Re: [ARSCLIST] Memorex CDs

[Don Cox <doncox@xxxxxxxxxxxxxx>]

Hello Mike

On 16/01/06, Mike Richter wrote:
> steven c wrote:
> 
>> We can assume that "digital" will clue the party that the pits and
>> absences represent ones and zeros. The next problem (see above) is
>> trying to figure out how the sequence (which wouldn't produce
>> anything like a sound signal if read directly) relates to the
>> sequence of digital values that represent a stepwise representation
>> of a sound waveform.
> 
> The problem appears to me to reduce to solvable and indeterminate
> parts - as follows.
> 
> 1. Trace the spiral of the CD to get the sequence of recorded bits -
> not at all difficult given reasonable tools.

Note that the pits do not directly represent 0s or 1s. The signal is
modulated using EFM modulation. 

http://www.ee.washington.edu/conselec/CE/kuhn/cdaudio2/95x7.htm


> 2. Guess that the disc represents audio and use persistence of tone to
> decrypt the encoding algorithm for the byte information. Not trivial,
> but not especially hard.

Very hard from pure guesswork. I'm not sure what you mean by
"persistence of tone". Audio CDs use an interleaved Reed-Solomon error
correction encoding so the data for any sample will be spread along the
length of the "groove".

http://web.usna.navy.mil/~wdj/reed-sol.htm

> 
> 3. The process of 2. will lead to the breakdown of samples to bytes. 
> That one is trivial.
> 
> 4. Determine sample rate - a classic impossibility from the audio disc
> alone.

There would have to be a major breakdown of civilisation for everyone to
forget 16/44.1.


Regards
-- 
Don Cox
doncox@xxxxxxxxxxxxxx