Purely Theoretical Musings on Resolutions and Formats...

[Scott Rosenberg]

i have a question regarding the amount of information captured on film, as i am new to such ponderings. i have spent the past many years blissfully shooting sheet film and happy as a clam in mud doing so. however, due to some recent overseas travels i have also started working with a mamiya 7 system. the 6x7 outfit was bought for times when i could sacrifice a little image quality for the convenience of shooting with the small RF. at least that's what i initially thought. however, much to my astonishment, the prints i was able to make from the M7 held up to enlargement as well as 4x5 even when printed at a decent size (30" in the long dimension). i was quite surprised by this so started doing some thinking on the matter and came up with the following reasoning...

the mamiya 7 lenses at f/8 are capable of resolving somewhere around 110 lp/mm
most large format lenses, such as the 110xl, at f/22 are capable of resolving somewhere around 60 lp/mm

i choose these as these are common working apertures for me. now, taking the approximate film size into account, one gets the following...

4" x 5" at 60 lp/mm = 46,451,520 lp
6cm x 7cm at 110 lp/mm = 50,820,015 lp

am i correct in assuming that the relative closeness of the amount of information captured by the respective formats explains the similar quality of the prints? it seems to my simple mind that if there is a comparable amount of information on the film then the enlargements are going to be close, regardless of the size of the original.

but then i am likely oversimplifying things here and hoped to elicit corrections from those more knowledgeable than i.

i appreciate any inputs,

[clay harmon]

Don't forget that another variable is the ability of the film itself to resolve the line pairs. Tech Pan in 6x7cm is likely to outresolve Royal-X Pan in 4x5.

[Donald Qualls]

Faster lenses (three stops faster common working aperture) in the 6x7 more than offset the advantage of larger film with respect to film grain -- ISO 100 in 6x7 has about the same grain (as found in the final print the final print) as ISO 400 in 4x5, but the 6x7 can shoot one stop faster shutter even with the slower film, if both are at their common working apertures.

I got a similar shock today, scanning some 6x9 on 120 negatives I shot in my Speed Graphic with my recently acquired Adapt-a-Roll 620. The 120 (even at ISO 400) actually looked better than what I've been getting with ISO 400 sheet film. I finally realized it was different brand film; the rolls are Foma, while the sheets are Forte, and there's a big difference in grain between Foma 400 and Forte 400...

Do a similar comparison with still faster lenses and still slower film and you might find 35 mm even competes -- Tech Pan (or one of the occasional partial replacements that pop up, like Bluefire Police) in 35 mm at f/4 may well outperform Plus-X in 6x7 at f/8. And then where are you?

[Paul Metcalf]

am i correct in assuming that the relative closeness of the amount of information captured by the respective formats explains the similar quality of the prints?

But when you enlarge, you have to enlarge more for the smaller format, so you need more resolution. But I suspect your overall shoot-to-print process probably has something that is more critical for producing quality prints so the resolution differences don't really show up between MF and LF.

[Scott Rosenberg]

clay:
i haven't done any testing yet (planned for this weekend) but i've heard that the 25 speed rollei films are tops in resolution. i couldn't locate any tech pan, but the 25 pan is supposed to be very good, and the 25 ortho is supposed to be as good as tech pan with a claimed resolution of 330 lp/mm.

donald:
good point re: grain size relative to the overall area of the film - i hadn't considered that. as i mentioned, i'll be trying some 25 speed stuff this weekend to see how that influences what i'm seeing from 6x7. it's size relative to the area should be basically the same as fp4+ in 4x5... it sure is good to have options.

paul:
i understand what you're saying re: enlargements and you bring up an excellent point to which i don't know the answer. of course a 6cm x 7cm film will have to be enlarged more than a 4" x 5" film to get each to the same print size. the enlargement delta being comparable to the delta in the film area - assuming each piece of film contains the same amount of information. however, if instead of thinking in terms of the size of the original we think in terms of how much information the original contains, does it stand to reason that two pieces of film with the same amount of information enlarged to the size print will require the same amount of enlargement regardless of the area of the film from which they were derived? that is to say, if film A contains 10 units of info and film B contains 10 units of info, and if i am going to make a print that will require 20 units of info, then both will be doubled regardless of the size of the film?

again, i might be oversimplifying things, but if i think in terms of how much information i've got on the film versus how much i'll need to enlarge based on the ration of the areas, does it matter what size the original is? or does it only matter how much information each has? i have no idea, but will be testing for this.

in the case of the 4x5 with LF optics (~60 lp/mm) versus 6x7 with M7 optics (~110 lp/mm) the increased resolution of the M7 glass seemingly makes up for the reduced film size, and the amount of information contained herein is roughly comparable - 4.6 x 10^7 lp versus 5.0 x 10^7 lp. i would assume, therefore, that when enlarged to the same print size, the prints would be roughly identical. in fact, if this reasoning is sound, the 6x7 film would need LESS enlargement, as it contains slightly more information.

again, i've only recently become concerned with such matters and am hoping some of you with more experience and knowledge can correct any flawed logic.

