Author Topic: Batch Processing .svl annotation files  (Read 6416 times)

fabien

  • Newbie
  • *
  • Posts: 1
    • View Profile
Batch Processing .svl annotation files
« on: November 25, 2009, 10:44:07 »
Dear all,
I would like to batch process a series of .svl files with time indexes annotations into simple text files with these annotations.
I know it is possible to open an .svl file in Sonic Visualiser and then save the annotation as a text file, but I was looking for a more automated process that could deal with a bunch of files at once.
Is there a way to do that?
Thanks.
Fabien

davidf

  • Newbie
  • *
  • Posts: 10
    • View Profile
Re: Batch Processing .svl annotation files
« Reply #1 on: May 02, 2010, 09:37:30 »
Yes Please! You are right.

.svl and .csv files do not carry the same data. I am at the point now where I need both data sets to work with.

a .svl file carries precise time data at precise or as close as  possible to iincriments of .02 seconds each. It is matched with (in my own case) power levels or decibel levels. There is, nowever, no way to match this information with a bar count or any way to find a corresponding spot in a piece of music.

It would be most helpful to have the DTD  and the XSL for his file type. It  isn't that we don't have applications to read the file, the applications do not recognize the extension as a valid XML document.

I am working with Open Office, the extension for Lilypond and OpenOffice Calc at the moment and I am at a standstill until I can match power or decibel levels with precise points in a score and a sound file.

Does anyone have any idea of how to do this?

thanks.
davidf

cannam

  • Administrator
  • Sr. Member
  • *****
  • Posts: 273
    • View Profile
Re: Batch Processing .svl annotation files
« Reply #2 on: May 07, 2010, 14:46:49 »
Can you perhaps provide some examples (either as attached files, or by cut-and-paste from their contents) of the sorts of data you find you need that are present in the .svl export but not in .csv or vice versa?

It would not be hard to crank out a Perl script or similar to turn a .svl file into .csv, at least for files containing sparse data.  There's no formal DTD but the format is fairly obvious: just a dataset element containing a series of point elements each with frame (i.e. audio sample number), value, duration, label etc attributes.


Chris