Color images from the 20's and 30's

ZorkiKat · Jun 7, 2007

htmlguru4242 said:
You mention the AGFA approach. I realize that they made an early AGFAcolor film, but I have no idea how it worked.

Could you explain in a little more detail?

The additive 1932 Agfacolor plate used coagulated dyes instead of starch in the mosaic filter grid. Worked in the same principle as the Lumieres' Autochrome in terms of separating, recording, and reproduction of the hues.

The description described in Brian Coe's "History of Colour Photography" says that the filter grid starts out as a

"mixture of three solutions of dyes in colloidal suspension. The elements in suspension were immisible but would coagulate, so that when they were coated on glass or film and allowed to dry, the dyed droplets formed into individual red, blue, and green filters. Furthermore, because the elements dried in close contact with each other, there were no interstices to be filled in with carbon black as was necessary for the Autochrome. In fact, it claimed that whereas some 92% of incident light was absorbed by the filter layer of an Autochrome plate, only 86% was absorbed by the Agfacolor mosaic."

The less light lost from absorption made Agfacolor faster than Autochrome. The film version of additive Agfacolor was Agfacolor Ultra.

Agfacolor-Neu was the tri-layered subtractive version which had colour couplers in the emulsion layers.

AgX must be proposing that since the Agfacolor used different materials which no longer required the flattening and carbon filling steps in Autochrome,
the former may be more feasible to work with.

Another method would be forming the filter mosaic with red, blue, and green lines, like what Dufaycolor did.

Brian Coe's book also shows a comparison (at 50X magnification) of the various filter mosaics used by the different systems. Dufaycolor's (the grid made it more consistent in terms of filter distribution) and Agfacolor Ultra's mosaics were the finest in the lot.

Jay

Mark Antony · Jun 7, 2007

I don't know if anyone posted this guys work:
http://www.loc.gov/exhibits/empire/gorskii.html
Prokudin-Gorskii was a very (pre WWI) colour photographer.
Mark
http://www.photo-utopia.blogspot.com/

Michel Hardy-Vallée · Jun 7, 2007

Mark Antony said:
I don't know if anyone posted this guys work:
http://www.loc.gov/exhibits/empire/gorskii.html
Prokudin-Gorskii was a very (pre WWI) colour photographer.
Mark
http://www.photo-utopia.blogspot.com/

That's the name I was looking for a few posts ago!

htmlguru4242 · Jun 7, 2007

Intersting about the AGFAColor mosaic. That sounds like it would be a significantly easier method to carry out, as the methods are readily available with modern technology.

If I'm understanding correctly, it was just a suspension of 3 different color dyes which would coagulate together into a gapless filter mosaic, yes?

I am actually aware of the methods with lines; I've read everything I can online about Dufay color (and the Joly process, which preceeded it). I also made a colored line screen on my inkjet printer, and shot a piece of 4x5 film in front of it. The color quality was horrendous, due to my printer and alignment issues with the screen, but I could definetely see it working!

AgX, any more details on the AGFAColor methods would be wonderful!

AgX · Jun 7, 2007

Actually, the reason why I proposed that Agfa technique was because it could yield the chance to produce very small grains with a very limited range of sizes. This was the reason that I used the term emulsion and not suspension. The latter would mean that one brings already sized (resin?) particles into the coating fluid.

Concerning Coe, I faintly remember having read at some other author indeed that those resin droplets did not adhere to each other but that a flattening technique was employed.
Anyway, Coe's description does not reveal enough to even try to reproduce that technique.

A lot has to be found out...

I just found on the net a hint on a comprehensive German work on the restoration of this Agfa material. But it is only to be optained the classical way, as a hardcopy from the public library.

AgX · Jun 7, 2007

To my understanding dyes cannot coagulate.
The need a substrate. That resin in our case.

AgX · Jun 7, 2007

ZorkiKat said:
The additive 1932 Agfacolor plate...

Jay, that is not wrong in the actual sense, but still misleading concerning the birth of this Agfa technique:

1916
Agfa releases a color plate `Agfa-Farbenplatte´ basically similar to the Lumiere Autochrome plate (fully adhering resin grains in place of starch grains making the soot filling obsolete).

