Alberto:

I recorded some 578 seconds of audio to hard disk at 7200 samples per

sec.

[snip]

Seems really a good idea, which could attract also the
die-hard CW-must-be-copied-by-ears crop around.

Just one simple question : why 7200 samples/sec ? It is not a standard
sound card sampling rate. Is there a specific reason for this sampling

speed ?

Perhaps Bill was using an external A/D converter. For the technique to
become widespread and accepted by many (all), it shouldn't rely on an
ad-hoc hardware. Modern sound cards have superb A/D converters,
so why not use them ?

Yes, your right sound cards today are at least good.  However Bill started
his work way back and he uses a small card that implements in discrete logic
a Sigma Delta A/D, published in QST in 1992.  The card runs the serial port
at 115200 baud with 10 of the 16 bits being a full async character with
start bit, 8 data bits, and a stop bit.  the blank 6 bits let the processor
get some work done between data characters.

http://www.ietc.ca/home/bill/bbs.htm

if you go and look and download some of the material at Bills web site you
can see how he does things.

For the technique to
become widespread and accepted by many (all), it shouldn't rely on an
ad-hoc hardware.

Fair opinion for you Alberto, possibly relevant in your terms.

Modern sound cards have superb A/D converters,
so why not use them ?

Computer sound cards did not exist when Bill started his work, since they
were not available they were not used.  Having been around A/D converters
for many years, the superlative of "superb A/D converters" seems to have a
degree of hyperbole from my perspective.

Bill will probably speak to the sound card issue at some point in the
future, that is his choice of course.

In the interim he has done DSP work that is superb for weak signal work and
has been demonstrated to perform excellently.

Larry
VA3LK