Encoding Data for Audio Tape?
Kris_J asks: "I've recently purchased and installed an audio cassette deck for my PC. It makes recording to and from tape particularly painless, but I'm looking for some funky other stuff to try. Along the lines of my new old SuperDisk FD32MB that can store 32MB on a normal 1.44MB floppy, and my Cuttle Cart that can load Atari 2600 games from encoded audio I'm wondering if there's any program that can encode a file as audio that can survive being recorded to audio tape or compressed as an MP3. I'll worry about 'why' later."
The device itself is quite simple. It's a 5.25" device with a standard power connection. It also has a ribbon cable that runs to a simple adapter that fits in a slot in the case (thought it doesn't plug in like a card) -- from there you connect it to audio in/out and a serial port. Software is seriously basic and I think Windows only, but you can just set a few settings and press a button to have it convert an entire tape, both sides, to an MP3 file. I then cut it up using MP3DirectCut. Just converted an old Vangelis tape yesterday.
So, yeah, it's basically a tape drive with Line In / Line Out and a serial port. However, I'm barely a programmer, just a bit of PHP and SQL, no C, no drivers, no comms, so I'm not up to the task of writing anything myself.
- for both the tape and the recording heads:
- Baseline noise level
- Effective response time
In other words, the quietest reliable volume and the shortest wavelength possible. You'll likely want to find or write an app that encodes ASCII data as 16-bit audio, choosing a high- and low-volume to record binary onto an analog medium. To counter your effective response, simply choose a good per-bit duration -- like cheap error correction. There are many, many ways to get this encoded, but your problem will be the low playback quality of the tapes. Another, more involved, solution would be to assign a specific frequency to each binary place-position, so you can assemble the completed byte-sound, record to tape, and disassemble into the component frequencies. This method will yield a higher density, but I'm afraid you'll probably encounter those quality issues I mentioned before... I'd say it's time to hit the google! Good Luck!