Sample of the MedleyDB dataset: http://medleydb.weebly.com/

MedleyDB was curated primarily to support research on melody extraction, addressing important shortcomings of existing collections. For each song we provide melody f0 annotations as well as instrument activations for evaluating automatic instrument recognition. The dataset is also useful for research on tasks that require access to the individual tracks of a song such as source separation and automatic mixing.

Audio files for the second release of the MedleyDB multitrack dataset (MedleyDB 2.0). Annotation and Metadata files are version controlled and are available in the MedleyDB github repository: Metadata can be found here, Annotations can be found here.

*THE MULTITRACKS IN THIS RELEASE ARE NEW. THIS DOES NOT CONTAIN ANY OF THE MULTITRACKS FROM THE ORIGINAL RELEASE OF MEDLEYDB!*

For detailed information about the dataset, please visit MedleyDB's website.

If you make use of MedleyDB 2.0 for academic purposes, please cite the following white paper:

Rachel M. Bittner, Julia Wilkins, Hanna Yip and Juan P. Bello, "MedleyDB 2.0: New Data and a System for Sustainable Data Collection" Late breaking/demo extended abstract, 17th International Society for Music Information Retrieval (ISMIR) conference, August 2016.

Audio files for the MedleyDB multitrack dataset. Annotation and Metadata files are version controlled and are available in the MedleyDB github repository: Metadata can be found here, Annotations can be found here.

For detailed information about the dataset, please visit MedleyDB's website.

If you make use of MedleyDB for academic purposes, please cite the following publication:

R. Bittner, J. Salamon, M. Tierney, M. Mauch, C. Cannam and J. P. Bello, "MedleyDB: A Multitrack Dataset for Annotation-Intensive MIR Research", in 15th International Society for Music Information Retrieval Conference, Taipei, Taiwan, Oct. 2014.

Subset of MedleyDB: 103 solo monophonic stem audio files and corresponding manually annotated pitch (f0) annotations.

For further details, refer to the MedleyDB website.

Further Annotation and Metadata files are version controlled and are available in the MedleyDB github repository: Metadata can be found here, Annotations can be found here.

For detailed information about the dataset, please visit MedleyDB's website.

If you make use of MedleyDB for academic purposes, please cite the following publication:

R. Bittner, J. Salamon, M. Tierney, M. Mauch, C. Cannam and J. P. Bello, "MedleyDB: A Multitrack Dataset for Annotation-Intensive MIR Research", in 15th International Society for Music Information Retrieval Conference, Taipei, Taiwan, Oct. 2014.

MedleyVox is an evaluation dataset for multiple singing voices separation, which consists of 381 segments (1.1hour), containing 23 songs from MedleyDB v1 and v2 (https://medleydb.weebly.com).

MedleyVox contains 1) unison, 2) duet, 3) main vs. rest (folder name 'rest') , and 4) N-singing ('unison' + 'duet' + 'rest', please check our dataloader code for N-singing in https://github.com/jeonchangbin49/MedleyVox/blob/main/svs/data/test_dataset.py) categories explained in our ICASSP 2023 paper. For more details, please check our paper (https://arxiv.org/pdf/2211.07302.pdf) and code repository (https://github.com/jeonchangbin49/MedleyVox).

Acknowledgment

We are grateful to Rachel Bittner, the author of the original MedleyDB data, for allowing us to publish the MedleyVox dataset.

License

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.

Subset of MedleyDB: Mix audio files and Melody Annotations for the 108 files in the MedleyDB multitrack dataset containing melody.

There are 3 types of melody annotations released:

• melody1: "The f0 curve of the predominant melodic line drawn from a single source"
• melody2: "The f0 curve of the predominant melodic line drawn from multiple sources"
• melody3: "The f0 curves of all melodic lines drawn from multiple sources"

For further details, refer to the MedleyDB website.

Further Annotation and Metadata files are version controlled and are available in the MedleyDB github repository: Metadata can be found here, Annotations can be found here.

For detailed information about the dataset, please visit MedleyDB's website.

If you make use of MedleyDB for academic purposes, please cite the following publication:

R. Bittner, J. Salamon, M. Tierney, M. Mauch, C. Cannam and J. P. Bello, "MedleyDB: A Multitrack Dataset for Annotation-Intensive MIR Research", in 15th International Society for Music Information Retrieval Conference, Taipei, Taiwan, Oct. 2014.

Estimated notes for a subset of the MedleyDB dataset (pitch tracking subset) [1], created using the pyin vamp plugin [2].

[1] Bittner, Rachel M., et al. "Medleydb: A multitrack dataset for annotation-intensive mir research." ISMIR. Vol. 14. 2014.

[2] Mauch, Matthias, and Simon Dixon. "pYIN: A fundamental frequency estimator using probabilistic threshold distributions." 2014 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 2014.

