I heard a report on NPR, reporting how a new method of playing back old recordings has been developed: you use a computer to scan the old analogue grooves (on a laquer record disc, or old edison wax cylinder). The computer can then translate the shape of the grooves back into audio-a much better method than playing the old discs (which would destroy them). My question: a lot of old recordings sound pretty “tinny”-can you restore the lost high frequencies? Or is this info lost forever?
great development though-I’m sure a LOT of these old recordings were about to crumble into dust! :eek:
I heard that story also, and it seems really cool. However, I would assume that the tinny sound is a result of the equipment used to make the original recordings After the recording process was created it was probably a while before the fidelity got any good.
IOW, the stuff was never on the recording to begin with. I suppose a good sound tech may be able to restore what ‘might’ have been played at the recording session, but I doubt we’ll ever know for sure.
Here is a link to the NPR story, with a set of before-and-after audio files. The system is called IRENE, and, as you note, it works by taking extremely detailed images of the grooved surface of the record or cylinder (or Dictabelt or what-have-you). There have been other optical scanning systems for records (including the ELP laser turntable), but this seems to be a more archivally viable option.
As far as the “tinny” sound of old recordings–this system, in and of itself, could not really address that. Most older recordings are sharply rolled off in both the low- and high-end (generally around 8K, maximum), and IRENE, I would think, is not designed to read frequencies that aren’t there in the first place. The most effective way to address such issues is in the mastering stage, wherein a mastering engineer uses compressers, EQs, and other tools to sculpt the audio file. Even this will not make a cylinder recording sound like a modern, slammed recording–but that is a good thing, as they shouldn’t be made to sound that way, anyway. Besides, I doubt any of these IRENE scans ever see a mastering suite.
There are certainly some issues with the technology that occur to me:
-
How well does this system read translucent recording surfaces? The ELP turntables, for example, fell down when trying to read anything but black wax. After all, there are many vinyl and shellac records that are presed on colored or translucent materials.
-
I’d be curious to know the source of the hiss in the IRENE sound files, which seems to be more prominent than the non-IRENE files. Is there a higher noise floor?
In general, however, this seems like a really interesting way for sound archivists to create surrogate sound files of rare recordings. As long as they don’t get any dumb ideas–like disposing of records once they’ve been scanned. I would love to know more about the costs of the system, too…
Wow, then CSI might not have been totally cuckoo after all. In one ep, they use something similar to scan a piece of pottery to extract voice recordings made while the clay was being shaped on the wheel. OK, it’s just a tv show
Actually, it was cuckoo. Mythbusters tried. Busted. http://kwc.org/mythbusters/2006/10/episode_62_killer_cable_snaps.html