our three color reproduction systems would work.  Because of metamerism we are able, using only three colors, to cause the human vision system to perceive a match between this tristimulus reproduction system and full spectral pigment. Metamerism is not a fault in a given color reproduction system.  Metamerism is a good thing.

And in terms of the Iris, it was used in pre-press for quick and temporary proofing of photos; not signs. The inks were vegetable dyes, you could (and in the day I did have the opportunity) lick them clear off the print. The history of ink jet printing of photograhy, and experience of some here are factual.

[dkaufman]

I've been following discussions of ImagePrint on this site and others for several years now, without feeling there is definitive evidence one way or the other about its print quality compared to manufacturer printer drivers. I acquired an Epson Surecolor P-9000 in December 2019, so I finally decided to download ImagePrint 10's trial version and see for myself. I might add that for at least five years before acquiring my own large format printer, all my prints, including for numerous exhibitions, were printed on Epson printers through ImagePrint at a service bureau, with final files prepared and supplied by me. I loved the quality.

My testing was concerned solely with print quality, not with the myriad of other attractive features that ImagePrint offers. My standard exhibition prints are architectural images, 24 to 40 inches in height by varying widths,  produced from very large Phase One stitched files. I tested images at native resolutions as I was not testing for uprezzing or interpolation either. I also tested small prints from Nikon DSLR files, portraits of family members at normal small sizes--8" x 10". I used Epson Premium Luster (260) paper for convenience for all  the tests, as it emerges from the printer very flat and it makes it is easy to see fine detail in the images.

I looked primarily at resolution, fine detail, and dithering. I normally use Epson's finest photo settings:  2880x1440 dot placement, unidirectional printing and normally finest details off (as recommended for photos in the Epson manual). I used my own custom profile for the paper and and ImagePrint's supplied profile for 2880 dot placement and daylight viewing.

The results: For large images which I printed at native 360 dpi input resolution, I saw virtually no difference between the Epson driver and ImagePrint. If anything, the dither pattern in fine detail (visible to my eyes only through an 8 or 10 times loupe) was smoother with the Epson driver. Slight advantage to Epson here.

For small images which I could print with native 720 dpi input resolution, at 2880x1440 dot placement, ImagePrint produced a smoother dither pattern with more accurate detail (but again only easily seen with a magnifying loupe of 8 or 10 power). Advantage to ImagePrint here. However, when I did what Epson advises against, and turned on Finest Detail, the ImagePrint advantage was erased and the Epson dither pattern and detail rendering matched the ImagePrint version of the print. I still think ImagePrint in this case had a very slightly smoother dither, but the difference was barely visible under a magnifying loupe.

I also looked at profile differences. My own custom profile for the Epson Luster (260) paper was made with an XRite iOnePro2 measuring system using a 2033 patch set. The ImagePrint profile was very slightly different and this was visible most easily on large prints. The ImagePrint profile produced slightly lighter prints (equal to a few points higher midway on a Photoshop curve) but with less red saturation throughout the test image. I think the ImagePrint profile more closely matched the screen image on my calibrated wide gamut NEC monitor in terms of image brightness. However, the profile differences are very small.

I also tried to examine claims that ImagePrint produces a more neutral B&W image. I normally print my B&W images using an RGB workflow which I find easier to control than Epson's ABW workflow. I printed a large, full tonal range, architectural B&W image three ways: With my normal Epson RGB workflow, with Epson's ABW system (using the Dark setting), and with an ImagePrint's gray profile for the Luster paper. I found no difference in image tone visible between the three resulting prints. It was impossible to pinpoint any difference in the three prints. (As a side test, I also used Epson's print cost utility to examine the ink usage of the Epson RGB print versus the Epson ABW print. Both images used all of the inks, many in very small amounts. The Epson RGB print uses more Cyan and less Light Cyan than the Epson ABW print, and more Light black and less Photo Black as well, plus some difference in Magenta usage. But when you look at the prints it is very difficult to impossible to see any difference in print colour. Sometimes I thought the Epson ABW print was very slightly less bluish in daylight but I'm still not sure.) I saw no basis for the claim that ImagePrint's gray drive produced a more neutral image.

So to sum up, looking at the issue of print quality alone, ImagePrint may have a very tiny advantage in dithering pattern in very fine detail for high resolution images (native 720 dpi images) but it's one that would be very hard for most people with normal eyesight to see. And, of course, the supplied profile(s) from my very limited experience appears to be very good.  So while I was prepared to spend the considerable sum to acquire ImagePrint, I did not find it offered any advantage in print quality that would lead me to acquire the software. Nevertheless, those who would use its many other significant features, such as layout ease, multiple profiles, etc., may come to a different conclusion than I did.

