If, back in the 1920s, you could have been the decider of the 35mm format ...

[David Lyga]

24x43mm would have been nice. 16:9 aspect ratio. But nobody was into that back in the 1920s!

To construct lenses for that (to me, anyway) wasteful format would have been very difficult. - David Lyga

 

[pentaxuser]

I too would have liked to see standardization on a 24mm x 32mm frame.
Same aspect ratio as 110 and micro 4/3 .
And 6x4.5 as you mentioned.

Matt, I had just given this thread a little thought before looking at it again and was going to say that my prints from 35mm negs on 5x7 paper and my 645 negs always about right then I see your almost identical thoughts

pentaxuxer

 

[koraks]

To construct lenses for that (to me, anyway) wasteful format would have been very difficult. - David Lyga

Well, plenty of lenses around in that day and age for half plate, full plate etc.

 

[Donald Qualls]

So:
35x35mm and no sprocket holes.

You'd need to have some rebate, for handling and edge protection (828 has about 1.5 mm on the non-perf side, 127 has 2mm or so on each edge, 120 is 3mm edge rebate). So you'd wind up with about 32x32 or 33x33 at most on a 35mm strip. You'd have essentially created a little brother for 127, in a market that was still mainly aimed at contact prints (127 was called "small format" back then; full frame 4x6 was about as small as a print could be and be easy to view in hand). With no-perf film, and no cassette, you'd have had to give it a paper backing, and that would have made it fiddly to load (I have an 828 camera and cut 120 for it -- I know). Then it'd be fiddly again, winding by red window, or the camera would have been expensive with a friction roll length counter or turns-counting cam system, and you'd be locked into a single roll length (8-12-16 exposures at most). Why not just shoot 127?

To construct lenses for that (to me, anyway) wasteful format [24x43mm] would have been very difficult. - David Lyga

Nothing difficult about it. Every 127 camera had a lens with larger coverage than that, never mind the plethora of cameras for larger film formats -- you must be mentally applying other restrictions. Perhaps "to be part of a simple, cheap camera (lots of 120 cameras with very simple, cheap lenses)", "to be very compact (127 folders smaller than a Barnack Leica)", or you're considering a narrower format "wasteful" of image circle -- which isn't really a limited resource. Our 35mm cameras would have had a larger lens mount, for those with changeable lenses, to give an opening that wouldn't vignette the wider frame, and there would have been fewer frames on a given length of film, but there's nothing particularly wasteful about that.

In the 1920s to early 1930s, the real issue is that there was no Cinerama, few extra-wide format cameras of any kind (banquet cameras from glass plate era were already mostly gone, and motor-panning panoramic cameras are whole different genre). No one had learned to "see" in 16:9 at the time -- the 3:2 ratio of 35mm double frame ("eight perf") was as wide as anyone used (same, roughly, as 6x9 on 120 or 4x6 on 127).

 

[BradS]

To construct lenses for that (to me, anyway) wasteful format would have been very difficult. - David Lyga

The diagonal of 24x43 is 49.25
It does not seem like that would have been very difficult at the time....at least no more difficult than the 5cm Elmar that they did use.

 

[Helge]

You'd need to have some rebate, for handling and edge protection (828 has about 1.5 mm on the non-perf side, 127 has 2mm or so on each edge, 120 is 3mm edge rebate). So you'd wind up with about 32x32 or 33x33 at most on a 35mm strip. You'd have essentially created a little brother for 127, in a market that was still mainly aimed at contact prints (127 was called "small format" back then; full frame 4x6 was about as small as a print could be and be easy to view in hand). With no-perf film, and no cassette, you'd have had to give it a paper backing, and that would have made it fiddly to load (I have an 828 camera and cut 120 for it -- I know). Then it'd be fiddly again, winding by red window, or the camera would have been expensive with a friction roll length counter or turns-counting cam system, and you'd be locked into a single roll length (8-12-16 exposures at most). Why not just shoot 127?

Yes and no.
The cheapest and easiest option, is regular guide rails and pressure plate.
But you could have flat glass and even alternatively a suction plate.
The trouble with glass pressing the film flat is of course as always keeping it clean and dust free (as anyone using glass slides will tell you).
But that could have be solved well enough then, and even better during the latter half of the twentieth century.

