DougGorius

LucasGodzilla said:

Although I doubt this is a useful bit or anything since probably someone else is covering this base on your team of restorators, I believe that this movie will be easier to fix the mild fading issue than most others for the basic fact that it provides hard captions as well as much of the footage being hard matted. Theoretically, one can make a white balance using the hard matte as an obvious black point and the captions as a white point.

I made a little test real fast to demonstrate my point (although please note, I forgot the title was “Return” not “Revenge” so minor whoops lol): http://screenshotcomparison.com/comparison/132337

I’m willing to bet this scan has a fade-corrected grade on top of it already. Look at how the dirt is blue-green instead of black - that is a side effect of stretching the red channel.

Fading doesn’t affect just the highlights or just the shadows in isolation, or one part of the frame more than another. It is a uniform reduction in contrast of the affected dye layer or layers, edge to edge, throughout the entire, exposed image. In prints the shift is more obvious in the blacks since that’s where the most dye is. Fortunately, the magenta layer - which has the most sharpness - is usually the last one to go, and just from personal experience with this stuff, if you see a little yellow and blue and green left, there is enough dye left to recover a color balance close to the original. It would be really interesting if the scanner supplied snippets of the film without fade correction whatsoever.

You can’t always assume that the frameline or hard matting was originally perfectly, solid black throughout or that the most transparent areas were perfectly clear. The highlights may have a slight tinge to them due to the processing. I’ve noticed this in LPP color from around the same time as G85. Depending on how each scene was timed, the darkest information could swing between slight green or brown or red. This makes it impossible to determine exactly what the original color balance was, so you will always have to settle with an approximation… the Australian Roadshow VHS is a good reference since it’s a transfer of an original theatrical print, but you still need to keep in mind that Roadshow’s telecine may affect the look of the film along with the fact that VHS has extremely limited chroma.

Digitization is always going to be an approximation, so it’s good to keep in mind that the scan’s color is going to be affected by the sensor (i.e. its dynamic range, SNR, color separation) the type of light source the scanner uses (diffuse or collimated) and even the monitor that you’re viewing the scan on. It’s also worth remembering that the film would have had a different color quality from screening to screening depending on the color temperature of the lamp and slight variations in color quality from print to print during the same print run.

It’s starting to dawn on me that almost every official DVD release of an MST3K episode has errors of some kind, like missing audio or video, or serious video glitches that have accumulated in the master tapes. Some episodes have been released twice - from Rhino and Shout! Factory - with each release featuring egregious errors the other doesn’t have and vice versa. And then you have eleven episodes that have flat out never been released on DVD except for their host segments on the most recent volume, and some of those released segments have video issues. And then there are the fan-traded copies, which have their own tradeoffs. It is utterly maddening.

Would there be any interest in a project with the goal of definitively reconstructing the unreleased episodes in their complete, originally broadcasted form using the best quality sources, while also identifying released-but-problematic episodes and reconstructing them with the same amount of care? I have edited video for years and have the obsessive eye for detail needed to reconstruct these episodes as precisely as possible, so I would like to take on the actual reconstruction, but I don’t really have that much good source material for MST3K episodes besides a handful of Rhino and Shout! volumes.

The pamphlet you linked to is for identifying edge codes of Kodak stocks only. Your film is DuPont stock, which didn’t begin production until 1920. This link should be more useful:

https://erikpiil.files.wordpress.com/2014/06/dupontedgecode.pdf

Please read the link ScruffyNerfHerder posted and be careful with this stuff. I urge you to study this data sheet closely, too:

https://amianet.org/wp-content/uploads/Resource-Nitrate-Safety-Data-Sheet.pdf

Because your film is nitrate, there is and will always be a risk of combustion. Cellulose nitrate film is classified as dangerous goods by the UN and cannot be shipped through the post or carried on public transportation. Handled and stored in ideal conditions, nitrate can last over a hundred years, but when it isn’t, the consequences can be tragic, to say the least. The film should be kept far away from sources of heat or ignition, no matter its apparent condition. It needs to be kept cool and dry. Nitrate film fires cannot be put out because the film itself contains an oxygen supply. The smoke is extremely toxic and potentially fatal if inhaled. Nitrate doesn’t combust randomly, it combusts due to ignorance and human error. This is and has always been the case. Please heed my warnings and treat this matter as seriously as possible.

And it wasn’t merely one studio that lost many silent films due to nitrate fires. Many studios and archives worldwide suffered from major nitrate fires through the 1980s, when it began to be understood how to safely store and handle it.

Side note - DuPont film was a bit lower grade than Kodak film and is known to decompose quite badly, so it’s cool to see such an early specimen of theirs in apparently great shape, aside from the mechanical damage.