Large Format Resolution

By David A. Goldfarb

Resolution isn't everything in large format (LF), and often the most distinctive LF effects involve using classic lenses or techniques that reduce resolution but produce a look that can't be produced any other way.

That said, I posted a demo a while back on my website to give some idea of what an 8x10" neg is capable of —at the link, use the links underneath the picture to click up through successive increases in resolution to get an idea of how much information is really there. Bear in mind that this was scanned on an old Agfa Duoscan that has an optical resolution of 1000 ppi (about 1/5th of what is practical on a drum scanner), and the largest image in the demo is 50% of full resolution. Even so, you can easily see individual bricks in the building next to the school bus. The film is T-Max 100, and the lens is a single-coated Goerz 12" (305mm) ƒ/6.8 Gold Dot Dagor from the 1940s or 50s—a fine lens, but not as sharp as a modern plasmat. So I haven't done anything really to go out of my way to maximize resolution, and still...well, most people who haven't seen an 8x10" negative are impressed by the demonstration.

And just to show you the interesting "low resolution" side of large format, below is a scan of an 8x10" albumen print from an 8x10" neg, made with a classic 360mm Voigtlander Heliar lens at a wide aperture (about ƒ/5.6).

__________________________

David

Comments

[Awake]

Your big B/W photo of the bridge has quite a few "Hot Pixels"... I would send the camera back to the manufacturer for repair. (Just kidding!!!)

An interesting aside, your largest sample (50%) displays on my screen as a smaller image than I get out of a Canon 5D when looked at as a 100% image. I guess this is a good illustration of how we can easily lose perspective of the true image size that we get from digital cameras and the way that they display on our computer screens. I am so tired of providing an 8x10x300 file to a client, only to have them complain about some small blurring somewhere when they look at it at 100%, not realizing that if their screen is 96ppi, they are seeing the image as a 24x30 print, and if the screen is 72dpi they are seeing the image as a 32x40 print. With large files everyone becomes a pixel peeper.

[christopher]

Awake: That 50% sample is well over 16MP, how does your 5D produce a bigger file than that?

[Fazal Majid]

It's really seeing an exhibition like "In the American West" by Avedon that brings it home, more so than Ansel Adams, in fact. Seeing the amazing level of detail on people is stunning.

[David A. Goldfarb]

Bear in mind that with a drum scan, I could get a file about ten times bigger than the 50% resolution file of the viaduct, and somewhat sharper with more dynamic range.

I should have mentioned also that I've used a fair amount of front rise on this shot to keep the lines straight without getting too much foreground. I wished I could have gotten a little more to avoid clipping the arch on the front right, but I was at the limit with this combination of lens and camera.

[JBrunner]

While a print (or pixel) sniffer might gravitate towards the incredible resolution, for me, the second image tells the tale. A lens much older than me, that offers a smoothness that can't be defined in normal terms, with no electronic machinations needed. Nor can the true communication of the lens be readily observed on a monitor. Only in a contact print from that giant neg can one really experience it. I have this lens as well. Using it is similar to sex.

[Niels Henriksen]

The new digital cameras are great technology wonders and really do produce great pictures and in many cases not better than the 35mm.

It is only with my 4x5 negatives that I get that magical feeling.

I don’t know why but handling the negative, using the loupe to see fine detail, just bring on a special feeling that I never get with the DSLR, even though I love the photos that come out.

I wish I had gone to 8x10 for the ability to contact print full size (8x10) as enlarging a 4x5 is a problem.

Niels Henriksen

[dwross]

David:
Your albumen portrait is stunning. I was thinking about asking you to try your superb craftmanship on handcrafted silver gelatin, but now I can't decide if that would be a legitimate distraction from the excellence that is your albumens. Thank you for setting the bar so high.

[David A. Goldfarb]

Thanks, Denise. I've seen part of one of Ron Mowrey's handcoated silver gelatin workshops here in New York, and it's definitely something I want to try at some point. I like the idea of getting an Azo-like tonality, getting more control over the contrast and paper base through handcoating. Meanwhile, though, I've been setting up the darkroom in a new apartment, so I feel I need to catch up to where I left off in the old place before trying to make room for a new process and a couple of extra gizmos for dealing with gelatin.

[RC]

Detail is one thing. One can get a lot of detail from smaller format, but the best expression I've seen to describe what you get more of as the format size increases is "tonal integrity". But as to just what that is in objective measurable terms, I'd be hard pressed to explain. Does anyone know better?