 

[Donald Qualls]

The trouble with glass pressing the film flat is of course as always keeping it clean and dust free (as anyone using glass slides will tell you).
But that could have be solved well enough then, and even better during the latter half of the twentieth century.

But this thread is about different paths someone in Oskar Barnack's position could/should have taken, with the technology of 1925-1930, and the knowledge of the market that was available then.

How would a film consumer load a roll of film without paper backing (for that matter, how was a Leica loaded before their proprietary cassettes, and later the interoperable Zeiss ones, both of which preceded the universal Nagel cassette we still use)? Sure, daylight rolls existed by that time for movie cameras, and they didn't require a spliced-on opaque leader -- the outer layers of film would protect the rest. But with only five feet or so of film on what had to be an open spool? Just letting the tension off would fog the whole roll, at least at edges and through sprocket holes. Wrapping the film in backing paper was the 20th century solution (until Leica, Zeiss, and Nagel).

 

[David Lyga]

The diagonal of 24x43 is 49.25
It does not seem like that would have been very difficult at the time....at least no more difficult than the 5cm Elmar that they did use.

What I meant was the the coverage would have had to be 43mm X 43mm. That would make FAST lenses difficult to make. 35mm thrives on 'fast'. - David Lyga

 

[Donald Qualls]

Coverage would actually need to be about 50+mm diameter for a 49+mm diagonal. And no, FAST lenses wouldn't be "difficult", they'd just be larger and more expensive. And more complex than a Tessar or Heliar (though a 55mm Tessar would cover that circle, and could be made at f/4.5, even f/3.5 by 1930).

Once again, in the 1920s and 1930s, f/3.5 was a fast lens -- most cameras ran from f/4.5 to f/6.3, if not slower (despite slow films -- few worried about 1/25 shutter at the time). The push to f/2.8 and beyond didn't occur until after Leitz and Zeiss had created and sold (incredibly expensive, for the day) lenses in that class for their 35mm cameras.

 

[AgX]

What I meant was the the coverage would have had to be 43mm X 43mm. That would make FAST lenses difficult to make. 35mm thrives on 'fast'. - David Lyga

The necessary coverage for 24x43 would have been lesser than that for 43x43, as the respective diameter always is same as a format's diagonal.
Whether the differnce in this case is substantial for achievable lens speed, I doubt.though.

 

[BradS]

What I meant was the the coverage would have had to be 43mm X 43mm.

Well, no. That's not correct...but even if it were correct, the diagonal of 43x43 is only 60.8 so, still not really 'very difficult'.

That would make FAST lenses difficult to make. 35mm thrives on 'fast'.

...and yet, f/3.5 was apparently fast enough.

SLR's need a fast lens but rangefinders certainly do not.

 

[Bill Burk]

I think the real frame size of 35mm film is 24 x 30. (The crop necessary to make equal border prints on 8x10). Isn’t that the sweet spot you mentioned?

 

[Bill Burk]

Your thread also made me realize there might be a certain audience-demanded quality and it might be around the image quality you get when cinema uses 5222 and when still shooters shoot Tri-X.

I’ve been wondering why 5222 was the cinema standard.

 

[Helge]

But this thread is about different paths someone in Oskar Barnack's position could/should have taken, with the technology of 1925-1930, and the knowledge of the market that was available then.

How would a film consumer load a roll of film without paper backing (for that matter, how was a Leica loaded before their proprietary cassettes, and later the interoperable Zeiss ones, both of which preceded the universal Nagel cassette we still use)? Sure, daylight rolls existed by that time for movie cameras, and they didn't require a spliced-on opaque leader -- the outer layers of film would protect the rest. But with only five feet or so of film on what had to be an open spool? Just letting the tension off would fog the whole roll, at least at edges and through sprocket holes. Wrapping the film in backing paper was the 20th century solution (until Leica, Zeiss, and Nagel).

It would have been perfectly doable in 1920 to have a sheet of glass holding the film flat.
You have had to instruct people in being extra careful in keeping things clean and dust free when film was loaded.
It wouldn’t be more of a problem than the sensor on modern SLRs and mirrorless.
Perhaps even less because you’d be able to clean the glass pressure plate more easily.

The main extra complication would be in having the pressure on the plate released and reapplied as a part of the winding of the film. Either manually or automatically.

