[mirtoolbox] mirplayer & mireventdensity bugs
- From: David Hirst <dhirst0@xxxxxxxxx>
- To: mirtoolbox@xxxxxxxxxxxxx
- Date: Mon, 8 Dec 2014 16:15:55 +1100
Dear Olivier,
I was just wondering if the 1.6.1 update addresses the bugs I reported recently
with respect to:
mirplayer
and
mireventdensity
See my previous posts below...
Cheers,
David
<snip>_____________________________________________________________
Hi,
I’m new to the MIRToolbox & it looks like a wonderful thing, but in the testing
phase of working through the Primer I think I’ve come across a bug in mirplayer:
For example, on page 8, the recommended sequence of commands generates the
following errors when I click on the “spectral flux” checkbox in the Player
window:
>> s = mirspectrum('ragtime.wav', 'Frame')
Computing miraudio related to ragtime.wav...
Computing mirsum related to ragtime.wav...
Computing mirspectrum related to ragtime.wav...
Elapsed time is 0.109951 seconds.
s is the Spectrum related to ragtime.wav, of sampling rate 44100 Hz.
Its content is displayed in Figure 9.
>> mirflux(s)
Computing mirflux related to ragtime.wav...
The Spectral flux related to file ragtime.wav is displayed in Figure 10.
>> f=mirflux(s)
Computing mirflux related to ragtime.wav...
The Spectral flux related to file ragtime.wav is displayed in Figure 11.
>> mirplayer(f)
Computing miraudio related to ragtime.wav...
Computing mirsum related to ragtime.wav...
Elapsed time is 0.078186 seconds.
Error using axes
Object is not parented to a figure yet
Error in mirplayer/selectFeature (line 917)
axes(featureAxes{nAxes});
Error in mirplayer/selectFeatureCallback (line 869)
selectFeature(songInd, selectedFeature);
Error while evaluating UIControl Callback
>>
Is there something I’m doing wrong, or is this a code error? It is such a pity
it isn’t working because the mirplayer functionality - to be able to play the
sound file while the graph is being displayed with a moving playback head - is
so useful. This happens with mirplayer no matter which other mir function one
chooses.
My system is Mac OS X Yosemite and I have the latest MATLAB install
MATLAB_R2014b. The Machine is an iMac from 2012.
BTW: I’m exploring the MIRToolbox for applications in the analysis of
Electroacoustic Music.
Best Regards,
DavidUpon further testing, ‘mireventdensity’ doesn’t seem to produce any
graphic output after appearing to run successfully, see:
>> mireventdensity('ragtime.wav', 'Frame')
Chunk 1/4...
Computing miraudio related to ragtime.wav...
Computing mirsum related to ragtime.wav...
Computing mirspectrum related to ragtime.wav...
Computing mirenvelope related to ragtime.wav...
Computing mironsets related to ragtime.wav...
Computing mirsum related to ragtime.wav...
Chunk 2/4...
Computing miraudio related to ragtime.wav...
Computing mirsum related to ragtime.wav...
Computing mirspectrum related to ragtime.wav...
Computing mirenvelope related to ragtime.wav...
Computing mironsets related to ragtime.wav...
Computing mirsum related to ragtime.wav...
Chunk 3/4...
Computing miraudio related to ragtime.wav...
Computing mirsum related to ragtime.wav...
Computing mirspectrum related to ragtime.wav...
Computing mirenvelope related to ragtime.wav...
Computing mironsets related to ragtime.wav...
Computing mirsum related to ragtime.wav...
Chunk 4/4...
Computing miraudio related to ragtime.wav...
Computing mirsum related to ragtime.wav...
Computing mirspectrum related to ragtime.wav...
Computing mirenvelope related to ragtime.wav...
Computing mironsets related to ragtime.wav...
Computing mirsum related to ragtime.wav...
Computing mirenvelope related to ragtime.wav...
Frame length longer than total sequence size. No frame decomposition.
Computing mirpeaks related to ragtime.wav...
Computing mireventdensity related to ragtime.wav...
Computing mirpeaks related to ragtime.wav...
Elapsed time is 1.127756 seconds.
The Event density related to file ragtime.wav is 4.6569 per second
… and then NO Figure is displayed.
cheers,
David
Associate Professor David Hirst, PhD
Honorary Principal Fellow
School of Contemporary Music | Faculty of the VCA and MCM | The University of
Melbourne
234 St Kilda Road, SOUTHBANK 3006
Ph: 0413 325 001 Fax: 9035 9461 www.vca.unimelb.edu.au
<http://www.vca.unimelb.edu.au/>
http://davidhirstmusic.wordpress.com/about/
<http://davidhirstmusic.wordpress.com/about/>
</snip>
> On 8 Dec 2014, at 9:14 am, Olivier Lartillot <olartillot@xxxxxxxxx> wrote:
>
> The bug has finally been addressed in the new update 1.6.1 just released.
>
>> On 30 Jun 2014, at 11:47, Haus, Michael Andreas <michael.haus@xxxxxx
>> <mailto:michael.haus@xxxxxx>> wrote:
>>
>> I'm trying to calculate high-level audio features with
>> miremotion('song.wav').
>>
>> But I get the following error:
>>
>> ----------------
>> Computing convolution, please wait...
>> Computing miremotion related to 1005382_AC_DC - Rock n Roll Dream.wav...
>> Error using cat
>> Dimensions of matrices being concatenated are not consistent.
>>
>> Error in cell2mat (line 75)
>> m{n} = cat(2,c{n,:});
>>
>> Error in miremotion>happy (line 413)
>> ha_f(3) = 0.4047*((std(cell2mat(cp)) - 8.5321)./2.5899);
>>
>> Error in mircompute (line 29)
>> res = algo(vk{:});
>> ----------------
>>
>> How can I avoid the error? Should the audio files are preprocessed before
>> using the miremotion method?
>>
>> The audio files are in wav format and has a sampling rate of 44.1 kHz.
>>
>> I'm using the version 1.3 of MIRToolbox.
>
Other related posts:
- » [mirtoolbox] mirplayer & mireventdensity bugs - David Hirst