MDB-bass-synth
==============

MDB-bass-synth (c) by Justin Salamon, Rachel Bittner, Jordi Bonada, Juan Jose Bosch, Emilia Gómez and Juan Pablo Bello.
MDB-bass-synth is licensed under the Creative Commons Attribution-NonCommercial 4.0 International License (CC BY-NC 4.0).
You should have received a copy of the license along with this work. If not, see http://creativecommons.org/licenses/by-nc/4.0/

Created By
----------

Justin Salamon*, Rachel Bittner*, Jordi Bonada^, Juan Jose Bosch^, Emilia Gómez^ and Juan Pablo Bello*.
* Music and Audio Research Lab (MARL), New York University, USA
^ Music Technology Group, Universitat Pompeu Fabra, Spain
http://synthdatasets.weebly.com/
http://steinhardt.nyu.edu/marl/
https://www.upf.edu/web/mtg

Version 1.0.0

Description
-----------

MDB-bass-synth contains 71 songs from the MedleyDB dataset (http://medleydb.weebly.com/) in which the bass track
has been resynthesized to obtain a perfect bass f0 annotation using the analysis/synthesis method described in the
following publication:

J. Salamon, R. M. Bittner, J. Bonada, J. J. Bosch, E. Gómez, and J. P. Bello. "An analysis/synthesis framework for
automatic f0 annotation of multitrack datasets". In 18th Int. Soc. for Music Info. Retrieval Conf., Suzhou, China,
Oct. 2017.

This dataset includes:
* 71 stereo wav files of song mixes where the original bass stem (track) has been replaced with a resynthesized one
* 71 mono wav files containing the resynthesized bass stem (track) only
* 71 csv files containing a perfect f0 annotation of the bass stem (track) obtained via the analysis/synthesis
method described in the paper

The data come in three folders, the contents of which is described below.

audio_mix
---------
Contains 71 stereo wav files of song mixes where the original bass stem (track) has been replaced with a
resynthesized one. All other stems (tracks) in the mix are the original ones (not resynthesized). The resynthesized
bass is obtained via the analysis/synthesis method described in the paper, and is automatically mixed together
with the rest of the stems as described in the paper.

Naming convention:

Example:
AClassicEducation_NightOwl_MIX_basssynth.wav

audio_bass
----------
Contains 71 mono wav files of the resynthesized bass stem (track) only. The resynthesized bass is obtained via the
analysis/synthesis method described in the paper.

Naming convention:

Example:
AClassicEducation_NightOwl_STEM_01.RESYN.wav

annotation_bass
---------------
Contains 71 csv files containing a perfect f0 annotation of the bass stem (track) obtained via the analysis/synthesis
method described in the paper.

Format:
Each file contains two comma-separated columns, the first containing timestamps and the second containing the bass
f0 in Hz. The first frame in the annotation is zero-centered. Silence is indicated as 0 Hz. The hop size of the
annotation is 128/44100 seconds (~2.9 ms).

Naming convention:

Example:
AClassicEducation_NightOwl_STEM_01.RESYN.csv

Please Acknowledge MDB-bass-synth in Academic Research
------------------------------------------------------

Please cite the following publication when using MDB-bass-synth:

J. Salamon, R. M. Bittner, J. Bonada, J. J. Bosch, E. Gómez, and J. P. Bello. "An analysis/synthesis framework for
automatic f0 annotation of multitrack datasets". In 18th Int. Soc. for Music Info. Retrieval Conf., Suzhou, China,
Oct. 2017.

For information about the original MedleyDB dataset please see (and cite):

R. M. Bittner, J. Salamon, M. Tierney, M. Mauch, C. Cannam, and J. P. Bello. MedleyDB: A multitrack dataset for
annotation-intensive MIR research. In 15th Int. Soc. for Music Info. Retrieval Conf., pages 155–160, Taipei, Taiwan,
Oct. 2014.

Conditions of Use
-----------------

Dataset created by Justin Salamon, Rachel Bittner, Jordi Bonada, Juan Jose Bosch, Emilia Gómez and Juan Pablo Bello.

The MDB-bass-synth dataset is offered free of charge under the terms of the Creative Commons
Attribution-NonCommercial 4.0 International License (CC BY-NC 4.0): http://creativecommons.org/licenses/by-nc/4.0/

The dataset and its contents are made available on an "as is" basis and without warranties of any kind, including
without limitation satisfactory quality and conformity, merchantability, fitness for a particular purpose, accuracy or
completeness, or absence of errors. Subject to any liability that may not be excluded or limited by law, NYU is not
liable for, and expressly excludes, all liability for loss or damage however and whenever caused to anyone by any use of
the MDB-bass-synth dataset or any part of it.

Feedback
--------

Please help us improve MDB-bass-synth by sending your feedback to: justin.salamon@gmail.com
In case of a problem report please include as many details as possible.