Counting the frames would be a bit harder and would have to rely on either counting the turns of end wind handle or having a roller on the back of the film.
Spacing wouldn’t be quite as tight as on a sprocket camera, but you should only loose about a frame or so.
If you had to allow for one millimeter more of slack space between frames, on a roll of 36 that would happen to be one frame lost, from the one gained by the slightly less wide frame of the 35x35mm.
Though still with a bigger negative.

 

[David Lyga]

I think the real frame size of 35mm film is 24 x 30. (The crop necessary to make equal border prints on 8x10). Isn’t that the sweet spot you mentioned?

Yes, MOST people want prints in the order of 24 X 32 aspect ratio. Your '"30" is also valid but I think that my "32" is better. It will not happen; the glue is set. Of course there are many exceptions but, if one wanted a 'standard' to use as an anchor, I think that the aspect ration of 4 X 3 is the best, overall.

I am very confused about one thing, however. If a lens for a 35mm camera is designed to cover a 24 X 36mm film plane area, how could that same lens NOT also be ready, also, to cover a 36 X 36mm square area? The lens's elements are circular and MUST cover the left and right (long) sides of a 24 X 36mm film plane. But that lens "does not know" that the upper and lower limits of coverage requirements are less severe (only 24mm film plane width). Again, that lens's optics are circular, so there is no built-in bias for truncated coverage up and down the film plane's width. In summation: I do not see how anyone can truthfully say that a normal lens for a 35mm camera will NOT cover a potential 36 X 36mm format THE SAME WAY that it covers a 24 X 36mm standard format? If it can cover "left to right", if can cover "north to south" the SAME WAY. - David Lyga

 

[David Lyga]

It would have been perfectly doable in 1920 to have a sheet of glass holding the film flat.

Sure, on the back, but what about the front of the emulsion? If the glass were on the front and the back was secured by a metal pressure plate, the introduction of "in focus" dust would have had the ready potential to be rather profound. And "telling" people to be "careful" would not be the same as telling them to imagine themselves as ants who can see dust as giant boulders. - David Lyga

 

[BradS]

I am very confused about one thing, however. If a lens for a 35mm camera is designed to cover a 24 X 36mm film plane area, how could that same lens NOT also be ready, also, to cover a 36 X 36mm square area? The lens's elements are circular and MUST cover the left and right (long) sides of a 24 X 36mm film plane. But that lens "does not know" that the upper and lower limits of coverage requirements are less severe (only 24mm film plane width). Again, that lens's optics are circular, so there is no built-in bias for truncated coverage up and down the film plane's width. In summation: I do not see how anyone can truthfully say that a normal lens for a 35mm camera will NOT cover a potential 36 X 36mm format THE SAME WAY that it covers a 24 X 36mm standard format? If it can cover "left to right", if can cover "north to south" the SAME WAY. - David Lyga

In fact, most users insist that the lens cover all the way out to the corners of the frame....the circle of coverage must be large enough to enclose the full rectangle.

 

[David Lyga]

In fact, most users insist that the lens cover all the way out to the corners of the frame....the circle of coverage must be large enough to enclose the full rectangle.

Good, corner to corner is MORE than 36mm in length. What I am getting here is the fact that "corner to corner' within a 36 X 36mm square is MORE than "corner to corner" within a 24 X 36mm rectangle. The manufacturers HAD to make the coverage match that extended diagonal distance emanating from the 36 X 36mm square, NOT the diagonal emanating from the 24 X 36mm diagonal. Do I err and why?

UPDATE: Thinking about this has made me think a bit differently. Maybe, since the slightly smaller diagonal is ALL that must be covered with the 24 X 36mm format, then that is ALL that must be covered. If so, I stand corrected and am happy that a bit of common sense has finally been allowed to venture into my old, old brain. - David Lyga

 

[BradS]

Good, corner to corner is MORE than 36mm in length. What I am getting here is the fact that "corner to corner' within a 36 X 36mm square is MORE than "corner to corner" within a 24 X 36mm rectangle. The manufacturers HAD to make the coverage match that extended diagonal distance emanating from the 36 X 36mm square, NOT the diagonal emanating from the 24 X 36mm diagonal. Do I err and why? - David Lyga

Yes, (I think) you err.

