Tape Disintegration Threatens Historical Records, But Chemistry Can Help (nautil.us)

An anonymous reader writes: Modern storage methods are designed with longevity in mind. But we haven't always had the scientific knowledge or the foresight to do so. From the late 60s to the late 80s, much of the world's cultural history was recorded on magnetic tapes. Several decades on, those tapes are disintegrating, and we're faced with the permanent loss of that data. "The Cultural Heritage Index estimates that there are 46 million magnetic tapes in museums and archives in the U.S. alone—and about 40 percent of them are of unknown quality. (The remaining 60 percent are known to be either already disintegrated or in good enough condition to be played.)" Fortunately, researchers have worked out a method to determine which copies are recoverable. They "combined a laptop-sized infrared spectrometer with an algorithm that uses multivariate statistics to pick up patterns of all the absorption peaks." Here's the abstract from their research paper. "As the tapes go through the breakdown reaction, the chemical changes give off tiny signals in the form of compounds, which can be seen with infrared light—and when the patterns of reactions are analyzed with the model, it can predict which tapes are playable."

Comments

Cryogenic storage

[frnic]

Freeze them all and wait until a 3d Printer can scan and reconstruct them at the atomic level...

Not a new problem, of course

[cirby]

I ran into a related issue about 25 years ago.

I was working in a college media library, and there were several stacks (over 70 tapes in total) of 2" reel-to-reel video tape from the 1960s and 1970s - recordings off air from Public Television, mostly. Some of them were of local shows nobody even seemed to remember, and others were from live performances at the Dallas station or of live feeds from PBS. There was a live Alvin Ailey dance troupe local show from the late 1960s, if I recall correctly.

The problem was that they were recorded in a rare two-inch format - and only four machines that used it were ever even built (no, it wasn't 2" quadruplex, there were still lots of those at the time). I couldn't find a working machine, and the only one I could dig up was missing major parts (like the heads). So unless someone builds a new one from scratch just to read those tapes, all of that is going to disappear - if it hasn't already.

Re:Not a new problem, of course

[Tapewolf]

couldnt the tape be still framed one at a time in a modern scanning format to bring it back? (the video portion at least) im not sure how to pull the audio but being analog wouldnt there be a way to pull that as well?

The audio would be pretty easy to pull off - it's going to be a straight linear audio track so you could probably just stick it in a regular 24-track studio recorder. Pulling the video is the hard part because practically all 2" video machines use a segmented scanning technique with the head-wheel angled at 90 degrees to the tape. If these are helical scan, the tracks are going to be laid down at 15 degrees or something weird like that, and you'd need to build a custom video head for it. Maybe it's possible to take a C-format head and machine a suitable drum for it, I don't know.

Earlier I asked if it was an IVC recorder - however, reading it again he said that only 4 existed so I'm pretty sure they were recorded on an Ampex 8000, a 1961 helical scan machine that Ampex made prototypes of but never went into full production with or something. So yes, that's going to be a rare bird indeed.

It's time for a global public digital archive

[presidenteloco]

A technical and logistical and financial project whose primary goal is longevity (in the multi-hundred-year sense) of that which it stores.
It should not be accomplished by individual media that are designed to last.
Rather it should use network redundancy cleverly and have protocols designed to ensure enough geographically distributed copies always exist.
It would have to carefully consider "readability, interpretability" assurances, such as very standard simple formats and protocols, and the methodology of storing the displaying / interpreting environment and code as well as the data. Emulated 1980s arcade games, now available and playable online, are good examples of this.
Sort of an Internet Archive on steroids. Crowdfunded?