Subject: Re: [linux-audio-dev] Safe-Ears/software volume
From: Tom Pincince (stillone_AT_snowcrest.net)
Date: Wed May 24 2000 - 06:21:57 EEST
I think that all you really want is a stereo limiter plugin inserted
across the monitor output.
I usually crank up the volume to hear fades and I often forget to turn
it back down. Fortunately I rarely ever monitor with headphones so my
ears are still ok.
An automatic software volume control would cause problems, particularly
when listening to very dynamic recordings. It would essentially act as
a dynamic volume maximizer followed by a fixed attenuator, causing loud
passages to become softer and soft passages to become louder. All of
the music would come out at the same volume and all dynamic information
from the recording would be lost. Audible pumping would also be
introduced and, while there would be no danger of loud signals, the
listener would not know what he is really listening to.
A zoomed in waveform graph often displays regions of 1 second or less.
Having the zoom level dynamically adjust itself durring playback would
be confusing. It seems to me that you are almost thinking in floating
point mentality and letting the graph display only the mantissa.
I do support your concept of protecting hearing. A simple aproach would
be to play a soundfile that has many peaks at or near 0dbfs, crank up
the volume to maximum, insert a stereo limiter to the monitor output,
and adjust the limiter so that even the loudest possible signal is still
safe. This configuration would not alter the sound of fades, which is
the only thing that I listen to at full volume. The limiter should have
a bypass function for those times when the listener wants to hear the
entire mix, as is, with peaks unchanged. This will require the user to
remember to turn it back on but I don't see how a fully automatic system
can be musically useful.
Tom
This archive was generated by hypermail 2b28 : Wed May 24 2000 - 06:42:47 EEST