Robert Orban wrote:
In article ,
says...
we listen to each others recordings at monthly meets. My
contribution that month was a sample of intentionally clipped audio
with differing amount of clipping. The material was classical choir.
The first 2 dB of clipping was inaudible to all, I just made the
demonstration to get people to understand that it is possible to
worry too much about technical perfection, so I didn't make any
notes, but I think it was around some 4 dB of clipping that it was
perceived as audibly deteriorated. What matters is not the amount of
clipping, but rather the duration. A couple of milliseconds tends to
be inaudible, this based on information provided at some AES event
here in Copenhagen.
It's *very* program-dependent.
Agreed. Choir in a reverberant room is not what I would consider "forgiving"
in this regards.
You can't hard-clip classical piano at
all without hearing clicks. Indeed, the original BBC delay line
limiter was inspired by the observation that even a fast attack time
limiter would cause audible clicks on piano.
How many milliseconds is "at all"?
The delay line limiter
allowed them to decrease the slope of the limiter attack to the point
where the modulation distortion was psychoacoustically masked.
Thank you!
Bob Orban
Kind regards
Peter Larsen