1932
A similar coating is done on film base (sheet and roll-film)(35mm only in 1935). These films are called `Agfacolor´, and the glass plate Farbenplatte also gets this new name.
To make things even more confusing Agfa releases the same year another type of additve material: their lens grid film. (for Contax and Leica cameras with 3colour stripe filters on apt lenses) This film is called Agfacolor too…

Thus additive grain grid films and plates and an additive lens grid film bear the same name.

1936
And then their subtractive chromogenic three-layer film is released. Yes, as you would expect, called Agfacolor… To make things a bit more clear it was called in the beginning `Agfacolor-Neu´.

ZorkiKat · Jun 8, 2007

AgX said:
1936
And then their subtractive chromogenic three-layer film is released. Yes, as you would expect, called Agfacolor… To make things a bit more clear it was called in the beginning `Agfacolor-Neu´.

Thanks AgX. The other references did say 'coloured resin' particles, not coagulated dyes. I could not find any of these to quote anymore so I used the readily available one, Brian Coe's book. The illustration of the magnified Agfacolor screen does show tiny coloured particles. Now if we could only find that type of resin....

I'm very interested in the Agfacolor subtractive process. I have a DVD of the 1943(?) "Baron Munchhäusen", restored of course. But even then, the charming colour palette of Agfacolor still shows through. It was said that cinematographers often disliked it (and its American equivalent, Anscolor) because of "muddy colours". The "muddy" description appears to be synonymous to subtle- since the garish (as in 'glorious Technicolor) was in fashion then.

I have an old Leica book which described two Agfacolor films- one which yielded colour transparencies and the other, colour negatives for prints. Quite advanced since they had colour negatives for 35mm long before Eastman Kodak made their own version. "Agfacolor" as a name seems to be really both abused and misused: too many types appearing under the same name!

Jay

AgX · Jun 8, 2007

Well, Agfa continously worked on their chromogenic film those first years. Which makes speaking of the capacity of that film a bit tricky.

Further, most images we see made from those early colour films are printed in books of that time have a second `photographer´, the lithograph...

Film based images which have survived time and are printed today are sometimes stunning. Well, perhaps time made them better... What I nwant to say is that looking back into time is always a difficult endeavour.

Going back into time ourselves. My preleminery thoughts on coughting a an additive mosaic.

I think in attempting to make a random mosaic yourself one would be faced by the following problems:

-) concerning the Agfacolor resin mosaic it could be that it hampers wet processing of the film. This can be tackled by putting the sensitive layer onto the mosaic. No problem for the industry; to us it would mean that we could not use industrial made film to coat our mosaic on.

-) concerning the Autochrome, how can a single-starch-grain layer be obtained? By wetting the film to make it sticky, powder our coloured grinded starch onto, brush it off and hope that just one layer keeps sticking?

-) concerning the Agfacolor resin mosaic: coating it with that emulsion, hoping that the sticky resin adheres to the base (base in most general sense), letting the rest run off, letting the emulsion dry, letting the solvent evaporate from that resin. Will that result in a single-resin-grain layer?

htmlguru4242 · Jun 8, 2007

You bring up some very interesting and very key points there, AgX. I think that for all intents and purposes, it will not be possible to coat any type of a mosaic over the emulsion of a film. The only thing that I can think of would be actually embedding waterproof dyes into the emulsion. This, however, poses a problem in two respects. First is actually finding suitable dyes and coating them. Second is the fact that, being waterproof, they will likely interfere with the processing of the film.

I think the only way to do this would be as it was done originally: coat the screen on a substate, seal it, and then overcoat the emulsion. This, obviously, poses another issue in finding a coatable panchromatic emulsion.

As far as getting a single layer, I have been able to do this using potato starch dyed black. If the starch is poured onto a sticky layer and then shaken off & brushed (I'll look up my exact procedure), it adheres in a fairly good monolayer. The issue is getting the grains to sit atop the adhesive, rather than being embedded in it.

I do have another idea, based on a suggestion by an APUG member awhile back. What about small plastic or resin pieces that melt? These could be placed in a layer on the plate, and then heated so they melt into a solid layer. I know this may have an issue with mixing of colors, but it is worth a try, IMO.

