I appreciate your help.
Good point about LF contacts. Morgan and Lester may not be right. I can't prove what those two Leicaphiles said about 1/30th degree, nor do I see it mentioned elsewhere on the web yet. I agree with the gentlemen above who suggest that there's definitely more to this than a simple digi/film comparison.
I suspect, but haven't tested, the eye's ability to sense much finer detail than scientists claim. It takes a lot of halftone dots per inch to make a reproduction that feels much like the original photograph. Of course other factors than resolution are involved in such a comparison. If the 1/30 degree resolution cited above is absolutely valid, there would be less need for LF contact prints.
I note that up-thread you wondered if the early Elmar might have had about 40 line pairs per millimeter. That would translate to about 80 pixels per mm. So, then, (24mm X 80 pixels) times (36mm X 80 pixels) would suggest a capability of the camera in the neighborhood of 5.5 Mpxls ... if the film then available could resolve it also, and if all those other variables were not overbearing.
If so, I am left thinking that it might be that the early Leicas which were thought to be "sharp enough" and "enlargeable enough" to enter the service of National Geographic, LOOK, and LIFE, just might have been capable of what we now think of as 5 or so megapixels.
Consider Lester and Morgan's idea about 100 lines per inch in the finished print, viewed at no closer than ten inches. At 40 lines per mm in the negative, that implies potential for a 14.4 inch print in the long dimension, and which would be "perceived as sharp" when viewed at ten inches away. I think I would probably view a 14 inch print further away than that. I will guess that is about as large as an image would have appeared in LOOK, National Geographic, or LIFE.
I'm assuming that the "line" of 1956 is the "line pair" of today, and that a line pair is approximately the same as 2 pixels. The "lines" described by Lester and Morgan have white spaces between them, but they do not use the term "line pair" as we do today.
Last edited by T42; 12-28-2010 at 12:32 AM. Click to view previous post history.
A Certified Dinosaur
Nikons F, F2, D700, Leica M3, & Kiev 4a
The question you raise can be generalized in how many pixels a film camera, with high quality optics and high resolution film, would have if compared with a modern digital camera. It is an interesting question indeed for the technically minded.
I'll jump very quickly over all the obvious stuff about the fact that resolution does not art make and post here some interesting observations that somebody else made on the subject.
Norman Koren has an extensive essay:
Ken Rockwell makes some interesting considerations:
Another simplified text quoting Ken Rockwell:
I find this text by Vitale to be extremely interesting:
I would like to add two personal considerations:
My slides scanned at 4000 dpi with a Nikon Coolscan 5000 ED yield scan files of around 108 MB 16 bit per channel. That equates to 54 MB at 8 bit per channel which corresponds if I get it right to a JPEG obtained with a 18 megapixels camera. I see no "wasted pixels" on my scans. When I view images along diagonal edges at 600% or so on the monitor, so that I can see each scanned pixel as a square on the monitor, I can clearly see no "double pixels", each pixel is different from the ones near to it and they all contribute to the image. I see that each pixel is useful to the definition, there are not pixels which do not contribute to the definition of the image, so my personal take is that 100 modern high-resolution slide film like Astia 100 is certainly capable of deliverying a resolution comparable to a 18mp camera. That's comparing digital to scan, which puts scan at a disadvantage (nobody tried to compare digital and film the other way, by obtaining a slide from a digital image, and comparing it with a film slide).
On the other hand, stock agencies like Photolibrary which would not accept images below 12 mp go on accepting scanned 35mm images, not necessarily drum scanned. That poses film scans above the 10 mp league.
So I think the equivalent resolution of film, by my impression and judging by industry practice, is more or less in the 15 - 20 mp range.
Questions like the ones raised here are only meaningful when considering prints. Given how undemanding even the highest resolution monitors are, it seems ironic that the questions are even asked when you consider how (relatively) few digital files ever end up getting printed.
“Photography is a complex and fluid medium, and its many factors are not applied in simple sequence. Rather, the process may be likened to the art of the juggler in keeping many balls in the air at one time!”
Ansel Adams, from the introduction to The Negative - The New Ansel Adams Photography Series / Book 2
About eyesight resolution, I found in several astronomy books (concerned with optical instruments) that it went down to 3/4 of one minute arc (one minute arc is 1/60 of a degree) so approximately 1/80 of a degree. This is meant to be the "grain" of the human retina at the spot called the "fovea" - and which is considered to be relatively constant from one person to another, and gives the limit of the best sighting when our eyes are optimally focused.
Last edited by polka; 12-29-2010 at 04:04 PM. Click to view previous post history.
Thanks for your experience based insight and for those links. That was very useful reading.
Good point. I guess I am just a curious person wanting to get a grasp on "how good" the early cameras were which made those first 35mm pictures in LOOK, LIFE, and National Geographic. I suspect that they would not translate into much in terms of megapixel equivalence, but it's just a guess.
That's very good data which I had not heard about. 1/80th of one degree. The authors of the Leica Manual of 1956 say it's 1/30th of a degree. And they conclude that most folks can only resolve about 100 lines per inch (25.4mm) when viewing a print at ten inches (254mm). 1/80th of a degree seems to turn that idea on its head. I'm assuming that the "lines" in the 1956 Leica Manual are equivalent to today's "line pairs," since the "lines" described had bands of white space between each of them.
Thanks again for the help.
A Certified Dinosaur
Nikons F, F2, D700, Leica M3, & Kiev 4a
Sponsored Ad. (Subscribers to APUG have the option to remove this ad.)
It would be interesting to know how such maximum eyesight translates to the opticians recommendations. In Sweden - and probably elsewhere - they strive for 1.0 eyesight when adjusting eyeglasses. It means that one can resolve the smallest line of letters on that traditional poster on the wall. 0.9 or 0.6 means that one can resolve the next smallest or the sixths smallest line.
Myself, I can see none of those letters without glasses or contact lenses, but with them I exceed the recommendations considerably (1.5 eyesight they told me.)
The eye constantly moves, and perhaps superimposes image upon image much as video does. If large projections of grainy movies appear reasonable sharp, perhaps our eye and brain accomplishes the same. Just a thought, with no scientific citations to support it.
Originally Posted by polka
At my age all these numbers have merely academic value...
That's very good data which I had not heard about. 1/80th of one degree. The authors of the Leica Manual of 1956 say it's 1/30th of a degree. And they conclude that most folks can only resolve about 100 lines per inch (25.4mm) when viewing a print at ten inches (254mm). 1/80th of a degree seems to turn that idea on its head.
having used a 1938 Leica I can honestly say that the quality is perfectly fine and although not as painfully sharp as a cannon 50D when scanned on an imacon you get a very fine 55meg file, I have no idea about the teccy stuff but I do know that the look of the Leica is perfectly good for most things, and has a fingerprint that is a lot more sesitive than most if not all digital captures to give you an idea of what I mean the files on this link below are all taken with a Leica II
just googled this which agrees with what I knew from my astronomy books : see his definition of Blackwell's "critical acuity" as the resolution of a spot as a non-point source (which is the thing that matters most to astronomers).
... and a lot of other very interesting related data in this document
Last edited by polka; 12-30-2010 at 03:31 PM. Click to view previous post history.