[rescue] Current collections...

[Joshua Boyd]

On Tue, Apr 06, 2004 at 08:04:21PM +0000, Lionel Peterson wrote:

> Well, let's do the math - TRS-80s saved programs and data at 1,500
> bits per second (bps), now, if I assume 10 bits per byte of data
> (allowing for handshaking/etc.), that comes to 150 bytes per second -
> right? 
> 
> I record MP3s at 160 Kb/s - so if I didn't change anything else, I'd
> be using 20,000 bytes of MP3 data to store 150 bytes of program data -
> each state change on the cassette would there have aprox. 20,000/150
> samples = 133 samples per bit... I think that would be fine. If you
> step down the sample rate on the MPs to, say 48 Kb/s, then you are
> looking at 4,800 samples per 150 bytes of data, or 32 samples per
> byte. 
> 
> The tolerances on the old cassette interfaces was far from precise,
> IIRC - so it should work, but oh the humanity! It would stillt ake 5
> minutes to store 5 minutes worth of data... ;^) 

I wonder if anyone has written software to convert an audio stream into
the data it carries, then later take that data and generate audio from
it.  That would be near the ultimate compression for this purpose.