Missing Formants

[Haible Juergen]

	>>... and Digital Reverb
	>
	>But this is more interesting. One bit sigma delta ADC/DAC by the
looks
	>of it driving a one bit FIFO. Could be done using discrete stuff,
by
	>discrete I mean MSI.

Deltalab showed the way. I built something like that some 10 yeras ago as
well,
based on an article published in Elrad called "Delta Delay". I don't have
the
box anymore (sold it for 50 Dollars to a guy who brought it back a week
later
and said it was broken. I'm quite sure it wasn't when I sold it, but what
the heck.
Payed back the money, threw the box in the "removal of bulky refuse" (is
this
actually the English word for Sperrmüll ??).
Ok, it was a mistake. But those were the times of the DX7 and D50, and
Midiverbs.

Btw, it was *not* a sigma delta converter (Neither was the elektor circuit
if memory
serves), but it's predecessor, a "delta modulator". No sigma - no coding of
DC.
Worked ok on audio signals, though. It was adaptive, i.e. the step size was
varied according to the signal, forming a sort of compander system.
Reportedly
better than the elektor single-chip solution, as these chips were optimized
for
telephone speech rather than audio.
I must have the article somewhere, but not at hand. Better go for the
original Deltalab
schemos anyway. Make a patent search for "Deltalab" and find everything you
need,
including detailed schemos.
Or even better, buy a used Effectron.

JH.