Chucking it and going to DSLR?

[Oren Grad]

Are you talking about inkjet? I'm referring to C-prints. Real c-prints. As in RA-4.

There's no way RA-4 comes remotely close to 5 lp/mm. This is from direct experience, both with high-end Lambda output and direct prints from negatives.

Ctein has tested many papers, B&W and color, using a contact-printed high-resolution bar target. He reported that Portra III RC paper recorded a resolution of 100+ lp/mm and Ultra II RC recorded 125+ lp/mm, and that these were typical of color print materals introduced from the 80's onward.

See "Is Your Print Paper Sharp Enough?", Photo Techniques, Mar/Apr 2002.

Paul, in the same article he reports why resolution greater than 10 lp/mm matters, up to a point. The gist of it is that it's possible to tell the difference between a square-wave pattern and a sine-wave pattern at 10 lp/mm. Mathematically, the bulk of that difference is contributed by the third harmonic, which means that somehow information at 30 lp/mm is being perceived. This is consistent with his tests of the long-discontinued Ektaflex print material, which recorded 18-22 lp/mm but nevertheless was perceived to be less sharp than conventional Ektacolor paper. But 30 lp/mm is about as far as makes any perceptible difference.

Achieving greater than 10 lp/mm in pictorial prints, of course, requires a good negative to start with as well as very careful printing technique, whether for contact printing or enlargement.

[Ben Syverson]

Ctein has tested many papers, B&W and color, using a contact-printed high-resolution bar target. He reported that Portra III RC paper recorded a resolution of 100+ lp/mm and Ultra II RC recorded 125+ lp/mm, and that these were typical of color print materals introduced from the 80's onward.

Oh, Ctein said it so it must be true. Nonsense. The paper may give a (very) smudgy rendering of the bar chart, but it really will come down to what you define as "detail."

A 300 DPI file can resolve 5.9 lp/mm. If color papers could resolve way beyond that, you would expect any Lambda/Lightjet/etc print to reproduce every single tiny detail, down to individual stuck pixels or dust from scanning. The reality is that the detail visible looks very close to the same file viewed at 50% in Photoshop. That would mean the print is really only about 150 DPI, or 3 lp/mm.

This is consistent with my experience printing color photochemically. Looking at the image through the grain focuser, you can see details from the negative that just never make it to the print the way they do with B&W.

You can see it with your own eyes, really. Every C-print I've ever seen in my entire life, from the Guggenheim to the gallery, has the same smudginess up close. I freelanced for the curator of a corporate collection and spent time examining Crewdson and Barney C-prints up-close. Guess what? Same story -- in the 3-5 lp/mm range.

Personally I don't really care about microdetail in prints. I care more about texture close-up, and I've found that C-prints are smoother than inkjet, which have a distracting stippling texture.

[Oren Grad]

Oh, Ctein said it so it must be true. Nonsense. The paper may give a (very) smudgy rendering of the bar chart, but it really will come down to what you define as "detail."

Have you replicated his test, with a careful contact printing of a precision test target and inspection of the prints under a microscope?

I'm not disputing what you've seen in practice - chromogenic prints usually look fuzzy to me too. It's just that I think you're way underestimating the extent to which the losses are due to the enlarging process, whether traditional conventional optical projection or Lambda/Lightjet etc., rather than the inherent properties of the print material.

And yes, to some extent the distinction is academic. For the print "consumer", it's a systems problem, not a paper problem - if the detail can't be gotten on to the paper with the tools and methods commonly available to the operator doing the printing, then for practical purposes it's unavailable. What somebody else can achieve under highly controlled laboratory conditions doesn't matter.

[Brian Ellis]

Awesome B&W photography, but the examples on his website look way too over sharpened to me, almost to the point of looking like illustrations.

I agree, some of the detail - in the trees for example - looks oversharpened to me as well. That could be a function of viewing a computer monitor rather than a print though I would have thought offhand that the opposite would occur (i.e. that any over-sharpening would show up more on a print than on a monitor).

It also could be caused by the fact that he sharpened the entire photograph by the same amount (i.e. he didn't sharp different areas of the print differently) because to me the areas that look oversharpened are mainly the areas of fine detail such as the leaves on the trees, which is where oversharpening is often most obvious.

I hate the over-sharpened look that you sometimes see in digital prints (not saying these are over-sharpened, they just look that way to me on my monitor). To me nothing says "this is a digital print and I didn't know what I was doing" quite so clearly as an over-sharpened print. When I do the final sharpening I get it to the point that I think is right and then back off about 10%. I try to be a certain as possible on the first try that no parts of the print look over-sharpened.

