In the sixties, videotape used to film television programs was expensive. When a program had been shown as many times as the contract required the tape was wiped and reused, unless someone requested it be saved for some reason. At least, this was the BBC’s doctrine. Many episodes of the BBC’s programs have gone missing due to this reuse of the videotapes but sometimes the films of these episodes are found in an attic or storage facility. [Cplamb] brings us the story of the salvation of some episodes of British comedians Morecambe and Wise’ first series on the BBC, their first color series.
Do make duplicates, the BBC would film a television playing one of the videotapes. This film duplicate would be sent out to television stations around the world, rather than the tapes. The Morecambe and Wise film was found in the humid basement of a television station in Nigeria. Due to the conditions, the film was “diseased” and was in danger of decomposing into soup.
A series of hacks was used to restore the episodes from the rotting film stock. X-ray microtomography was used to scan a roll of film to see if this could be used. This worked because the film has a layer of silver oxide emulsion(the image) on one side and plastic (the film stock) on the other. A program was written so that the resulting voxels could be remapped into two dimensions in order to see the original frame. However, the volume that the machine could x-ray was small – using it on an item the size of a full roll of film would probably destroy the film, if it could be done. The next hack was to cut the film into small blocks using a laser cutter. This itself seems destructive but if you can either cut it up and scan it or let it turns into soup the choice is easy.
A second part of the story has been published, but the third article in the series hasn’t been yet, so we don’t know how the resulting film looks. But this is a pretty cool story involving scanning, x-rays, programming, and laser cutters — all hallmarks of the great hacks we see on Hackaday. Check out this article on the mechanics of film projection and this one on automatically scanning 8mm film for similar style hacks.
eeeh?
part 3: http://www.bbc.co.uk/rd/blog/2017-12-morecambe-wise-film-recovery-processing-algorithm
The third part is linked at the end of the second. It’s dated the 31st December, maybe this HAD post was in a queue for a few days.
Second paragraph starts with a word / typo: the word To seems intended
The really brilliant part will be when all the images are collected and they use a neural network to correct damage, distortion and degradation. It’s pretty amazing what you can do using neural networks.
Not sure if joking…
If not: why would a neural network be better than the collected knowledge from testing and experiments done by clusters of high-performance biological neural networks++ condensed into mathematical and algorithmic form?
IOW why not use programs designed for the purpose designed by people with extensive knowledge of types of degradation and the physics of film stock?
probably in reference to this (https://motherboard.vice.com/en_us/article/gydydm/gal-gadot-fake-ai-porn) Just set it up to map an approximation of the original onto the damaged scenes and let it run, Probably not going to be as good as what you are saying but will probably be quicker and cheaper.
Martha Stewart will never be the same
Where do I send the fees for transfer documents, enabling shipment?
This.
The first part will be deleted when the last part is posted to make room on the Internets.
Don’t worry, I’m sure we’ll find it on a rusty flash drive in someone’s basement in Nigeria in 50 years.
Hello – I’m one of the researchers who did the scanning.
The results published so far are good, but you ‘aint see nothing yet.
Now find some of the missing Dr. Who episodes!
B^)
Nothing would make me happier.
I’m hoping an archivist will turn one up and deliver it to the lab for scanning one day.
Some of the film found in Nigeria was Dr Who episodes. The Dr Who archive has been working on recovering lost episodes for a long time, and I believe that nearly all the colour ones have been found and most of the B&W. Once they have finished this Morecambe & Wise project there will be a new system for recovering “diseased” film, and stuff which was thought completely lost will be recoverable. This is a truly amazing project and a great hack. Kudos to all concerned.
I myself can not wait. This is largely good news.
https://www.themarysue.com/lost-doctor-who-episodes-found-in-space/
That’s just no way this is true. A parody site perhaps?
Kind of reminds me of the work done to retrieve text off of scrolls, documents, and other surfaces.
Yeah, I was really amazed when I first heard of this technique to scan ancient religious scrolls.
I saw the documentary about the Antikethera Mechanism which used this technology to look inside the corroded lump that was recovered from the sea bed. The results allowed them to make an accurate replica. One of the things that really got me was the engineer who made accurate gears from scratch with ancient tools!
The text from scrolls thing was done by at least one of the same people (Dr. David Mills) – I think the BBC approached him because of that work
Yep – that’s what happened. We didn’t think film would show up, so just put a bit on to scan one weekend to so we could show it wouldn’t work and they would stop asking us. We ended up getting a very nice result from that scan, and then agreed to work on the BBC film.
This video shows the result from one of the test scans we did. It’s just a show reel on a 35mm film can, good quality film.
https://www.youtube.com/watch?v=3AD3_ABDjT0
Truly fascinating. This will open the way to recover a lot of other nearly destroyed films. While Dr. Who seems to be the one show that generates the most interest, I’m sure that there are plenty of other good shows that will be saved with these methods.
“videotape used to film” The only thing used to film stuff is FILM. Tape is used to record stuff.
Is that like the motor vs. engine thing?
Wow, lots of typos… this article was pretty hard to read relative to the average grammar on HAD articles