+ - X / and ~

[Tim Walters]

At 6:38 PM 12/12/96, Kevin Lightner wrote:
>Isn't the formula that for every 3db, the "audible" perception of loudness
>is doubled also?

Not quite.  While mixing two uncorrelated signals of equal strength tends
to yield a 3dB increase, we don't really hear it as "twice as loud".  That
point is generally considered to be around 10dB.  3dB is more like a
"noticeable increase", with 1dB being roughly the "minimum noticeable
increase".  Of course, this is all quite subjective, but backed up by
psychoacoustic studies.

Imagine the following:  a violinist is playing an Irish reel.  For the B
part, she is joined by another violinist in unison.  Louder?  Yes.  "Twice
as loud"?  No.

And of course, to get the same relative increase again, you'd have to add
two violinists, then four, etc.  Which is why symphonies have a zillion
fiddlers, but the soloist can still be heard.


-----------------------------
"This is digitally conceived New-Pagan vomit.  And I've never tasted vomit
this good." --yet another review of my CD, this time in _Sonar Map_

Tim Walters  --  walters at digidesign.com  --  http://www.slip.net/~coredump