Ríomhclar Ceol

Here you can read about and download my software for traditional Irish musicians. Some of this work was developed towards my PhD on the topic of Machine Annotation of Traditional Irish Music.

"as we enjoy great advantages from the inventions of others, we should be glad of an opportunity to serve others by any invention of ours; and this we should do freely and generously" - Benjamin Franklin

UPDATE: My paper "Compensating for Expressivness in Queries to a Content Based Music Information Retrieval System" - Was awarded the best presentation prize at ICMC 2009, Montreal, Canada.

Try out tunepal.org - a query-by-playing search engine for traditional tunes. (no install needed, but you will need a recent version of Java installed on your computer). Read Siobhan Long's coverage of tunepal.org in the Irish Times.

Examples of tune forms and ornamentation in traditional Irish dance music.

My PhD test audio (130 pieces) in uncompressed WAV format

My PhD Thesis: Machine Annotation of Traditional Irish Dance Music

Presentation on my PhD Thesis 16 November 2009 for DIT staff and students

A Wordle of my PhD Thesis:

TANSEY - Turn ANnotation from SEts using SimilaritY profiles

A set in traditional Irish music is a sequence of two or more dance tunes in the same time signature, where each tune is repeated an arbitrary number of times. A turn in a set represents the point at which either a tune repeats or a new tune is introduced. Tunes in sets are played in a segue (without an interval) and so detecting the turn is a significant challenge. The TANSEY algorithm is a novel algorithm for identifying turns in sets of traditional Irish music. TANSEY works on digitised audio files of monophonic flute and tin-whistle music. If you want to try out this algorithm yourself download and install MATT2. Edit the file matt.properties and set the value tansey=true.

TANSEY Test Audio

The Tunes. Warning! These are 16 bit, 44100 mono uncompressed WAV files and hence are very big!

Publications on TANSEY

(TANSEY is called MATS in my ISMIR paper)

Bryan Duggan, Brendan O'Shea, Mikel Gainza, Padraig Cunningham, "The Annotation of Traditional Irish Dance Music using MATT2 and TANSEY", The 8th Annual Information Technology & Telecommunication Conference, Galway Mayo Institute of Technology, Galway, Ireland, October 2008
B. Duggan, B. O'Shea, Mikel Gainza and P. Cunningham, "Machine Annotation of Sets of Traditional Irish Dance Tunes", Ninth International Conference on Music Information Retrieval (ISMIR), Drexel University, Philadelphia, USA, September 2008. Paper: Poster:

MATT2 - Machine Annotation of Traditional Tunes

The primary goal of MATT2 (Machine Annotation of Traditional Tunes) is to annotate recordings of traditional Irish dance music with useful metadata including tune names. MATT2 incorporates a number of novel algorithms for transcription of traditional music and for adapting melodic similarity measures to expressiveness in the playing of traditional music. It uses a matching algorithm tolerant to errors which aligns short queries with longer strings from a corpus of known tunes, meaning that the algorithm can match entire tunes, incipits and phrases from any part of tune with equal success.

Publications on MATT2

Bryan Duggan, Brendan O'Shea, Mikel Gainza, Padraig Cunningham, "Compensating for Expressivness in Queries to a Content Based Music Information Retrieval System", Accepted for publication at the 2009 International Computer Music Conference (ICMC 2009) in Montreal, Canada, August 2009 - Winner of the best presentation prize at ICMC 2009.
Bryan Duggan, Brendan O'Shea, Mikel Gainza, Padraig Cunningham, "The Annotation of Traditional Irish Dance Music using MATT2 and TANSEY", The 8th Annual Information Technology & Telecommunication Conference, Galway Mayo Institute of Technology, Galway, Ireland, October 2008
B. Duggan, B. O'Shea, and P. Cunningham, “A System for Automatically Annotating Traditional Irish Music Field Recordings,” the Sixth International Workshop on Content-Based Multimedia Indexing, Queen Mary University of London, UK, Jun. 2008

Installing MATT2

Install the latest version of the Java Runtime Environment.

Download MATT2.

Using ABC files transcribed by Henrik Norbeck
http://www.norbeck.nu/abc
Uses ABC4J by Lionel Gueganton
http://code.google.com/p/abc4j

Unzip all the files to a folder on your computer and click on PhD.bat to run. To annotate a live query:

Press Live Query and play a tune, part of a tune or a set of tunes.
This will save a WAV file of the recording to the folder tunometer/wavs. This file is likely to be big as its
recorded in mono 44100Khz 16 bit resolution.
Click Transcribe to transcribe an approximation of the tune in ABC notation.
Click Find to search for the transcription in the corpus.
Search results get displayed in order of similarity to the transcription.
Double click on a file to see the ABC, or click Play to play it.
You can add more tunes to the corpus by dropping ABC files into the folder searchCorpus. Don't forget
to click Reindex in the program otherwise MATT2 won't search the files.
This software should be considered Alpha, in other words not extensively tested. It works really well with Flute
and whistle queries.

Try editing the file matt.properties and clicking on the other buttons to change how the program works.

Source code for MATT2 & TANSEY is available under GNU Public License Agreement V2 and can be accessed via Google code:

You can also try out an online version of MATT2 without installing any software. This requires you have a recent version of Java installed (it wont run on a Mac as Mac's dont support Java 6 yet).

A Combinational Creativity Approach to Composing Traditional Irish Reels

In this work we present a system that uses a corpus of 864 traditional Irish reels as input into an algorithm that composes new tunes. The system performs a structural analysis of the tunes in the corpus and also counts n-gram note sequences in the tunes. It then recombines n-gram note sequences together in structures from the corpus to generate new tunes. The generated tunes were evaluated by 29 domain experts.

Listen to and evaluate the generated tunes!

Publications

Zheng N, Duggan, B: A Combinational Creativity Approach to Composing Traditional Irish Reels, 18th Irish Conference on Artificial Intelligence and Cognitive Science 29th - 31st August 2007, Dublin Institute of Technology. Paper: Poster:

TunePal

TunePal is a tool which facilitates convenient audio access to archive transcriptions of traditional tunes on a PDA. The main feature of TunePal is its ability to locate a tune by name or musical phrase from thousands stored on a PDA. TunePal enhances not only the recall of a musician using TunePal in a traditional music session, but also the acquisition of new repertoire.

Publications

Duggan, B: Enabling Access to Irish Traditional Music Archives on a PDA, Eight Annual Irish Educational Technology Users Conference, DIT Bolton St. Ireland, May 23rd - May 25th 2007. Paper:Poster:
Duggan, B: TunePal: A Portable Tune Teaching Tool for Traditional Musicians, DIT Annual Showcase of Learning & Teaching Activities, January, 2007
Duggan, B.: Learning Traditional irish Music using a PDA, IADIS Mobile Learning Conference, Trinity College, Dublin, Ireland, July 2006,

To download TunePal, see the TunePal homepage.

MATT1 - Machine Articulation of Traditional Tunes

MATT1 uses a combination of case based reasoning and wave table synthesis to simulate the creative interpretation of traditional Irish tunes on the wooden flute. This was early work on on my PhD, but I have abandoned this idea for now.

MATT1 Publications

Duggan, B., Zheng, C., Cunningham, P.: MATT - A System for Modelling Creativity in Traditional Irish Flute Playing, Third Joint Workshop on Computational Creativity, ECAI'06, Italy, August 2006,

Last Updated: 18 October 2009
Contact me at: bryan.duggan@dit.ie