Silvet, or Shift-Invariant Latent Variable Transcription, is a Vamp plugin for polyphonic music transcription (from audio to note times and pitches).

What can I use it for?

In conjunction with a Vamp plugin host like Sonic Visualiser, you can use Silvet to help you work out what notes are being played in a piece of music, for example if you want to learn to play it yourself.

You might also use it to study performances for musicological features such as timing and dynamics.

Silvet also serves as a useful stable baseline for comparative purposes, for researchers working on other methods related to note transcription.

Silvet uses a high-quality and quite flexible method, but it has various limitations which you can find described in the README file. Although you can easily get interesting and useful results for many kinds of music, don't expect it to take you straight from the audio to a complete and readable score!

How does it work?

It uses the method described in A Shift-Invariant Latent Variable Model for Automatic Music Transcription from Emmanouil Benetos and Simon Dixon.

See the README file for more information.


Silvet is Free Software published under the GNU General Public License (GPL) v2 or later.

In addition to the formal licence terms, if you use Silvet in any public or commercial context, we ask you to kindly acknowledge the authors and Queen Mary, University of London in your user-visible documentation.

If you use Silvet in academic work, please cite the publication listed in the CITATION file.

Related projects

Issue tracking

View all issues

Related publications

E. Benetos and S. Dixon, “A Shift-Invariant Latent Variable Model for Automatic Music Transcription,” Computer Music Journal, vol. 36, no. 4, pp. 81–94, 2012.
[More Details] [BIBTEX] [URL (ext.)]
E. Benetos and S. Dixon, “Multiple-F0 Estimation and Note Tracking for MIREX 2012 using a Shift-Invariant Latent Variable Model,” in Music Information Retrieval Evaluation Exchange (MIREX), 2012.
[More Details] [BIBTEX] [URL (ext.)]
E. Benetos and S. Dixon, “Multiple-instrument polyphonic music transcription using a temporally constrained shift-invariant model,” Journal of the Acoustical Society of America (JASA), vol. 133, no. 3, pp. 1727–1741, 2013.
[More Details] [BIBTEX] [URL (ext.)]
E. Benetos and A. Holzapfel, “Automatic transcription of Turkish microtonal music,” Journal of the Acoustical Society of America, vol. 138, no. 4, pp. 2118–2130, Oct. 2015.
[More Details] [BIBTEX] [URL (ext.)]


Manager: Chris Cannam, Emmanouil Benetos