[geoff]

On 22/02/2020 6:30 am, Scott Dorsey wrote:
If I were you I'd try to do some processing to improve intelligibility rather
than worrying about fidelity. High pass everything below 200 Hz or so,
then low-pass everything above maybe 6KHz, and consider sticking a presence
boost in there. You might then consider an expander and fiddling with the
threshold on the expander to try and boost the voice out of the noise.

I think you are worried about entirely the wrong thing here.
--scott


Yeah, the source media content is the only thing getting in the way of
intelligibility. Unfortunately most of the extraneous clutter is not too
far removed from the vocal frequencies.

Looks like more of a job for Spectral Layers, or a Mac or Linux
equivalent, to edit out the unwanted noises. Arduous fiddly work, but if
the content is really that important ....

geoff