[Michael Alpert]

i would assume, therefore, that when enlarged to the same print size, the prints would be roughly identical. in fact, if this reasoning is sound, the 6x7 film would need LESS enlargement, as it contains slightly more information. . . . i've only recently become concerned with such matters and am hoping some of you with more experience and knowledge can correct any flawed logic.

The same print size would mean a higher degree of enlargement for the smaller negative. Small prints (enlarged to about 4x5 inches from the 6x7 negative) will look the same because, in this size, the limiting factor is the resolution of the photographic paper, assuming that the negatives are very sharp.

Larger prints will always favor the larger negative. The greater resolution of the medium-format lenses and finer-grain film are will not save prints made from smaller negatives from becoming soft when compared with larger negatives. The measure of resolution, in practical terms, is the contact print. When making enlargements one needs to learn what to accept (i.e., what one's personal quality-standards are) by comparing enlargements against contact sheets.

[Ron Marshall]

Photodo used to have, and may still have, a comparison of the resolving power of 35mm, 6x6 and LF, under optimal conditions. They measured the resolution obtained on TMY of each format. LF won, but not by much. The greatest benefit, according to them, for a larger format is smoother tonal gradation and less grain, assuming constant print size.

[Bruce Watson]

The same print size would mean a higher degree of enlargement for the smaller negative. Small prints (enlarged to about 4x5 inches from the 6x7 negative) will look the same because, in this size, the limiting factor is the resolution of the photographic paper, assuming that the negatives are very sharp.

Larger prints will always favor the larger negative. The greater resolution of the medium-format lenses and finer-grain film are will not save prints made from smaller negatives from becoming soft when compared with larger negatives. The measure of resolution, in practical terms, is the contact print. When making enlargements one needs to learn what to accept (i.e., what one's personal quality-standards are) by comparing enlargements against contact sheets.

Indeed.

And there's more to photography than just sharpness. One of the other ways a bigger negative shines is in tonality. Having more film area helps make smoother tonal transitions and gives a larger range of tones. IOW, when you make same size prints the one from the larger negative usually looks smoother. That combination of excellent sharpness with excellent smoothness is one of the reasons I like LF so much.

All that said, there's no reason not to use a medium format camera when traveling if it will do what you want. And it sounds like it works for you, so go for it.

[Greg Miller]

Are you talking about wet prints or scanned / digital prints? If you are scanning, are you using the same scanner for both formats?

[Emmanuel BIGLER]

Since i'm using both 6x9 and (recently) 4x5" film sizes, I feel concerned by this discussion and an Idea I had was that it depends whether you scan and print your images digitally or if you enlarge them with a traditional enlarger.

My feeling is that even with a good scanner when you scan a certain area of film with more DPI the effect of grain increases like in a classical micro-densitometer analysis (definition of the RMS value, the smaller is the aperture, the bigger is noise). So in fact even if you can have a higher resolution 6x9 image, if you get more noise/grain in your scan, you are more or less cheating in the contest vs. 4"x5 where smaller noise will be recorded in the digital scan of a bigger image with the same final number of recorded samples. For exactly the same number of samples bigger samples in a 4x5" film-image generate a much less noisy digital file. And and it is easier with bigger samples to get an acceptable image with an amateur-grade flatbed in 4x5" than in 6x9 : in a sense for the amateur, scanned 4"x5" is the amateur medium of choice, scanned 6x9 being the choice of professionals who can afford a top-class scanner !

In traditional enlargements and prints, the grain pattern is actually transformed depending of the king of illumination (point source, classical bulb + condensers, diffuse/cold-heas/color head) and image transfer properties in the optics. Final noise in te analog image might or might not be visible in the enlargement, we are quite accustomed to incorporate the grain pattern in our analysis of a traditinal silver halide image. The issue of grain pattern and their transformation in a classical enlargement is more complex than in a micro-densitometer or a scanner, but the idea is : since we are speaking about resolution, can we actually get rid of noise in the final image assessment ?