Hi!

Thank you for your reply. I fear you are correct that the front end
is most important - the mike and preamp.

It's going to be a good starting point. But almost any portable cassette
recorder I've used did a good job of picking up sound from all the room.
It's a compromise of course...the more room noise you pick up, the less
chance you have of being able to focus on a specific source.

In defense of digital vs mechanical, the Sony has 1 1/2 hr of
'perfect' MP3 and more than 340 hours of plain recording

A tape won't match that, but my thought is to wait until the noise occurs
and start recording immediately. Announce the time of day and then just let
it roll. Or get a tape recorder with voice operated recording. I have a
Realistic Minisette 20 here right now that does this with two sensitivity
levels. Its built in speaker is nothing to write home about, yet the
recordings are pretty good. I'd guess it was made before the heyday of
microcassettes (and handily beats those in terms of audio quality).

[even speech recognition software]

Which would probably work from a tape recording as well, although you're
likely to be able to do a much better job by transcribing it yourself if the
need arises. Speech recognition is still an imprecise concept.

I just realized I was playing the sound back using the built in 1 inch
diameter speaker. Duh! Will try using earphones with better fidelity
to see if the lower spectrum exists.

Consider going further than that -- play the unit into your stereo receiver.

so I can do some very esoteric manipulations using a Matlab clone, octave.

While that sounds very interesting, why do you want to do that? It sounds
like a lot of work to me, and I don't understand the value of the outcome.

William