Cellulose nitrate, the original material used for motion picture films, had a couple problems: it would self-decompose chemically, often unpredictably, and it was exceptionally flammable being chemically very close to modern gunpowder.

Cellulose acetate, a replacement first used around 1910 and improved somewhat in succeeding decades, is comparatively non-flammable. Unfortunately, like cellulose nitrate it still has problems with chemical breakdown.

Polyester filmstocks are probably pretty good, but they've not been around long enough to establish their archival properties.

For truely long-lived film archiving of modern films, therefore, I would suggest applying the concept of color separations to a very old technology: printing motion picture films onto (non-wood-based) photo paper.

Although paper-roll films cannot be projected conveniently as can transparent films, archived movies generally shouldn't be projected anyway--their use should be to generate viewing copies of movies (whether in film or digital format) which can then be viewed without risk of damage. And scanning paper roll films should work just fine for that.

To avoid issues of color dye stability, I would suggest printing films onto paper rolls as color separations--one frame each red, green, blue, etc. throughout the film. For recording multi-channel sound, a roll of paper wider than the 35mm film could be used; since the linear speed would be three times that of standard 35mm projection (for which optical sound has historically been deemed adequate) audio fidelity for even analog soundtracks should be adequate.

Although storing films in this manner would require somewhat more space than storing them in a composite-color format, this approach should have a few advantages over current archival methods:

-1- Because the films would be printed using silver-halide chemistry without the use of dyes, the emulsion should be exceptionally stable. Hundred-year old films printed this way survive today with minimal degredation, so there's no reason not to expect another hundred.

-2- Use of color separations would ensure that if any part of the film does fade for some reason, the different color components could be extracted easily and accurately.

-3- The format would be reasonably self-apparent with the media. One of the major problems with many types of modern storage--especially with video--is that as formats become obsolete there's no good way to get at the old material. One can't very well look at a videotape and see what needs to be done to read it. By contrast, someone examining a motion picture stored as described could easily determine what needed to be done. The recurring patern every three frames would clearly suggest a color separation, and the wavy lines would be clearly recognizable as soundtracks. Although some experimentation might be required to determine what order the color components belong in, and which soundtrack corresponds to which channel, it should not be hard to figure such things out.
-- supercat, Apr 18 2005

Silver halide? So, these separations would just be monochromatic representations of the opacity of each color channel?

My approach would be to scan the film at grain resolution (say 4 or 8K) and store digitally but to each her, or his, own.
-- bristolz, Apr 18 2005


The paper grain seems a problem compared to a transparency.

This seems really complicated for the problem at hand. I liken it to digging a hole through a concrete floor to get to the other side of a door when you could have just opened the door.
-- bristolz, Apr 18 2005


Sounds like it would be possible to extract the information easily enough, but doesn't really sound like film any more. Leaving aside concrete, seems more like keeping printed sheet music instead of recordings. Also, wouldn't it be awfully bulky?

A magnificently hare-brained scheme. [+] in an intellectual kind of way.
-- moomintroll, Apr 18 2005


bris, I thought we have all coverd this ground before. If there is nothing to be fixed, or improved upon, but someone does, and it costs more, looks differently, acts differently, and fixes no apparent defect in anything, that is what makes it a halfbakery ta-da.
-- blissmiss, Apr 18 2005


mmm. Forgive me if I'm wrong. But what's wrong with the current idea of storing on Laser Disk and duplicating every 5 years.
-- DRudge, Apr 18 2005


You may have covered this somewhere, bliss, but it's a ta-da for me if it makes sense (plausible but inobvious) or is ridiculously over-the-top and this isn't either.
-- bristolz, Apr 18 2005


The difficulty with storing to digital media and copying periodically is that anything that for some reason doesn't get copied at one pass may end up being effectively lost, even if the media itself hold up just fine. There are videotapes from the 1950's which may contain the only existing copies of television shows, but nobody can really tell because nobody has equipment to play them.

As I see it, all that would be required to convert a paper roll back into a movie would be a 1200dpi or 2400dpi scan of its length. Even if one knew nothing of the details of how the film was stored, it shouldn't be too hard to figure out.

The archival stability of photo paper should be adequate for 100+ years, since existing paper prints have already lasted that long.
-- supercat, Apr 18 2005


Have they, though? I thought that traditional photo prints suffer some fairly severe degradation after only 25-30 years but maybe that's all dye problems.
-- bristolz, Apr 18 2005


//I thought that traditional photo prints suffer some fairly severe degradation after only 25-30 years but maybe that's all dye problems.//

Many color dyes have problems over a few decades; some have problems even faster than that. If a film is properly developed and fixed or bleached (depending upon whether a neagive or positive is desired) there shouldn't be a whole lot to go wrong. The dark areas of the paper are silver; the light areas are nothing.
-- supercat, Apr 18 2005


I know Cibachrome is very colorfast.
-- bristolz, Apr 18 2005



random, halfbakery