[Helcio Tagliolatto]

JBrunner
I think you got it: the second image really shows the best of LF photography.

Concerning resolution, take a look at this test made by portuguese LF photographer Nana Dias. Words in portuguese, but this don't causes any problem:
http://forum.fotografiabrasil.com/index.php?topic=22601.0

Helcio

[Q]

It is spelled Albumin. You don´t need to post this but seeing "Albumen" just makes me go nuts. It is the most abundant blood plasma protein.

[Michael Francis]

I second RC, one of the things I notice with large format images is that they tend to have a tonal smoothness that is rarely met with smaller formats - oddly this is often preserved even when reduced on the web. Perhaps it is a function of the information in the image? Is the signal to noise ratio just that much higher with LF images?

[Mike]

Q,
The photographic term is "albumen," meaning egg white and its protein, used in the making of the emulsion. The terms are related but they're not identical--albumen is an albumin, but not all albumin is albumen.

Mike J.

[ben]

The S/N ratio is the same in LF and smaller formats (given the same film, of course). The difference is that you typically don't need to enlarge LF as much as you with, say, 35mm, so the noise (grain) is much less apparent.

For example, if you blow up a 36x24mm section of an 8x10 negative to 8x10" and display it next to an 8x10" enlargement of a 35mm shot, they'll have identical grain characteristics. They're both 7X enlargements.

The difference is that with that amount of enlargement, the 8x10 negative could print at 56x70" if it wasn't cropped. If you printed the 8x10 neg at 24x30" (3X), the grain would be MORE than twice as fine as an 8x10 print from 35mm. In other words: extremely fine.

The comparison to the 5D's filesize earlier is funny. With a cheap flatbed, you can scan 8x10 negs at 2400 DPI and wind up with a 24000 x 19200 file. For the megapixel counters, that's 460 megapixels. Needless to say, it'll be a decade or three before you can afford a digital camera of that caliber. :) Hint: it won't take EOS lenses. :)

[Bob Tilden]

What a great portrait! With such a shallow DOF, how do you control the subject to keep the eye(s) in focus? Hot glue? Or is the key to put the subject in a stable pose?

[David A. Goldfarb]

If you're interested in the albumen process, take a look at http://albumen.stanford.edu

[David A. Goldfarb]

Focus? Good question. The process involves focusing, and then you have to insert the filmholder, close and cock the shutter, stop down the lens, and remove the darkslide, and wait for the right expression to make the photograph without being able to see the image on the groundglass. That's plenty of time for the subject to move out of the narrow DOF range.

I attach a string to the camera mount or the tripod with a knot on the end. The subject holds the knot to his or her nose or forehead while I set the focus. Then when I'm ready to make the exposure, I can check the distance with the string right before clicking the shutter. It feels a little silly sometimes, but it's very reliable.

Another method that I use when I use strobes and modeling lights is to pay close attention to the shadows on the face when I'm focusing. I'm usually looking at where the nose shadow falls on the upper lip. If the subject is seated or standing in one spot, they can't move far, and just by adjusting the head to cast the same shadows, I can usually be sure the subject will be in focus.

There are other options with other types of cameras. Cameras like the Linhof Technika or Speed Graphic have rangefinders usually. The Graflex reflex cameras are SLRs that can be focused just like a 35mm or medium format SLR.

[Q]

I apologize, Albumen indeed seems to be commonly used, on the other hand, in a recent exhibition of Atget´s work, they explicitly called it Albumin Paper.

[David Dyer-Bennet]

It's fun poking around at the XL image from the 8x10 scan. Thanks! As people have said, the image has a nice smooth quality; probably, as people have said, from low enlargement. Also I wonder if the lenses have different characteristics that make a difference as well.

[David A. Goldfarb]

Thanks, David. In that photograph, I was shooting around f:32 or maybe even smaller, past the point at which diffraction due to the small aperture becomes more important than the qualities of the lens as far as resolution goes, but the Dagor is a nice lens. Here's a portrait made with the same lens at about f:14--

http://www.echonyc.com/~goldfarb/photo/imng2002.htm

In a photograph with a lot of front rise, as in the shot of the viaduct, a small aperture with a wide lens like a Dagor reduces falloff of illumination in the corners of the image, and that's usually a worthwhile tradeoff for the ability to keep the lines straight and for a little loss of resolution.