[mirtoolbox] Re: mirchromagram with 'nbins' option

  • From: MJ McVicar <Matt.McVicar@xxxxxxxxxxxxx>
  • To: mirtoolbox@xxxxxxxxxxxxx
  • Date: Mon, 10 May 2010 09:09:21 +0100

Ah thanks Olivier this is just what I needed!

I'm pretty sure this is the case, but using the ('tuning',t) option, t is the centre of the frequency bin right? Not the endpoint.

Matt

--On Monday, May 03, 2010 1:00 PM +0200 Olivier Lartillot <olartillot@xxxxxxxxx> wrote:

Dear Matt,

It was not possible so far to get the central frequency of each chroma.
But I just modified slightly mirchromagram. Unpack the enclosed file, and
replace your previous @mirchromagram subfolder with the new one, and
restart Matlab.

Now you can do the following:

c = mirchromagram(..., 'Tuning', t)   % where t is the frequency
associated to chroma C get(c, 'ChromaFreq')   % will give you the central
frequency for each chroma (in the unwrapped representation)

M McVicar kirjoitti 2.5.2010 kello 22.32:

Dear list,

when I call 'mirchromamgram' with the 'nbins' option (let's say I set
nbins=3) which frequency bands are the resulting 12*3 dimensional feature
in?

In your case, you should use 'Nbins', 12*3   (instead of 3)


are they grouped like this:

...
A4 + 2 thirds of a semitone +octaves
A4 + third of a semitone +octaves
standard A4, 440 Hz + octaves
...

or is it symmetric, like this:

...
A4 + third of a semitone +octaves
standard A4, 440 Hz + octaves
A4 - third of a semitone +octaves
...

My problem is I have mis-tuned songs and I wish to use different
reference tunings for some of them. As far as I know I can't call

get(my_chromagram,'Frequency')

as the octaves are being summed right? I hope this is clear despite my
clumsy notation!

Matt
--
Matt McVicar, Bristol Centre for Complexity Science
mm4819@xxxxxxxxxx



Olivier




----------------------
MJ McVicar
mm4819@xxxxxxxxxxxxx

Other related posts: