I was wondering: On the past we used to store software on tape. Some cassettes even came with music AND software. What if one wants to become a really alternative hipster indie software developer, saving his program on a Vinil disk? Would it be possible to store information in the form of sound (or… 0s and 1s, that coud be translated as grooves and bumps on the record)? If this developer recorded the sound of a tape on a vinil (creating the press, etc etc etc) could he run the software by playing the record?
You mean Vinyl?
Sure, you could do that. An LP would store a little less than half as much data as a CD - 300 Mbytes or so.
Not only has this been done before, we’ve discussed it before.
Anyway, back when Microsoft was spelled Micro-Soft and they were best known for their BASIC interpreter for the Altair 8800, microcomputer (personal computer) software was most commonly distributed on audio Compact Cassette tape. It’s the same idea: Use audio-frequency sound to encode data, often by assigning one tone to represent 1 and another, very different tone to represent 0. (More complex encodings are possible, but if you’re distributing on consumer-grade media and decoding it on an Intel 8080 with a laughable amount of RAM, simplicity is a virtue.)
That’s the start of the OP …
Yes, the vinyl is even better than audio cassette …
The data on audio cassette is stored as sound… so the vinyl record company could just take the tape and transfer the sounds from it to vinyl… the record player would easily play the same sounds accurately. The interface on the computer demodulates the sounds to transfer them back to 1’s and 0’s.
It is interesting to consider the potential capacity of vinyl and tape. The hobbyist cassette standard was dreadful in both its simplicity and design flaws. Using a pair of harmonically related frequencies instantly meant that the effective signal to noise ratio was degraded.
Shannon tells us that the ultimate data rate is bounded by the product of bandwidth and signal to noise. Complicating matters, the signal to noise is not constant across the band available, but we simply integrate over the band to get the result. Actually making something that could make full use of the band wouldn’t be easy - but isn’t too far removed from what DSL does. Indeed it might even be possible to re-purpose a DSL modem for the job. Done carefully I could imagine getting many tens to maybe a hundred odd megabyes to of data onto an LP, maybe more with well crafted error correction and detection. You would need a pretty high quality turntable setup, but it is likely viable.
I could imagine setting this as a final year engineering project.
I think sound was the basis of short term memory too, in some devices. I think they made a sort of echo chamber, like a pipe that a speaker would squeak or squawk into and from the other end of which a microphone would listen. This would store a set of really short squeaks and squawks that are going around in a short loop. Their quality would not deteriorate in the sense that they would be freshly reformed each time around.
You are thinking of delay line memory. This comes from a much earlier era in computing. The first were mercury filled lines, and later there were wire based ones that transmitted a torsional wave along a spiral of wire. The data wasn’t encoded onto a carrier, but the raw bits as pulses ran down the lines. It was for a while an economic form of storage relative for working data, but slow. Typically held a few kilobits of data.
I used the best of the best, was carefull to get best audio quality on my chrome cassettes in our recording studio in the early 80s, years later, they have no sound quality at all even being kept well in a room temp safe,
Punch Cards, metal ones.
In the future, our robot overlords will groove to binary music on records.
Write it out on paper, in binary, by hand.
I can see a couple technical problems with saving data on vinyl. Sure, there’s no issue on saving 1s and 0s on vinyl. First of all, you’d need a cutting lathe to record the data. And they didn’t cut the master on vinyl, they used something else, maybe a hard plastic. They’d use the master to make molds, and press vinyl copies with them.
You could probably save data at a faster speed than 33 1/3 rpm. During the cutting process, you’d have to be there with a brush to sweep away the slivers of plastic the cutting “needle” carved out.
I think it would be a lot of effort to no real advantage.
Naw. Use parchment or vellum and a quill. Or a brush. Store them rolled as scrolls. Maybe apply some artwork to the first bit of each section, so you’d have illuminated data scrolls.
Impractical and silly, but cool: The latest rage is storing data in DNA. Slow and expensive to read, ridiculously slow and expensive to write, but the density is higher than any other human technology, making copies is easy and read/write technologies are improving daily.
You mean like this?
I think that storing data in henges should be explored.
Not that much storage available I would guess, the write speed would be pretty low, but the long term storage should be excellent
Record it at 16 2/3 rpm and double the content! What a breakthrough!
That, and write an encoder and then translate the data into encrypted format, by hand.
You can speed up the process considerably by using a CARDIAC computer.