[paulr]

Paul, in the same article he reports why resolution greater than 10 lp/mm matters, up to a point. The gist of it is that it's possible to tell the difference between a square-wave pattern and a sine-wave pattern at 10 lp/mm. Mathematically, the bulk of that difference is contributed by the third harmonic, which means that somehow information at 30 lp/mm is being perceived. This is consistent with his tests of the long-discontinued Ektaflex print material, which recorded 18-22 lp/mm but nevertheless was perceived to be less sharp than conventional Ektacolor paper. But 30 lp/mm is about as far as makes any perceptible difference.

Achieving greater than 10 lp/mm in pictorial prints, of course, requires a good negative to start with as well as very careful printing technique, whether for contact printing or enlargement.

I thinkg that's about right. It fits with the general research on visual perception. Although 30 lp/mm goes way beyond what anyone can see!

In my own tests, looking at the difference between prints band-limmited to 7lp/mm and 14 lp / mm, there's a slight difference in perceived detail. No difference in sharpness, all else equal. The biggest improvement is freedom from aliasing with slightly diagonal lines.

[paulr]

Oh, Ctein said it so it must be true. Nonsense. The paper may give a (very) smudgy rendering of the bar chart, but it really will come down to what you define as "detail."

I haven't read the ctein tests, but I've seen MTF tests of gelatin silver, chromogenic, and ciba papers. These materials are essentially very slow photographic films. The emulsions themselves can capture stupefyingly, uselelessly fine detail. The only limitations on their resolution come from surface texture, which in glossy materials is irrelevent ... any of these papers can resolve finer detail at much higher MTF than any negative that will be printed on them.

The reality is that the detail visible looks very close to the same file viewed at 50% in Photoshop. That would mean the print is really only about 150 DPI, or 3 lp/mm.

Most likely this has to do with the lambda machine's optical resolution.

This is consistent with my experience printing color photochemically. Looking at the image through the grain focuser, you can see details from the negative that just never make it to the print the way they do with B&W.

I'll deffer to people who have contact printing experience with c-print paper to comment. I've personally looked at c-print contact prints with a loupe and see more than with the naked eye ...

Personally I don't really care about microdetail in prints. I care more about texture close-up, and I've found that C-prints are smoother than inkjet, which have a distracting stippling texture.

I'm not interested so much in microdetail either. But I'd define that as detail you can't see without a loupe. A print that's band-limmited to 5 lp/mm is capable of looking very sharp (the most critical detail is in the 1 lp/mm to 5 lp/mm range). But if the subjact matter includes finer detail, you'll miss it. It would be fairly obvious in a well made print, under good lighting, viewed up close and side by side.

Ink jet prints made with the newest printers don't have visible stippling anymore. They tend to look different from c-prints, but my ability to guess which is which has dropped well below 100%.

And the monochrome ink systems are amazing. I can't see ink dots on my piezo prints even with a loupe. The tones are as smooth as platinum prints.

[Scratched Glass]

After many years with my D100 I'm tired of the resolution, and can't afford the high end camera and lenses that might make me happy. My solution was to buy a speed graphic, lenses and accessories and a high end point and shoot for my casual photos. It is very early in this experiment and I'll see how it goes.

[Paul Kierstead]

I really think that for the vast majority of shooters, hand them a DSLR and an LF camera, get them to go on separate shoots with each, they will come back with pretty different work.

The exception might be for those with a highly 'refined' vision or style who actively strive to maintain that style regardless of equipment. And even then, let them use the stuff for a while, their vision will alter a little.

I'm not saying one or the other is 'better' or 'best' (how I hate that word in most contexts), but that the differences in results extend beyond the technical merits/aspects. You can't *really* sub one for the other; the results will change.

[John Brady]

I have just recently stated shooting with digital along side my 8x10. Each is a tool capable of doing amazing things. I am able to do things with LF film that I can't do with digital and vice a versa.

With digital I am using a canon 5d II and the 17mm tilt shift lens. I am stitching 3 exposures of 3 stitches for a total of 9 exposures. I am blending the exposures, I don't like hdr. Here is an example.
www.timeandlight.com

[Ben Syverson]

I'll deffer to people who have contact printing experience with c-print paper to comment.

That would be me.

Trust me, C-prints are not finer grained than film.