David Kaufman
(www.davidkaufmanphotography.com)

[John Hollenberg]

I acquired an Epson Surecolor P-9000 in December 2019, so I finally decided to download ImagePrint 10's trial version and see for myself.

Very helpful information.  I assume you tested ImagePrint Black (the Red version uses the Epson driver)?

[dkaufman]

Yes, in reply to the above query, I tested ImagePrint Black. Furthermore I work with 16 bit files in Prophoto colour space, part of my standard workflow.
About the profile comparison: Looking more carefully at large images printed with my own custom profile and ImagePrint's appropriate profile for the paper and printer dot placement resolution, I would say the ImagePrint profile produced a slightly bluer print and perhaps a tiny bit lighter as well, which explains why the ImagePrint image looked a little lighter and had less red/yellow saturation. I would now say that the ImagePrint profile was not as good as my custom profile, but that would not necessarily be unusual.

David Kaufman
www.davidkaufmanphotography.com

[dkaufman]

Further to my just posted reply, I would like to add some further thoughts:

I was happy for many years to use Imageprint at the service bureau where I made my prints prior to acquiring my own printer last December, for longevity reasons. ImagePrint reputedly used much less of the yellow pigment in the previous pigment colour set used by Epson for its 7900/9900 series printers, a yellow pigment that was reputed to fade faster than other colours in the set.. However with the new pigment ink set developed for the Surecolor P7000/9000 incorporating a yellow pigment that held its own for a longer period without fading (for particulars see the Aardenburg Imaging site) that became less of an issue. Whether or not ImagePrint's former advantage in reportedly using less yellow pigment still applies is hard to know without knowing how much or how little of the yellow pigment from the current P-9000 ink set is used by the ImagePrint 10 Black driver. According to the Aardenburg site, Epson's ink set is now significantly more permanent than Canon's current inks, although still lagging significantly behind HP's inks.

Second about the B&W images I compared and tested: Comparing a B&W image printed through a 16 bit RGB workflow and through Epson's ABW workflow, the following ink usages resulted:
Epson RGB: CY: .07  OR: .03  Y: .03  LCY: .09  PB: .20  VM: .06  LB: .54  G: .06  LLB: .38   VLM: .14
Epson ABW: CY: .02  OR: .02  Y: .03  LCY: .18  PB: .32  VM: .02  LB: .25  G: .02  LLB: .21   VLM: .24

Analyzing the above, the ABW workflow uses less of the stronger CY, VM, and G, and more of the LCY and VLM, also different proportions of the blacks. As I said, I saw no difference in print colour, but perhaps the different ink usage contributes to greater longevity for the ABW prints. As I reported I could see no difference in the ImagePrint B&W image using the gray profile, either in colour or in any other aspect of the image. When looking at the ImagePrint print under a ten times loupe, one could see flecks of cyan throughout the image but the loupe was not strong enough to show other colours that may be used. I don't know that ImagePrint has published any data on what ink colours it uses for B&W images printed with its gray profiles.

David Kaufman
www.davidkaufmanphotography.com

[Richh]

David, thanks for sharing your findings using ImagePrint vs. your custom profile. Did you happen to also evaluate prints made using the canned Epson ICC profiles?

I like to print using a wide variety of paper surfaces, so having access to all the profiles available through IP Black has been great.

[dkaufman]

In reply to the question about the Epson profile: I tested the Epson profile before I made my own custom profile, not only for Luster paper but also for several other papers. (However, I did not print through ImagePrint Black using the Epson profile instead of theirs.) For my own printer I generally have found Epson supplied profiles and Canson profiles to be very good for their respective papers. My own custom profiles seem to be consistently less blue and have better and more accurate yellow saturation for some papers for my printer (which shows most readily in skin tones and yellow objects) but otherwise the differences are not easy to see and hard to pinpoint without doing the sort of objective measurements done by Mark Segal when he tests papers. This may be a failure of perception on my part or an indication that supplied profiles for newer printers  have become very good. The ImagePrint profile for Epson Luster (260) was more different from my custom profile than Epson's supplied profile. This is all very subjective and based on my perception which is not the same as rigorous measurement.

David Kaufman
(www.davidkaufmanphotography.com)