Matchmaker is a Python library for real-time music alignment.
Music alignment is a fundamental MIR task, and real-time music alignment is a necessary component of many interactive applications (e.g., automatic accompaniment systems, automatic page turning).
Unlike offline alignment methods, for which state-of-the-art implementations are publicly available, real-time (online) methods have no standard implementation, forcing researchers and developers to build them from scratch for their projects.
We aim to provide efficient reference implementations of score followers for use in real-time applications which can be easily integrated into existing projects.
The full documentation for matchmaker is available online at readthedocs.org.
- Available Python version: 3.12
- Fluidsynth
- PortAudio
First, install Fluidsynth, and then install the pyfluidsynth Python library. We recommend to install Fluidsynth using conda as well (see instructions below).
Note that pyfluidsynth only provides Python bindings for Fluidsynth; it does not install Fluidsynth itself. Be aware that there is also a fluidsynth Python library (without the py- prefix), but it is not compatible with matchmaker. We recommend installing Fluidsynth using conda
Setting up the code as described here requires conda. Follow the instructions for your OS.
To setup the experiments, use the following script.
# Clone matchmaker
git clone https://github.com/pymatchmaker/matchmaker.git
# Create the conda environment
conda create -n matchmaker python=3.12
conda activate matchmaker
# Go to matchmaker directory
cd ../matchmaker
# Install matchmaker in editable mode
pip install -e ."[dev]"
# Install GCC
conda install -c conda-forge gcc=12.1.0
# Install glib and fluidsynth
conda install -c conda-forge glib fluidsynthBecause of the dependency of partitura, which uses MuseScore_General.sf3 (free soundfont provided by MuseScore) as the default soundfont, the soundfont will be installed automatically inside the partitura package. This might take a while for the first time.
The solution seems to be to download vs_BuildTools.exe from https://visualstudio.microsoft.com/visual-cpp-build-tools/ and then execute
vs_buildtools.exe --norestart --passive --downloadThenInstall --includeRecommended --add Microsoft.VisualStudio.Workload.NativeDesktop --add Microsoft.VisualStudio.Workload.VCTools --add Microsoft.VisualStudio.Workload.MSBuildToolsOn Windows, pyfluidsynth expects fluidsynth.exe to be located in C:\tools\bin (other users have reported that it is expected in C:\tools\fluidsynth\bin). You can fix the issue by
- Get the ZIP file for your Windows version from https://github.com/FluidSynth/fluidsynth/releases/latest
- Extract the contents to
C:\tools(or wherever pyfluidsynth expects the executable to be).
We recommend to install Fluidsynth from conda in a dedicated environemnt. If however, you want to use the system-wide Fluidsynth installed with homebrew, you might run into an ImportError("Couldn't find the FluidSynth library.") with pyfluidsynth. Please refer to the following link.
To get started quickly, you can use the Matchmaker class, which provides a simple interface for running the alignment process. You can use a musicxml or midi file as the score file. Specify "audio" or "midi" as the input_type argument, and the default device for that input type will be automatically set up.
from matchmaker import Matchmaker
mm = Matchmaker(
score_file="path/to/score",
input_type="audio",
)
for current_position in mm.run():
print(current_position) # beat position in the scoreThe returned value is the current position in the score, represented in beats defined by partitura library's note array system.
Specifically, each position is calculated for every frame input and interpolated within the score's onset_beat array.
Please refer to here for more information about the onset_beat concept.
You can simulate the real-time alignment by putting a specific performance file as input, rather than running it as a live stream.
The type of performance file can be either audio file or midi file, depending on the input_type.
from matchmaker import Matchmaker
mm = Matchmaker(
score_file="path/to/score",
performance_file="path/to/performance.mid",
input_type="midi",
)
for current_position in mm.run():
print(current_position)To use a specific audio or MIDI device that is not the default device, you can pass the device name or index.
By default, input_type is set to “audio”. If you are using a MIDI device, you can change the input type to “midi”.
from matchmaker import Matchmaker
mm = Matchmaker(
score_file="path/to/score",
device_name_or_index="MacBookPro Microphone",
)
for current_position in mm.run():
print(current_position)The repository includes a ready-to-use example script that demonstrates the complete workflow:
# Run with audio input (default)
python run_examples.py
# Run with MIDI input
python run_examples.py --midiThis script runs a complete example with score following and evaluation, saving results to the results/ directory.
For testing with Audio input, you can specify the alignment method as follows:
from matchmaker import Matchmaker
mm = Matchmaker(
score_file="path/to/score",
input_type="audio",
method="dixon", # or "arzt" (default)
)
for current_position in mm.run():
print(current_position)For options regarding the method, please refer to the Alignment Methods section.
For options regarding the feature_type, please refer to the Features section.
Matchmaker currently supports the following alignment methods:
"dixon": On-line time warping algorithm by S. Dixon (2005). Currently supports audio input; MIDI support coming soon."arzt": On-line time warping algorithm adapted from Brazier and Widmer (2020) (based on the work by Arzt et al. (2010)). Currently supports audio input; MIDI support coming soon."hmm": Hidden Markov Model-based score follower by Cancino-Chacón et al. (2023), based on the state-space score followers by Duan et al. (2011) and Jiang and Raphael (2020). Currently supports MIDI input; Audio support coming soon.
Matchmaker currently supports the following feature types:
- For audio:
"chroma": Chroma features. Default feature type for audio input."mfcc": Mel-frequency cepstral coefficients."mel": Mel-Spectrogram."logspectral": Log-spectral features used in Dixon (2005).
- For MIDI:
pianoroll: Piano-roll features. Default feature type for MIDI input."pitch": Pitch features for MIDI input."pitchclass": Pitch class features for MIDI input.
Initialization parameters for the Matchmaker class:
score_file(str): Path to the score file.input_type(str): Type of input data. Options:"audio","midi".feature_type(str): Type of feature to use. Options:"chroma","mfcc","cqt","spectrogram","onset".method(str): Alignment method to use. Options:"dixon","arzt","hmm".sample_rate(int): Sample rate of the input audio data.frame_rate(int): Frame rate of the input audio/MIDI data.device_name_or_index(str or int): The audio/MIDI device name or index you want to use. IfNone, the default device will be used.
If you find Matchmaker useful, we would appreciate if you could cite us!
@inproceedings{park_matchmaker_2025,
title = {Matchmaker: {An} {Open}-{Source} {Library} for {Real}-{Time} {Piano} {Score} {Following} and {Systematic} {Evaluation}},
booktitle = {Proceedings of the 26th {International} {Society} for {Music} {Information} {Retrieval} {Conference} ({ISMIR} 2025)},
author = {Park, Jiyun and Cancino-Chacón, Carlos and Chiruthapudi, Suhit and Nam, Juhan},
address = {Daejeon, South Korea}
year = {2025}
}@inproceedings{matchmaker_lbd,
title={{Matchmaker: A Python library for Real-time Music Alignment}},
author={Park, Jiyun and Cancino-Chac\'{o}n, Carlos and Kwon, Taegyun and Nam, Juhan},
booktitle={{Proceedings of the Late Breaking/Demo Session at the 25th International Society for Music Information Retrieval Conference}},
address={San Francisco, USA.},
year={2024}
}This work has been supported by the Austrian Science Fund (FWF), grant agreement PAT 8820923 ("Rach3: A Computational Approach to Study Piano Rehearsals"). Additionally, this work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (No. NRF-2023R1A2C3007605).
This project is licensed under the Apache License 2.0 - see the LICENSE file for details.