“Talking rubber”, a forgotten audio format | Hacker News Reader

“Talking rubber”, a forgotten audio format

arstechnica.com

63 pointsshawndumas12 comments

"in 1954 the only way to record sound was to press a phonograph or record sound on film for a movie."

There was wire recording:

https://en.wikipedia.org/wiki/Wire_recording

AMPEX demonstrated magnetic tape recording in 1947, based on the German magnetophone technology used during WWII. Shipments to radio stations began in 1948.

http://www.aes.org/aeshc/docs/company.histories/ampex/leslie...

The Dictabelt also became available in 1947, according to Wikipedia: https://en.wikipedia.org/wiki/Dictabelt

An interesting article about archiving of Dictabelt and other audio format recordings: http://thequietus.com/articles/18004-sound-library-british-l...

Also the Blattnerphone (steel tape):

https://en.wikipedia.org/wiki/Ludwig_Blattner

Speaking of forgotten audio formats: Tefifon, old german analog loop of elastic, non-magnetic, tape.

More or less ``8-track meet vinyl''.

https://www.youtube.com/watch?v=nBNTAmLRmUg

https://www.youtube.com/watch?v=9tTURrAWVYE

I find it odd that the corporate research lab is popularly acknowledged to be dead when so many large companies maintain corporate research labs:

http://www.geglobalresearch.com/

http://research.ibm.com/

https://research.google.com/

https://www.microsoft.com/en-us/research/

https://www.cbinsights.com/blog/corporate-innovation-labs/ <- a giant list of more labs

...

Now, I will admit that corporate research has tapered off significantly since then, but 'tapered off' and 'dead' are two entirely different things.

You might think 'But wait! the kind of basic physics research that used to be done by Bell is dead'. I counter with two arguments:

- Basic computer science research is an equally valid, equally fundamental thing, and many of the labs listed above conduct it as their primary business

- Basic physics research does still happen. I included GE Global Research at the top of my list because they in particular do a lot of research into materials, chemistry, etc (which will eventually be applied to things like turbine blades, batteries, LEDs, etc.).

CS research, for the most part, is a fractional cost compared to basic physics research, materials research, chemistry, etc... mostly due to the capital costs associated with setting up proper labs.

The majority of science research, especially the risky exploratory kind vs. the incremental efficiency/process improvement kind, is mostly in universities and to a lesser extent, national labs.

IBM's research projects ( http://research.ibm.com/#featured-projects ) seem to include areas such as polymer research ( http://www.research.ibm.com/articles/industrialpolymers.shtm... ), semiconductors ( http://www.research.ibm.com/articles/chipchallenge.shtml#fbi... ) amongst other domains.

While there is certainly CSesque research in the list of areas, it is not the only area.

I think I once actually saw this. When I first started at the local telephone company back in the day I was shown the corporate PBX system which was interesting because it was a small step by step system in an era where such things were considered obsolete. There was a rubber belt running on pulleys in a transparent box. There were several magnetic recording/playback heads in contact with the belt. The bottom of the belt passed through a pool of lubricant. It picked up so much that the heads left a wake and had a kind of a bow wave. lubrication was apparently very important.

Likely that the heads rode on a film of oil. Thus was no actual contact between the rubber belt and the heads and so no wear. And also probably the oil film meant that belt moved smoothly under the heads. One could imagine if there was any friction the belt would tend to move unevenly (skips and jerks) leading to crummy voice quality.

thetechmoan on youtube often brings old audio formats into light:

https://www.youtube.com/user/Techmoan

not as old as talking rubber usually though

https://en.m.wikipedia.org/wiki/Speaking_clock

I believe some of these used the same tech

Reply on news.ycombinator.com