Time domain sample focused algorithms likely work better for many music
apps. In 10 weeks I would expect they'll cover these. The fact they've
made a course for this niche subject is a great thing in itself, kudos to
them.
________________________________________________
From:"Theo Verelst" <theo...@theover.org>
Sent:"A discussion list for music-related DSP"
<music-dsp@music.columbia.edu>
Date:Mon, September 8, 2014 7:50 pm
Subject:Re: [music-dsp] MOOC on Audio Signal Processing for Music
Applications
> Ming Li wrote:
>> Sorry for cross posting.
>> ...
>>
>> In collaboration with Prof. Julius Smith from Stanford
University, I
>> have
>> put together a 10 week long course on Audio Signal Processing for
Music
>> Applications in the Coursera online platform. ...
>
> Before everybody happily strings along with this fun enough
subject,
> consider the limitations of FFT once again (maybe more appealing to
tech
> students, but equally relevant for musicians, etc.):
>
> The FFT has various limitations, like the length of the FFT
interval,
> possible averaging schemes, bad dealing with transient information
in
> the intput signal, and various general accuracy limitations (how
many
> bins are there to "capture" certain frequencies ?)
>
> Sampling issues aren't solved by FFTs, and you might notice that
> processed transformed data being transformed back (by IFFT) into a
time
> domain sample sequence that is played over an audio card sounds
"weak"
> or "funny" in some ways: there are hard sampling
reconstruction reasons
> for that.
>
> If you'd use pro-audio of the traditionally well made kind as a
basis
> for these types of experiments, you may find that artists from long
agao
> already were able to "trick" you into measuring all kinds
of FFT things,
> and that those things are strangely (not so) correlated with the
music...
>
> T.
> --
> dupswapdrop -- the music-dsp mailing list and website:
> subscription info, FAQ, source code archive, list archive, book
reviews,
> dsp links
> http://music.columbia.edu/cmc/music-dsp
> http://music.columbia.edu/mailman/listinfo/music-dsp
>
--
Gadgets for your pockets, from our home-office
--
dupswapdrop -- the music-dsp mailing list and website:
subscription info, FAQ, source code archive, list archive, book reviews, dsp
links
http://music.columbia.edu/cmc/music-dsp
http://music.columbia.edu/mailman/listinfo/music-dsp