Roger Hicks · Jun 8, 2007

Martin Reed at Silverprint (an APUG sponsor) knows quite a bit about Autochromes and was at one time trying to re-create them, if I recall aright.

Cheers,

Roger

htmlguru4242 · Jun 8, 2007

Roger, that's good to know!

Any idea if he'd mind being contacted about it?

ZorkiKat · Jun 8, 2007

Lines instead of grains?

Going back to the "microfilter" issue- with the problems arising from making them stick and sit on the adhering surface, perhaps the Dufaycolor method is more feasible? It might be easier (relatively speaking) to use a reseau of evenly printed lines.

HTMLguru's inkjet-generated filter grid sounds interesting. I've been thinking about that method. The filter is not part of the film, and is used for both shooting and viewing. A suitable registration method would be needed. A similar device was shown and described in one of the books in the Time-Life Library of Photography. It was a filter which was installed in the film gate. Ordinary BW film would pass behind it. The same filter went with the film after it was processed to view the picture in colour.

Printed colour grids were also used by the last commercially-produced additive-screen colour film, the instant 35mm slide from Polaroid. I believe the reseau used it in resembled the layout found in a Trinitron-type colour CRT screen- tiny groups of RGB phosphors laid out in an array.

Jay

htmlguru4242 · Jun 8, 2007

ZorkiKat, I do like that idea. As I [may have] mentioned, I've tried it once, and it worked with a degree of success. With modern (or even older) printing techniques, it would be extremely easy to print a regular grid or arrangement of colored lines on almost any transparent substrate. Hell, your common 800DPI color laser could do it.

From a technical standpoint, this is significantly easier, and is probably a good first step.

I like the idea of a separate filter. There was a system called Paget plate that did exactly this. There was a separate exposure and viewing mask. It's discussed briefly at:
http://www.awm.gov.au/captured/colour/paget.asp

All examples that I see have online have low color saturation. To me, it looks significantly lower than even Autochrome, which isn't terribly saturated to begin with. If memory serves, this arose because the exposure mask was printed in extremely light colors, in order to shorten exposure times. I'm sure that they'd look better IRL, as most photos I've seen, regardless of process, do.

Identical screens could be used today for both exposure and viewing, due to the increased film speeds available to us.

There really are only 2 or 3 issues that I can think of with this. First is registration. (well, re-registration). After development, re-registering the filter will take a little bit of work. A regular screen will make this easier, but it will still take some time.

Second, I'd imagine that a regular grid or line pattern would give rise to some Miore effects. This is minor compared to other technical / practical hurdles.

Third, keeping the filter-plate and film in tight contact will require some type of clamping mechanism. I don't see this being an issue for testing, but for practical use, new film holders would need to be made.

Does anyone here have access to a high-resolution color laser that can print on transparencies?

htmlguru4242 · Jun 8, 2007

If either myself or someone else [bad grammar there] could prepare a "screen-plate" on a transparency, initial testing could be conducted in a contact-printing frame with projections from a slide projector or an enlarger.

Processing would be standard black and white reversal methods.

AgX · Jun 8, 2007

-) Instead of powdering on starch grains, one also could make a suspension of them in an hydrophobic volatile solvent, and coat this suspension onto the film.

-) Bringing dyes embedded in whatever substrate into the emulsion would hardly yield a sort of filter layer we all have in mind. (Though perhaps I’m lacking fantasy.)

-) Heating a film covered with meltable resins would strongly endanger the emulsion. Using high temperatured air blown through a slit, moving the film by at high speed to produce a strong temperature gradient? I’m still sceptical.

-) All regular mosaics I remember! were made out of micro-stripes of a dyed substrate built up in a regular way as those chequered verniers (placing layers on top of each other, slicing them and rearranging them).

-) Instead of trying to print a regular mosaic, why not reproduce it on colour film whilst scaling it down? Panchro-film and that colour film emulsion side to emulsion side would make a fine additive system. Of course one could consider this use of a colour film weird; but what about this modern computer driven printer to print that mosaic?

htmlguru4242 · Jun 8, 2007

I was JUST going to suggest that color film idea! That would be the perfect use for a color slide film. (Obviously the mask on C-41 would interfere significantly). Regular E-6 slide film, in any sheet format would work wonderfully for this, as it has sufficient color saturation and resolution, in addition to excellent transparency. Some type of optical printer, or a camera with a high-resolution lens, would be required to do it. This would have the distinct advantage that new masks could be reproduced by simply contact-printing a master onto slide dupe film.

Reading back on what I'd said, I wasn't entirely clear as far as the meltable resins, etc. go. I was talking about putting those under the emulsion, in the manner done with Autochrome. However, due to the difficulties with that method that we have discussed endessly in the past, I think the idea of the removeable screen is far superior.

Roger Hicks · Jun 8, 2007

htmlguru4242 said:
Roger, that's good to know!

Any idea if he'd mind being contacted about it?

Dunno. If he does, blame me. Martin is incredibly enthusiastic and knowledgeable, but also, as the fons et origo of Silverprint, incredibly busy. If he has the time, I'm sure he'll respond in detail. If he's too busy, he'll probably say so unless he's WAY too busy. Either way, he's an incredibly nice, straightforward, honest guy.

Martin: I present my apologies if you're reading this and ARE too busy.

Cheers,

R.

Kino · Jun 8, 2007

I have a rather extensive article on Dufay Color and its manufacture from a book, "Colour Cinematography" by Maj. Adrian Klein 1936. In fact, I have been in contact with fellows from the British War Museum and a British Lady who now resides in Kentucky who were attempting to locate the final resting place of the original manufacturing equipment for Dufay Color film. Seems the equipment was in Great Britain and, when Dufay ceased operations, a French company purchased it and it disappeared into the Continent and has yet to be located.

In any event, you want to read about a hellishly complex and beautiful color process? Try Gaspar color! I held what was perhaps the last pristine release print of Oskar Fishinger's "Colour Flight" in me hot little hands at the vaults of the Library of Congress Lab in Dayton Ohio.

It is unlike any filmstock I have ever seen before or since and it puts Technicolor and Kodachrome to shame for saturation and purity.

htmlguru4242 · Jun 8, 2007

Hey, thanks Kino. I'd heard about Gaspar, but I was never able to find any info on it!

Has anyone ever seen (or does anyone have) any examples of Dufay Color?

AgX · Jun 8, 2007

Gasparcolor is nothing special, I'm sorry.

It's a silver-dye bleach process as we still use one in the materials from Ilford Imaging.

In the Mees&James you can even find some chemical background on it.

greybeard · Jun 8, 2007

htmlguru4242:

If you are actively pursuing the re-creation of Autochrome using starch granules or the like, you might want to consider electrostatic charging as a mechanism for adhering a layer only one particle thick. Because of the field cancellation by opposite polarity charges, the tendency is for the first particle in a given spot to adhere, and subsequent ones arriving in the vicinity to be repelled....if you get conditions juuuuuuuuuuuust right! It isn't obvious that potato starch would have the right conductivity, etc., but there is probably a (dyeable) polymer that would have suitable properties.

htmlguru4242 · Jun 8, 2007

Interesting idea ...

I'm sure that electrostatic stuff like that could be done. I'm wondering how i would go about making the particles stick permanently afterwards, but that shouldn't be too hard.

htmlguru4242 · Jun 8, 2007

Speaking of Dufaycolor ...

http://cgi.ebay.com/Vintage-Boxed-I...oryZ4701QQrdZ1QQssPageNameZWD2VQQcmdZViewItem

holy crap, I've never seen that stuff on EBay before. I would bid on it if I thought I would win it ...

Based on the speed figures that I've seen, I'd not be too surprised if it's still good.

AgX · Jun 8, 2007

greybeard,

I'm not sure about this. Doesn't become the first particle layer electrically part of that base? At contact the polarity is cancelled between the first layer and the base; the base however is connected to a charging circuit, thus has reserves to build up polarity again towards the next upcoming particle front.
If there was a field-cancellation, then by means of electrostatic coating no thick powder layers could be gained.

Color images from the 20's and 30's

Membership Council

Privacy & Transparency

Privacy & Transparency