[music-dsp] Re: Declipping

Joshua Scholar joshscholar at yahoo.com
Tue Feb 22 17:59:31 EST 2005


----- Original Message ----- 
From: "Nigel Redmon" <earlevel at earlevel.com>
To: "music-dsp" <music-dsp at shoko.calarts.edu>
Sent: Tuesday, February 22, 2005 2:44 PM
Subject: Re: [music-dsp] Re: Declipping
> As far as your question to me: You are comparing a complete samplerate
> conversion process in the frequency domain (FFT/append/IFT) to half of
> my samplerate conversion process in the time domain (just the
> zero-stuff part, leaving out the low pass filter that follows and fills
> in the zeros). Also, my post addressed the subject matter in the post I
> was replying too--frequency domain wasn't involved--there's nothing
> wrong with it, just not what we were talking about.
>

Oh don't take this so seriously.  You slipped up and said "that's what's
really there" which made this an arguement about mathematical aesthetics not
about filtering.

I was just pointing out that looking at it from a different mathmatical view
point you can "zero stuff" and get a filtered signal!

Anyway we all know all of the math you're going over.  All we're really
finding out is that you do you math based on a visual image of impulses -
well _I_ do my math based on a visual image of a continuous, bandlimited
signal.

I think your model is a bit odd - we are listening to bandlimited sound
after all, but if you can keep from making mistakes while thinking about
samples as discrete impulses, then you don't need any help to think like the
rest of us.

Joshua Scholar




More information about the music-dsp mailing list