Resolution loss from transparency to print

[bob carnie]

I am lucky enough to see on a daily basis , prints made by inkjet, prints made by lambda laser, and yes optical prints. I scan and print each day , moving from one process to another , somedays printing all three.

After about 8 years of observing digital process images that has indeed gone into Museums, and Gallery's from our equipment I will say that all the forms of printmaking have their strong points and some weaknesses. I have yet to see a print made from either traditional enlarger or high end digital blow the other away.
In fact the differences are so subtle that even after 35 years of image making and being able to print sniff , I have a hard time saying one process is better than the other. I do look at the images supplied and make an educated guess as to what process will make the image work best. Then I do the tests to see if I was right.


Good negative exposure, good processing, good enlarging practices with good optics and flat film plane will give very sharp results.
Good image capture, good PS editing, good laser equipment with excellent sharpening skills will give very sharp results.

Paper surfaces have the most to do with percieved sharpness, a heavy rag paper will not seem as sharp as a high gloss cibachrome no matter if you print optically or digitally. I was lucky enough to bridge both methods of making cibas and can say that its the workflow that is important...

If you are lucky to live in a large market like Paul does and have the inclination to take advantage of the numerous shows on a weekly basis you will see that his argument is correct.
If you are making complicated contrast, colour masks for enlarger printing you will see Drews point.

I have been lucky enough to live in a large market (Toronto) and see a lot of work from various regions of the world , so I see Paul's point, and as well I do come from Drew's background of complicated mask-making to enhance colour images and can appreciate his viewpoint.

I think it boils down to good workflow , appreciating different print processes and understanding that some will have attributes others do not. I do not think we have ever seen a better time to be making photographs. This mixture of methods is mind boggling and to state one method is better than the other is IMO silly.

We just upgraded today our Cannon IPF 9000 to a Cannon IPF 9400 , a noticeable difference with improved ink sets is apparent .
This is a three year cycle that all users of inkjet will notice until the ink sets match Pro Photos gamut or beyond.
I see a different colour pallet with RA4 than that of our current ink sets, I appreciate both differences and based on the image and how I want to make the final print look, is more important to me than the type of optics or lasers produce the image. The image controls the process as I like both processes sometimes I have to try both to decide, just like I like a glossy print for some of my BW work and Matte surface for others.

[Drew Wiley]

They aren't inks, Paul - they're micro pigments originally designed for totally different applications, so would still need R&D for printing machine application, unless you were simply dispersing them into an assembly printing process analogous to
quad carbon or Fresson. The point is, it would be a vasty simplified system, and you would have true transparency comparable to organic dyes, yet with extreme permanence. Don't hold your breath. Maybe something could be done that resembles early color movie hues or an old Autochrome. But true process colors don't exist yet. In a second category they do, but would not be suitable to inkjet at this time, only layered assembly techniques.

[Drew Wiley]

Just one point, Bob, some of the very most expensive digital printing shown in NYC or even being put into permanent installations in NYC, Germany, or Seoul is being done right in this immediate neighborhood, basically by hired guns with
incredible equipment and backgrounds. Very little of that kind of thing is locally exhibited, though I might try to scheme up
something in the future, nor is it locally sold - we're talking about seven figure stuff. Rich guys around here tend to buy silly pop art and and ocassional AA or EW print. Rich geeks don't generally become good collectors. And the most expensive lenses in the world are made right here; and there are plenty of digital printing gurus around who I know one a first name basis. So cutting-ege techie exposure is no issue at all. Personally, I prefer the look of a platinum print made in a chicken
house by Julia Cameron. By I am currently in the school of hard knocks making the transition from Ciba to Fuji Supergloss both from internegs and current color neg originals - that is, in terms of optimizing this kind of thing to the level of quality
I previous expected with Ciba. A new challenge, but you know ... a step at a time.

[Nathan Potter]

From the previous post, 300 dpi = 0.0033" dot.

At this moment in digital print technology, I'm not yet convinced it is mature in the same way film technology is. There is a long ways for digital technology to go.
When digital print technology can produce random sized dots on the order of 0.000001" to 0.000005" then we are nearing what is film is capable of. The orderly nature of digital image sensors and print technology adds to the problem of information reproduction. It is why Moiré patterns happen in digital and not with film.


Bernice



Bernice

Bernice, your dimensions of 0.000001 to 0.000005 inch for equivalent film resolving power is much too small if you speak of equivalent grain size to dot size. Modern films after development have a typical grain size from 0.5 µm to maybe 10 µm for large clumps. Your figure implies size 0.025 µm to 0.125 µm. (20,000 lp/mm to say 5000 lp/mm). Even the finest emulsions made by Eastman Kodak (Kodak High Resolution Glass Plates) could resolve about 2000 lp/mm. under extraordinarily precise exposure and processing conditions.

By the way I just saw an advertisement by Epson of a new printer with a dot size capability of 25 µm on paper (very smooth paper I assume). Thats 20 lp/mm, 1016 dots/inch. Seems almost overkill for what the human unaided eye can perceive at close viewing distance.
If the print is an enlargement from an original diminutive source then the source film needs to have greater than 20 lp/mm or the information conveyed is emply of the detail that would be possible.

Nate Potter, Austin TX.

[paulr]

By the way I just saw an advertisement by Epson of a new printer with a dot size capability of 25 µm on paper (very smooth paper I assume). Thats 20 lp/mm, 1016 dots/inch. Seems almost overkill for what the human unaided eye can perceive at close viewing distance.

The reason for the tiny dots in a screened printing process is to have as many dots per pixel as possible. The greater the number of dots you can fit in, multiplied by the number of shades of ink, corresponds with the number of discreet dynamic levels you can display. With a high ratio of dots to pixels and a high number of ink shades, it becomes very easy to linearize an inkset and produce tones that are perfectly smooth from paper white to full black. The lower these numbers, the trickier it becomes to get smooth results.

I asked Jon Cone why he upgraded the Piezography inks to six and seven shades of gray ... I thought the older 4 ink process looked amazing. He said it was mostly to make linearization easier. With four inks, it was extremely hard to get the curves and ICC profiles to work. Any deviation and the whole thing would fall apart. This fit my experience—I had had Cone make me a custom ICC profile for the old quadtone inks, and it took him three tries to get it right. The first two weren't even close. Now with the 6 and 7 ink systems, a generic profile works for most people. It's become easy for him to linearize the tones.

[Drew Wiley]

I'd be extremely skeptical that any kind of paper per se could hold micro-detail. Sound like advertising hype to me. You'd need polyester film at a minimum. Just because you might be able to spit it out that fine doesn't mean that's how it will
actually land. And I don't think any of us know what film actually can do. My brother was always of the persuasion that the
"military" was always twenty or thirty years ahead of the curve compared films available to the public. Having seen a number
of classified prints over the years, I'd say that was completely true. Whether it's still true during this era of increasing digital
survelliance, I cannot say either. Ultimately, for me at least, it's not about hypothetical resolution once you're below the
threshold of visual recognition, but about the smoothness of the tones, distribution of the hues (gamut), surface quality of
the print, etc. And in color work at least, inkjet and chemical photography just look different. With monochrome, it's more about toning versus whatever ... personal choices.

[Brian C. Miller]

By the way I just saw an advertisement by Epson of a new printer with a dot size capability of 25 µm on paper (very smooth paper I assume). Thats 20 lp/mm, 1016 dots/inch. Seems almost overkill for what the human unaided eye can perceive at close viewing distance.
If the print is an enlargement from an original diminutive source then the source film needs to have greater than 20 lp/mm or the information conveyed is emply of the detail that would be possible.

Nate Potter, Austin TX.

I took a look at the Epson site, and came across the 7900CTP, which is used for printing aluminum printing press plates. The printer specs claim 150lpi (plate, not paper). That could be the next new frontier, but I'd hate to try and send plates through the mail.

[paulr]

Yup, personal choices. I haven't said my digital prints are better than my darkroom prints. In some cases I think they are and in others not. Someone else looking over my shoulder might have completely contrary opinions. But I think everyone would agree that they appear sharper and more detailed. These are qualities that may or may not be important to any given work.

Film resolution is easy to know. The manufacturers have been publishing MTF curves for a long time. Of course it's a curve, not a hard limit, so it has to be interpreted on situational basis. Papers certainly hold as much detail as film. I've never seen anyone publish MTF data for papers, because they're all probably at 100% all the way up to the point where the detail is too fine to see. The exception would be highly textured art papers. No one reaches for those if they want to see a lot of detail.

[Drew Wiley]

Or course, detail is only one parameter. That's where polyester media like Ciba and Supergloss excel. I also fool around a little bit with dye transfer, which is certainly no winner in that category, but is in terms of gamut and saturation. A few folks are tinkering with the idea of running those dyes thru printers onto mordanted paper, but still some issues, and probably wouldn't compete with inkjet for permanence. But let's face it - something like eight to eleven inks in a set, whereas dyes can do it better gamut-wise in three. Probably an iris printer would work for something like that. But I like the tactility of actual darkroom work. I'm more interested in some of these various options conceptually, cause I only have time to work with
a very limited number of processes for real. Got my gear and methodology, so no sense to change now.

[Nathan Potter]

Paul I hadn't contemplated possible advantages of a 20 µm dot size for halftone dots but inkjet prints are not halftones so I assume you are only comparing the advantages of very fine halftone to what might be a similar advantage with inkjet.

But I've been thinking about the inkjet method of generating large format negatives for direct contact printing, a technique being more widely used for carbon and Pt/Pd contacts and employed by a number of workers here and other places. With 20 µm dot technology a significantly finer digital master could be generated on a transparency film such as pictorico. Such detail on film could even stand a 4X enlargement to paper and still yield around 300 DPI for acceptable print quality.

I tend to agree with Drew that such a tiny dot pattern may be almost nonsensical on most normal papers.

I see the 18 µm line machine is listed as a plotter by Epson so now I'm not so sure it would be suitable for full photographic images although a full color photographic image is shown.
The Epson SureColor T3000 is a 24 inch wide machine at $2995. 36 inch at 3995 and 44 inch at 4495. Advertised in the IEEE Spectrum from the Institute of Electrical and Electronic Engineers. They may be targeting the IC industry where, in particular, I have always been a bit frustrated with the quality from printed blowbacks of chip designs. Cartridges up to 700 ml are listed as available.

Nate Potter, Austin TX.