MDB-mf0-synth

MDB-mf0-synth
=============

MDB-mf0-synth (c) by Justin Salamon, Rachel Bittner, Jordi Bonada, Juan Jose Bosch, Emilia Gómez and Juan Pablo Bello.
MDB-mf0-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-mf0-synth contains 85 songs from the MedleyDB dataset (http://medleydb.weebly.com/) in which polyphonic pitched 
instruments (such as piano and guitar) have been removed and all monophonic pitched instruments (such as bass and voice) 
have been resynthesized to obtain perfect f0 annotations 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:
* 85 stereo wav files of song mixes where:
    * polyphonic pitched instruments (such as piano and guitar) have been removed
    * all monophonic pitched instruments (such as bass and voice) have been resynthesized using the analysis/synthesis 
      method described in the paper
* 85 csv files containing a perfect multiple-f0 annotation of all the (monophonic) pitched instruments in the mix, 
  obtained via the analysis/synthesis method described in the paper

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

audio_mix
---------
Contains 85 stereo wav files of song mixes in which polyphonic pitched instruments (such as piano and guitar) have been 
removed and all monophonic pitched instruments (such as bass and voice) have been resynthesized using the 
analysis/synthesis method described in the paper. Non-pitched tracks (percussion) are kept unchanged (i.e. the 
original stems are used). All the stems (tracks) are automatically mixed together as described in the paper.

Naming convention: 
<artist>_<songtitle>_MIX_mf0synth.wav

Example: 
AClassicEducation_NightOwl_MIX_mf0synth.wav

annotation_mf0
--------------
Contains 85 csv files containing a perfect multiple-f0 annotation of all pitched stems (tracks) in the mix, obtained 
via the analysis/synthesis method described in the paper. 

Format:
The annotations follow the MIREX multiple-f0 estimation (frame-basis) format:
https://www.music-ir.org/mirex/wiki/2018:Multiple_Fundamental_Frequency_Estimation_%26_Tracking#I.2FO_format
This format is also support by mir_eval: https://github.com/craffel/mir_eval

Each row in the annotation starts with a timestamp, followed by 0 or more tab separated frequency values in Hz 
representing the f0 of each active pitched instrument present in the time frame represented by the row. The first 
frame in the annotation is zero-centered. The hop size of the annotation is exactly 10 ms. 

IMPORTANT: no assumptions can be made as to the ordering of the f0 values in each row. The frequency values are NOT 
ordered neither by instrument nor by frequency, and should thus be treated as a "bag of frequencies" (a set) without 
any assumptions as to which frequency belongs to which instrument.

Naming convention:
<artist>_<songtitle>_MIX_mf0synth.csv

Example:
AClassicEducation_NightOwl_MIX_mf0synth.csv

Please Acknowledge MDB-mf0-synth in Academic Research
-----------------------------------------------------

Please cite the following publication when using MDB-mf0-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-mf0-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-mf0-synth dataset or any part of it.

Feedback
--------

Please help us improve MDB-mf0-synth by sending your feedback to: justin.salamon@gmail.com
In case of a problem report please include as many details as possible.