Subject: Re: [linux-audio-dev] Perl module for GUIDO music notation
From: Karl MacMillan (karlmac_AT_peabody.jhu.edu)
Date: Mon Jan 22 2001 - 22:40:21 EET
On Mon, 22 Jan 2001, Karl Ward wrote:
> Hi everyone. I'm nearing completion a Perl module that can read,
> manipulate, and write GUIDO score files. For more info on GUIDO, visit:
>
> http://www.informatik.tu-darmstadt.de/AFS/GUIDO/
>
Great!
> But I decided to use GUIDO because I feel it is better engineered.
> Unfortunately, it seems the GUIDO developers either don't respond to email
> or are not working on GUIDO right now, so I can't ask them for advice.
>
I know that they are still actively working on Guido as I spoke to them at
the Music Information Retrieval conference recently
(http://ciir.cs.umass.edu/music2000/). They both have University jobs
that are not related to Guido directly and they are probably busy with the
beginning of the semester.
> The main question is, should I go out of my way to extract generic
> subroutines and internal representations that would be useful for any
> notation language (abc, GUIDO, MusiXML) and place those in a Perl module
> called Music::Notation, leaving the specifics in Music::Notation::GUIDO ?
> This might be useful, but would be time-consuming and would probably delay
> my work on the GUIDO part considerably. If any of you are Perl hackers, I'd
> like for you to take a look at my code and tell me what you think. An alpha
> version of the code is available at
>
> http://www.columbia.edu/~karlward/GUIDO.tar
>
> It may be that a generic Music::Notation module wouldn't be of much use,
> because each music notation language has different parameters and such. I'm
> afraid that creating a generic Music::Notation module, or to be really
> generic, a Music module, might be the equivalent of creating yet another new
> representational scheme for music (i.e. reinventing the wheel). There was
> some brief discussion of doing this a while back (1998):
>
> http://www.xray.mpe.mpg.de/mailing-lists/modules/1998-11/msg00127.html
>
> but it appears that nothing ever came of it.
>
This is certainly a tough problem and one that a lot of people have
struggled with. A generic Music::Notation module would be a HUGE task and
one that you might not really want to solve. It seems like a reasonable
project to write a Guido module, but a good general purpose notation
module would be a task that would either a) take all of your time or b)
not really solve the problem that it is trying to solve (i.e. not be broad
and general enough).
This is based on my experience as one of the developers for the Levy
Project - http://levysheetmusic.mse.jhu.edu/ - which is a very large sheet
music that has been digitized (about 29,000 peices). I work on Adaptable
Optical Music Recognition software for use with the collection. This
software is divided into two parts - the symbol recognition part (which I
work on) and what we call Optical Music Interpretation (which my colleague
Micheal Droettboom works on). The OMI part has to solve exactly the same
problem as your Music::Notation module - it has to be able to represent
and interpret almost any type of western music notation and then output it
in a variety of file formats (currently Guido!). Mike has some papers
about how he has solved these problems in python available on his website
- www.peabody.jhu.edu/~mdboom. All of this software is 'alpha' (i.e. it
works for us but we are not brave enough to claim it will work for you
:)) The web site for the AOMR software is mambo.peabody.jhu.edu/omr. I
hope that this helps.
Karl
P.S. I am always eager to help another Karl with a K!
> Thanks for listening. All you Perl hackers let me know what you think at
> karlward_AT_columbia.edu.
>
> -k
>
>
_____________________________________________________
| Karl W. MacMillan |
| Computer Music Department |
| Peabody Institute of the Johns Hopkins University |
| karlmac_AT_peabody.jhu.edu |
| www.peabody.jhu.edu/~karlmac |
-----------------------------------------------------
This archive was generated by hypermail 2b28 : Mon Jan 22 2001 - 23:15:30 EET