Anders Andersson wrote:
> How can one easily do a 90 degree phase-shift on a signal?
> Please explain theoretical, and not just how to
> implement it in CSound, as I'm probably going to
> use it when programming in other languages too.

Being myself a complete neophite, i think in the analogue world this is
continuously done, e.g. in a loudspeaker. Consider the magnetic flow passing
through a coil pair. The induced current is 90º phase-shifted from the inducing
one, because it's its first derivate.

I am a bit surprised of reading about the importance of the phase in a sound. If
I'm right, the phase in the air soundwave is 90º from the one in the microphone;
this is another 90º from the one in the cables. Only God knows what phasing is
done in all the equipement stuff. Once again the phase is shifted at the
loudspeaker. Happily all the frequency components are to be considered senoids,
whose derivate is a cosenoid, whose derivate is a top-down senoid and so on. All
of this is producing a time delay of at least one or two wavecycles in each
frequency. So, why worry about phase?

I'd like to know if i'm right.

Saludos,
-- 
Francisco Vila. Badajoz (Spain)
E-mail: fravd@ctv.es
http://www.ctv.es/USERS/fravd