You write the Melody Computer Handles Harmony"As you add more pieces to the mix, these can dilute the style and wash out the fine details that define what's special about a piece," said Chew, who is herself an accomplished pianist and seemed ideally suited to mentor Chuan's interest in music.
Success in nailing individual style may come from having a music theory framework for the program. Most programs take a bottom-up approach that starts from scratch without the rules of music theory to go by.
ASSA first uses learning techniques based on previous experience to identify notes that form the spine of the main melody, and builds accompaniment chords around those. It also identifies checkpoints where harmonies are clear at certain points in each song.
The program then applies Neo-Riemannian music theory to create chains of accompaniment chords between the checkpoints, based on the particular smooth music transitions that a musician stylistically prefers.
"The music theoretical knowledge does come from the top-down," Chew noted.
Chuan and Chew eventually want to create a program that can produce an entire song from a hummed melody.
Music CompuMaestroThe original London paper lays out the steps in the ASSA attack on the
problem. “The system first determines the chord tones in the melody.
[One ASSA] module applies machine learning techniques to choose the
chord tones from the input melody, based on the [chosen stylistic]
example pieces. The system uses 17 attributes to represent the melody."
Chew and Chuan at the 2008 USC Viterbi School graduate commencement at which Chuan received her PhD.
Then, chords are prescribed at checkpoints in the melody where
[the system finds] the harmony unambiguous. "Using these checkpoints as
anchors, we use neo-Riemannian transformations to build chord
progressions between checkpoints. Finally, we use Markov chains to
generate the final chord progression."
The initial study took four songs by Radiohead:
High and Dry,
Fake Plastic Trees, Airbag, and
Creep
and applied the rules derived from analyzing three of the songs to
generate an accompaniment for a fourth. That is, the bare melody of
song four was the starting point, to which the stylistic rules derived
from the other three songs were applied to create a new, ASSA
accompanied version.
Creep and
High and Dry were both used as tests.
Analyzing, the "overall correct rate" of the chord tone choices made by
ASSA was 82 percent (for a 54-note sample) in
Creep; and 70.5 percent on a 61-note sample for
High and Dry.
The tests used to evaluate the effectiveness of its accompaniment were
at least partially subjective, based on listeners’ opinions in Turing
tests, in which they had to guess if the accompaniment was the original
or machine-generated.
ASSA is not the first attempt to create a robotic accompaniment.
I-Ring (by Hong-Ru Lee and Jyh-Shing Roger Jang of National Tsing Hua
University) and MySong (by Ian Simon, Dan Morris, and Sumit Basu at
Microsoft Research) generate accompaniment based on training sets of
150 and 298 songs respectively -- songs from various genres are treated
homogenously as a whole for training the systems. David Temperley and
Daniel Sleator created another system, the rule-based Harmonic Analyzer
in Melisma.
Only ASSA and MySong aim to emulate style, and in MySong, style is
classified into two broad modes: happy and Jazz. The ASSA approach
treats style at a much more individual and specific level – as the
property of pieces from a particular period in a band’s output, or as
features unique to one individual piece.
To hear samples of the accompaniments created, go to:
http://www-scf.usc.edu/~chinghuc/Site/Samples.html
The two papers referred to can be downloaded in pdf form:
C. H. Chuan and E. Chew, "A Hybrid System for Automatic Generation of
Style-Specific Accompaniment," 4th International Joint Workshop on
Computational Creativity, London, UK, June 17-19, 2007.
www-rcf.usc.edu/~echew/papers/CC2007/chuanchew_cc07.pdf
C. H. Chuan and E. Chew, “Evaluating and Visualizing Effectiveness of
Style Emulation in Musical Accompaniment,” the 9th International
Conference on Music Information Retrieval, Philadelphia, PA, September
14-18, 2008.
http://ismir2008.ismir.net/papers/ISMIR2008_179.pdf
EVOLUTIONARY SCIENCE & CULTURE HUMAN SYNCRONY