27 May
2008
27 May
'08
8:16 a.m.
On Tue, May 27, 2008 at 12:07:11PM +0200, Atte André Jensen wrote:
Fons Adriaensen wrote:
If the delay roughly corresponds to the path difference of one
ear to both speakers, and you do invert, then this is a primitive
form of crosstalk cancellation.
If you want to try the real thing (which can be quite good), get
jconv and use one of two EYCv2** configurations. More info in
the config files.
Ciao,
--
FA
Laboratorio di Acustica ed Elettroacustica
Parma, Italia
Lascia la spina, cogli la rosa.
Bah, what you get is no stereo image at all, but
some sort
of 'spatial soup'.
Ok, then we agree :-)
However now I reread the post more closely, and it seems his intentions
were a bit different. Here's what he wrote:
"feed the left channel back to the right with a short delay, inverted;
feed the right channel back to the left with a short delay, inverted;"
Actually it sounds to me like the inverting part becomes irrellevant. When
the delayed + inverted sound is output, the original audio has moved on and
the delayed audio is simply "something new" that's mixed in, inverted or
not.
Is this some kind of well known trick or am I right that inverting the
delayed signal doesn't do anything (good) for the result?