Draw yourself a diagram. Draw a rectangle inscribed in a circle. You will see that the diagonal line connecting the upper left corner to the lower right corner (for example) is the diameter of the circle and is larger than the length of either side of the the rectangle. This is the minimum dimension that the lens must cover.

For a lens to cover 36x36mm the diameter of coverage would have to be at least 51mm
but to cover 24x36mm the diameter of the circle of coverage is only 43mm

Does that make sense?

EDIT: I think that you're overlooking the fact that the diagonal of 24x36 big plenty big enough to cover the 36mm dimension ?

 

[David Lyga]

In other words, a 'corner' of a potential 36 X 36 format, ON STANDARD 35mm CAMERAS is lower than the 36mm height. Thus, the need for coverage becomes a tiny bit less, as the 'corner' has to be ONLY 24mm in length. Sorry, I could not understand this at first, but this is not crystal clear. - David Lyga

 

[AgX]

I am very confused about one thing, however. If a lens for a 35mm camera is designed to cover a 24 X 36mm film plane area, how could that same lens NOT also be ready, also, to cover a 36 X 36mm square area?

Before spending a thousand words: take a sheet of squared paper, draw some rectangles of different aspect ratio, take a divider and draw circles around them. Or draw circles and move cut-outs of various aspect ratios around.

 

[David Lyga]

Yes, (I think) you err.

Draw yourself a diagram. Draw a rectangle inscribed in a circle. You will see that the diagonal line connecting the upper left corner to the lower right corner (for example) is the diameter of the circle and is larger than the length of either side of the the rectangle. This is the minimum dimension that the lens must cover.

For a lens to cover 36x36mm the diameter of coverage would have to be at least 51mm
but to cover 24x36mm the diameter of the circle of coverage is only 43mm

Does that make sense?

You say, "Does that make sense"? That makes EMBARRASSING sense. This is an example whereby I somehow mentally obscured the fact that it is the DIAGONAL, not the "east / west" or "north / south" distance which counts. I, stupidly and naively, was thinking solely in terms of the distance from one side of the frame to the other!!! NO, that lens does not see distance that way. It sees distance as the furthest possible from its dead center, and from that dead center, THAT MEANS that the diagonal is ALL THAT COUNTS. Thank you for helping me to reinvent the wheel. I was not thinking properly. - David Lyga

 

[David Lyga]

Before spending a thousand words: take a sheet of squared paper, draw some rectangles of different aspect ratio, take a divider and draw circles around them. Or draw circles and move cut-outs of various aspect ratios around.

AgX, I think that you know by now, that David Lyga would too often rather spend freely with prolix verbiage than with simplistic common sense. - David Lyga

 

[koraks]

It would have been perfectly doable in 1920 to have a sheet of glass holding the film flat.

It would have brought additional challenges with flare and light piping. I know because I tried.

 

[MattKing]

I am very confused about one thing, however. If a lens for a 35mm camera is designed to cover a 24 X 36mm film plane area, how could that same lens NOT also be ready, also, to cover a 36 X 36mm square area? The lens's elements are circular and MUST cover the left and right (long) sides of a 24 X 36mm film plane. But that lens "does not know" that the upper and lower limits of coverage requirements are less severe (only 24mm film plane width). Again, that lens's optics are circular, so there is no built-in bias for truncated coverage up and down the film plane's width. In summation: I do not see how anyone can truthfully say that a normal lens for a 35mm camera will NOT cover a potential 36 X 36mm format THE SAME WAY that it covers a 24 X 36mm standard format? If it can cover "left to right", if can cover "north to south" the SAME WAY. - David Lyga

David.
You are analyzing the mid-points on the edges, when it is the corners that create the challenge.
Draw a diagonal line from corner to corner on a 24mm x 36mm frame. That line is ~43.3 mm long, and is the minimum diameter of a circle of coverage that minimally covers the 24mm x 36mm frame.
Now draw a diagonal line from corner to corner on a 36mm x 36mm frame. As that frame is taller, the diagonal ends up being longer - ~50.9 mm long, and is the minimum diameter of a circle of coverage that minimally covers the 36mm x 36mm frame.
When you compare the distances between the exact centre of the frame and the outermost corners (not edges) the distance on a 24mm x 36mm frame is about 8.8mm closer than on a 36mm x 36mm frame.
EDIT: I now see that others have help you clarify your quandary. Oh well!