Re: Newbie questions
Re: Newbie questions
- Subject: Re: Newbie questions
- From: Brian Willoughby <email@hidden>
- Date: Sun, 5 Apr 2009 23:40:16 -0700
Exactly what I am saying. A phase shift is a delay, albeit a
frequency-dependent delay. Polarity inversion involves no delay at
all. This whole discussion started when I pointed out that "180-
degree phase shift is misleading" because it implies a delay.
"Polarity inversion" is 100% clear and accurate, thus not misleading.
P.S. To answer James' question, yes, the math "OUTPUT=LEFT-RIGHT"
will work. Just Do It. We're only discussing the best terminology
to refer to this. The math is the same no matter what words you use
to refer to it.
BTW. Tahome's suggestion for a pseudo-stereo "edge" mix is not
compatible with industry standards for stereo. It won't survive a
mono mixdown because that would result in silence. This "edges of
the stereo field" is actually just inverting the one resulting
channel and putting it in the other. Many really cheezy guitar
pedals do this, and it's fine for cheap stereo because it will fool
the human brain, but it's not really recommended to use such a signal
because combining channels destroys the audio (not just the vocal,
but all of the audio). Might be fine for iPhone, though, but if your
users attach any monophonic output device, such as particular
BlueTooth speakers, earpieces, or headphones, then it won't work.
Brian Willoughby
Sound Consulting
On Apr 5, 2009, at 23:14, tahome izwah wrote:
You wouldn't even have to use a Fourier transform, you could simply
apply two Hilbert filters in series. Yes, that would result in a
polarity inversion and a 180 degree phase shift at any frequency
(within the passband of the filter).
You could argue that a 180 degree phase shift isn't the same as a
polarity inversion because you're also introducing a delay, but other
than that I don't see why a 180 degree phase shift would not be the
same as inverting polarity.
--th
2009/4/5 Hamish Allan <email@hidden>:
On Sun, Apr 5, 2009 at 6:56 PM, Brian Willoughby
<email@hidden> wrote:
Even though music can
theoretically be transformed into a series of sine waves that
repeat to
infinity, inverting the polarity of a piece of music is not going
to shift
each frequency by a certain number of degrees.
Would it not give you the exact same result as performing a perfect
Fourier decomposition, phase-shifting each frequency by 180 degrees,
and recomposing?
FWIW, I agree with you that it's not "a phase operation" -- it could
be viewed as an infinite sum of phase operations, but why look at it
that way, when you could just look at it as a polarity operation?
Hamish
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Coreaudio-api mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden