Beets media library management

Overview

MusicPlayerPlus includes integrated installation and extensive configuration for the Beets media library management system. Beets is an open source, extensible, highly configurable suite of tools and plugins that organizes and improves the structure, cataloging, and metadata of a music library.

Management of your music library with Beets is an optional feature provided by MusicPlayerPlus. Although optional, use of Beets can enhance your music library in ways that make it more useful and easier to access with the mpcplus MusicPlayerPlus MPD client. For example, one of the automated tasks that Beets performs is updating the tags in your music library. Beets queries online sources like MusicBrainz, Bandcamp, and Last.fm to update the music library with widely used metadata for songs and albums it can identify. Subsequent invocations of mpcplus will be able to use this rich set of tags to filter, search, and find items in your music library.

The primary source of online data for Beets is MusicBrainz, a collaborative music database. As of March 2022, MusicBrainz contained information on roughly 1.9 million artists, 3 million releases, and 26.5 million recordings. This extensive set of data, along with additional data from other online sources like Bandcamp and Last.fm can be used to significantly improve the metadata content and organization of your music library.

[NOTE:] Beets is NOT the now defunct music service purchased by Apple. It is an open source media library management system.

Getting started with Beets

MusicPlayerPlus has integrated Beets initialization and use into the mppinit initialization command and the mpplus primary user interface. Due to its complexity, flexibility, and power, Beets can be fairly difficult to manage. MusicPlayerPlus attempts to ease the task of managing your music library with Beets. The following few steps are all that is necessary to get started with Beets after installing MusicPlayerPlus and initializing with mppinit.

Quickstart with Beets

After installing MusicPlayerPlus and running mppinit some common additional setup steps that can be performed include:

Configuring the music library location
Converting WAV format media files to MP3 format
Importing a music library into the Beets library management system
Downloading album cover art
Downloading additional lyrics
Activation of YAMS scrobbler for Last.fm
Analysis and retrieval of audio-based information for media matching a query

Configure the music library location by editing ~/.config/mpd/mpd.conf and setting the music_directory to your music library location (default setting is ~/Music). If you change the music_directory setting in mpd.conf then run the command mppinit sync.

The following optional post-initialization steps can be performed individually as described below or they can be performed in two steps using mppinit:

mppinit import

When the import is complete

mppinit metadata

The mppinit import command downloads album cover art, converts any WAV format media to MP3 format, and imports the music library into Beets. The mppinit metadata command identifies and deletes duplicate tracks, retrieves album genres from Last.fm, and optionally analyzes and retrieves metadata for all songs in the music library.

Album cover art can be downloaded with the command mpplus -D.

Convert WAV format media files in your library to MP3 format files with the command mpplus -F. Conversion from WAV to MP3 allows these files to be imported into the Beets media library management system.

If you wish to manage your music library with Beets, import the music library with the command mpplus -I.

Download additional lyrics with the command mpplus -L.

Activate the YAMS scrobbler for Last.fm with the command mpplus -Y.

Analysis and retrieval of audio-based information can be performed with the command mpplus -X 'query' where ‘query’ is a Beets library query. The special query term ‘all’ indicates the entire music library, i.e. mpplus -X all.

Quickstart summary

To summarize, a MusicPlayer quickstart can be accomplished by:

Install the latest Debian or RPM format installation package
Run the mppinit command as your normal user
If the music library is not located at $HOME/Music:
- Configure the music_directory setting by editing ~/.config/mpd/mpd.conf
- Run the command mppinit sync
Optionally:
- Perform these three steps with the command mppinit import
  - Download album cover art with the command mpplus -D
  - Convert WAV format files to MP3 format with the command mpplus -F
  - Import your music library into Beets with the command mpplus -I
- Perform these steps with the command mppinit metadata
  - Remove duplicate tracks with the command beet duplicates -d
  - Rename tracks left after duplicate removal with beet move
  - Analyze and retrieve audio-based information with a command like:
    - mpplus -X 'query' where ‘query’ is a Beets library query
    - mpplus -X all indicating analyze the entire Beets library
- Activate the YAMS scrobbler for Last.fm with the command mpplus -Y
- Download additional lyrics with the command mpplus -L

After initializing the Beets media library management system try playing something with a command like:

beet play QUERY

Where ‘QUERY’ is a valid Beets query. This can be a simple string like “blue” or “love” or a more complicated expression as described in the Beets query documentation. The Beets play plugin should match the query string to songs in your music library, add those songs to the MPD queue, and play them. Use beet ls QUERY to see what would be played.

Create a music library

If you do not already have one, create a music library on the system where MusicPlayerPlus is installed. This is the music library that MPD will be accessing and its location is defined in $HOME/.config/mpd/mpd.conf. The default location for the music library is $HOME/Music/. If you wish to store your music library in another location, e.g. on an external drive, then there will be a couple of extra steps to perform, outlined below.

It is recommended to organize the music library with a structure like Artist/Album/songs but this is not necessary. One of the ways Beets tries to identify albums and songs is by examining filenames and paths. If the music library is organized with top-level directories being artist names, subdirectories being album names, and files within album folders being songs from that album then Beets can use that structure to match artists and albums with online data. Should your music library be organized in a different fashion then Beets will still identify artists, albums, genres, songs, and other metadata using intelligent matching algorithms.

The music library must be writeable by the MusicPlayerPlus user.

Configure the music library location

If you have created a music library in a location other than $HOME/Music/ then edit $HOME/.config/mpd/mpd.conf and set the music_directory configuration to the location of your music library.

After configuring mpd.conf with the proper music_directory setting, run the command mppinit sync. This will update other MusicPlayerPlus and Beets configuration files that need to know the music library location.

If your music library location is $HOME/Music then you can skip this step.

Initializing the Beets media library management system

[NOTE:] Beets is NOT the now defunct music service purchased by Apple. It is an open source media library management system.

MusicPlayerPlus includes the Beets media library management system and preconfigured settings to allow easy integration with MPD and mpcplus. Beets is an application that catalogs your music collection, automatically improving its metadata. It then provides a suite of tools for manipulating and accessing your music. Beets includes an extensive set of plugins that can be used to enhance and extend the functionality of the media library management Beets provides. Many Beets plugins are installed and configured automatically by MusicPlayerPlus.

To get started using the Beets media library management system, it is necessary to import your music library into the Beets database. This process catalogs your music collection and improves its metadata. The default Beets configuration provided by MusicPlayerPlus moves and tags files in the music library during this process. It adds music library data to the Beets database. To import your music library into Beets, issue the following command:

mppinit import

or to skip WAV format media conversion and just perform the Beets import:

mpplus -I

Try playing something with a command like:

beet play QUERY

[NOTE:] MusicPlayerPlus has configured the Beets play plugin to use the command /usr/share/musicplayerplus/scripts/mpcplay.sh to play media with this plugin. This script clears the MPD queue, adds any songs matching the query to the queue, and plays the MPD queue. In addition, two arguments are supported: --shuffle and --debug. These additional arguments are passed using the --args feature. For example, to play all media matching the string “velvet” and shuffle the order of play, issue the command beet play --args --shuffle velvet.

Example usage of the beet play command:

beet play velvet
beet play playlist:1970s
beet play --args --shuffle playlist:1990s
beet play --args "--debug --shuffle" green eyes

Additional Beets metadata analysis and retrieval

After completing the Beets music library import with either mppinit import or mpplus -I, additional Beets metadata can be retrieved with the command:

mppinit metadata

This will identify and delete duplicate tracks, retrieve album genres, download album cover art, and optionally analyze and retrieve metadata for all songs in the music library using the Essentia extractor and Essentia trained models.

MusicPlayerPlus uses Essentia for extracting acoustic characteristics of music, including low-level spectral information, rhythm, keys, scales, and much more, and automatic annotation by genres, moods, and instrumentation.

This is the same sort of thing that AcousticBrainz does but the AcousticBrainz project is no longer collecting data and will be withdrawn in 2023. MusicPlayerPlus provides the same functionality using pre-compiled and packaged Essentia binaries and models.

However, the process of analyzing, extracting, and retrieving metadata can be time consuming for a large music library. The mppinit metadata command performs several metadata retrieval steps in a non-interactive manner and in the background so it can be left unattended if desired.

The individual metadata retrieval steps performed automatically by mppinit metadata can be performed manually using the instructions in the following sections.

Remove duplicate tracks

Duplicate tracks in the Beets library can be removed with the command:

beet duplicates -d

The ‘-d’ option removes duplicate files from the music library location on disk as well as removing duplicates in the Beets library. Omit the ‘-d’ option to remove duplicates in the Beets library only.

The Beets configuration file at $HOME/.config/beets/config.yaml has been configured to use checksums to find duplicates. This can be a fairly slow process for a large library. Beets can also find duplicates using MusicBrainz IDs or other methods. MusicPlayerPlus prefers checksums but if this process is too time consuming, it may be preferable to use IDs.

Rename tracks that were duplicates

After removing duplicates it is often the case that filenames contain no longer needed characters (e.g. “Love Song 2.mp3” which was a duplicate of “Love Song.mp3”). To correct these filenames, simply run beet move.

Retrieve genres from Last.fm

The Beets lastgenre plugin can be used to augment the genre metadata of a Beets music library. The MusicPlayerPlus configuration of the lastgenre Beets plugin includes a customized local copy of the genre tree and whitelist. To retrieve genres using the Beets lastgenre plugin, run the command beet lastgenre.

Download album cover art

To download album cover art for all albums in your music library, run the command mpplus -D.

Analyze and extract additional metadata with Essentia

To analyze and extract a wealth of additional metadata directly from all of the audio files in the music library, run the command mpplus -X all. To limit the analysis and extraction to a subset of audio files in the music library, run the command mpplus -X [QUERY] where ‘QUERY’ is a Beets query. For example, to analyze and extract metadata for all audio files with beets-per-minute set to 0, run the command mpplus -X bpm:0.

The analysis and extraction of metadata with Essentia can be a time consuming process. For this reason, it is run non-interactively in the background with a terminal window being used to monitor the progress. The extraction process can be left unattended and may need to run for hours. If the length of time required to analyze and extract metadata using Essentia is prohibitive then it may be preferable to use the Beets acousticbrainz plugin to retrieve additional metadata. To use acousticbrainz, uncomment the acousticbrainz entries in $HOME/.config/beets/config.yaml and run the command beet acousticbrainz [QUERY]. Using the acousticbrainz plugin is faster (the analysis has already been performed) but incomplete, often incorrect, out-of-date, and will no longer be available in 2023.

Convert WAV format media files

If your music library contains WAV format media files, these will not be imported into Beets. The WAV format is an older media format and does not support metadata in a way that Beets can manipulate. If you wish to import these files into Beets it is necessary to convert them to a format that is compatible with Beets. MusicPlayerPlus provides a convenience command to convert all WAV format media to MP3 format.

If the music libary does not contain any WAV format media files then this step can be skipped.

Import the music library into Beets

If you wish to manage your music library with Beets, import the music library with the command mpplus -I.

mpplus -I

The Beets import defaults are controlled by the Beets configuration file at $HOME/.config/beets/config.yaml and command-line options to the beet import command. By default, MusicPlayerPlus imports the MPD music library into Beets, moving rather than copying files to conform with standard detected artist/album/song naming conventions, and writing detected metadata. A log of the album import is written to the file $HOME/.config/beets/logs/import.log. A log of the singletons import is written to the file $HOME/.config/beets/logs/import_singletons.log. A log of the import times is written to the file $HOME/.config/beets/logs/import_time.log.

Prior to performing the intial Beets import of your music library, examine the import section of the Beets configuration at $HOME/.config/beets/config.yaml. Modify the import settings to suit your preferences. The default import settings configured by MusicPlayerPlus are:

import:
    copy: yes
    move: yes
    write: yes
    incremental: yes
    incremental_skip_later: yes
    quiet: no
    quiet_fallback: asis
    resume: yes
    from_scratch: no
    default_action: apply
    detail: yes
    non_rec_action: ask
    # duplicate_action can be 'skip', 'keep', 'remove', 'merge' or 'ask'
    duplicate_action: skip
    group_albums: no
    autotag: yes
    bell: no
    log: ~/.config/beets/logs/import.log

You may prefer to set move: no if you do not wish your files and folders to be moved or if you wish that rearrangement to be performed by copying rather than moving. Moving conserves disk space, copying preserves a libary’s structure but can consume much additional disk space. The MusicPlayerPlus default preference is to move rather than copy.

[NOTE:] Some users prefer to use MusicBrainz Picard to tag their music library. If you have previously tagged your music library using MusicBrainz Picard then you may wish to modify the default Beets import configuration to preserve Picard tagging and speed up import. An example import section of the Beets config.yaml that would preserve Picard tags might look like the following:

import:
    write: no
    copy: no
    move: no
    resume: ask
    incremental: yes
    incremental_skip_later: no
    from_scratch: no
    timid: yes
    log: ~/.config/beets/logs/import.log
    group_albums: no
    autotag: yes
    bell: no

The import process, depending on the size of the music library and metadata analyzed, may take several hours. For this reason, it is performed in the background and non-interactively. For extremely large music libraries (e.g. over 200GB) the import may need to run overnight. As long as your computer remains on and connected to the Internet, the import process should run uninterrupted and without need for attention. You may continue working or leave the import unattended.

An interrupted Beets import can be resumed simply by re-running the mpplus -I command. The import process will skip over previously imported items and import items the previous import failed to reach.

A rough estimate of the time to import a music library into Beets can be obtained by calculating the size of the library. On a moderately equipped Ubuntu Linux system, a test import of an 11GB music library of over 500 songs took about 10 minutes. Import times will vary with network speed, how quickly metadata sources like MusicBrainz and Bandcamp can locate media matches, and many other factors. Unfortunately, it is not possible to speed an import beyond the frequency with which MusicBrainz permits API requests.

A very large music library of hundreds of Gigabytes and thousands of songs will take hours to import. For this reason MusicPlayerPlus disables many of the automatic retrieval of metadata and tagging during import that some of the Beets plugins are capable of. This minimizes the import time. For example, the auto fetching of cover art, lyrics, and genre during import is disabled. These and other music library metadata retrieval can be performed post-import using MusicPlayerPlus commands.

A good workflow might be: Kickoff a Beets import, Enjoy some Sun with a good book, Check the import progress, Return to reading, Check progress, Repeat.

When running in the background, monitor the progess of the import by examining the log file. For example, to view the progress of the album import in real-time:

tail -f $HOME/.config/beets/logs/import_time.log

To view the progress of the singletons import in real-time:

tail -f $HOME/.config/beets/logs/import_singletons.log

The mpplus -I ... command performs both an import of the music library albums and an import of singleton songs.

Music library folders that are skipped are likely those for which beet import did not find the entire album. These songs will be picked up and imported during the subsequent import where singletons are identified and imported.

If a manual beets import is desired, it may be performed by issuing the commands:

beet import -[w|W] <MUSIC_DIRECTORY>

and

beet import -[w|W] -s <MUSIC_DIRECTORY>

Where is the full pathname to your music library and the `-w` flag indicates 'write metadata' while `-W` flag indicates 'do not write metadata'. Omitting both the `-w` and `-W` flags will use the *write* setting in the *import* section of `~/.config/beets/config.yaml` to determin if metadata are written. The first import command above imports albums from the music library at . The second import command imports singleton songs.

To perform a faster import without auto-tagging, writing tags, or copying, run the command:

beet import -qAWC

If you wish to update the tags in your music library after having imported without auto-tagging or writing, use the mbsync plugin to update tags using MusicBrainz: beet mbsync.

If auto-tagging, writing, moving, or otherwise modifying the music library during import, you may wish to backup your music library prior to importing. To view help on Beets importing, run the command beet help import.

If new songs or albums are added to the MPD music library, the Beets import can be re-run with mpplus -I or manually and only new albums and songs will be added to the Beets database as the import is performed incrementally. Incremental imports is a configuration option set in the import section of the Beets $HOME/.config/beets/config.yaml configuration file.

List the currently imported music library items with the command beet list and the currently imported music library albums with the command beet list -a.

View music library statistics with the beet describe command. For example, to view library genre statistics run the command beet describe genre. To view library beats per minute statistics run the command beet describe bpm.

The beet command-line reference is available at https://beets.readthedocs.io/en/latest/reference/cli.html

Learn more about the Beets media library management system at https://beets.io/

Beets import followup steps

The Beets import process is very smart and able to detect most albums and song titles. It uses an extensive database and several algorithms. But it does fail to correctly identify some albums and songs. MusicPlayerPlus has done quite a bit of work behind the scenes to make the Beets setup process as simple as possible. The import process is run non-interactively and in the background as is the metadata analysis and retrieval. For this reason, the Beets import process has made intelligent choices for you during import. These are almost always correct due to the algorithms used.

After completing a Beets import it is recommended to manually review the resulting music library. If there are albums which Beets was not able to identify then they will be in folders named something like:

$HOME/Music/John Doe/[0000] Unknown Album (6_19_2007 4_56_26 PM)

The Beet list command can be used to display a list of all albums or tracks in the imported library. Run beet ls -a to list all albums and beet ls to list all tracks. A summary of music library statistics can be obtained with beet stats. Information on a particular item or items can be obtained with beet info -l [QUERY] where QUERY is a Beets query. An empty query returns the entire library. To view info on a particular item, provide enough of the item filename or path to identify it as the query. For example, to view info on the items in the above unknown album by artist John Doe: beet info -l 'John Doe'.

Find and review all albums in the music library with folders containing Unknown Album with the command beet list -a 'Unknown Album' or by listing the folders in the music library itself with a command like ls -d "/path/to/music_library/*/*Unknown Album*".

Often a simple interactive re-import of these folders can be performed to properly identify them. Using the above unknown album as an example, to perform an interactive re-import of this unknown album:

beet import "$HOME/Music/John Doe/*Unknown Album*"

Performing an interactive Beets import

The interactive import process will provide a list of candidates to select from, display the changes it would make, and ask to apply those changes. If prompted, enter the number of the candidate best matching the album you wish to identify and apply those changes when subsequently prompted for your approval. If none of the candidates displayed matches the album, display additional candidates or cancel the import if no match can be found. The MusicBrainz database contains information on roughly 1.9 million artists, 3 million releases, and 26.5 million recordings. But not everything. If one or more of the albums in the music library cannot be identified by MusicBrainz or Beetcamp (or Discogs if you have enabled that plugin) then it may need to be updated manually.

Use the beet modify command to selectively add titles, artists, track names, and other metadata to items in the music library for which there is no match in any of the configured metadata sources. For example, if the above unknown album by John Doe is not matched by MusicBrainz or Beetcamp but you know the album title is “Eldorado” then the beet modify command can be used to set the album title for these items:

beet modify -a john doe unknown album 'album=Eldorado'

Provide enough of a query to uniquely identify the album or track(s) you wish to modify. To verify that the query uniquely identifies the album in question, prior to issuing the modify command, list it with

beet ls -a john doe unknown album

Performing a Beets import with a MusicBrainz ID

Another method of identifying an album or tracks is by searching the MusicBrainz website, locating the release you have, and setting the MusicBrainz ID for that release during import. There are several ways to do this. Suppose you have a digital rip of the “Snow Crash” audio book by Neil Stephenson on cassette tape. The non-interactive Beets import will probably misidentify it as one of the more recent releases of the audio book as the cassette release is lower in the candidate list.

To correct the non-interactive import or to import this audio book for the first time, first visit the musicbrainz.org and search for “Snow Crash” in Releases. Locate the cassette release of Snow Crash at https://musicbrainz.org/release/d5f4b4eb-0b58-4886-872b-d538863941dc

The long string of digits and letters in the URL is the MusicBrainz ID for this release. Copy that string. Perform an interactive Beets import of the Snow Crash audio book digital media in the music library:

beet import "$HOME/Music/Neil Stephenson/Snow Crash"

At the Beets import prompt:

Enter search, enter Id, aBort, eDit, edit Candidates, plaY?

Enter ‘I’ to “enter Id”. At the following prompt for ID enter the MusicBrainz ID copied earlier from the URL of the release followed at the next prompt with ‘A’ to Apply these changes.

Manual modification of the Beets music library should be needed infrequently unless the library contains unreleased, unpublished, or otherwise unavailable digital audio recordings.

After manually identifying and updating an album or track that MusicBrainz could not match, it is possible to update the MusicBrainz database with that new information. Beets even has a MusicBrainz Submit Plugin to ease this process but that is beyond the scope of this document.

Generally, an interactive Beets re-import of unmatched, miscategorized or misnamed items in the music library is all that is needed.

Download album cover art with the mpplus command

Once your music library has been created and the library location set, you can download album cover art with the command mpplus -D.

This step is optional as album cover art is not required by either Beets or MusicPlayerPlus. However, cover art can be displayed by many MPD clients and enhances the user experience. Pre-existing album cover art is retained and only albums with no cover art are updated.

Note, for large libraries with many albums missing cover art this process may take a long time. Monitor the download process with the command tail -f ${HOME}/.config/mpcplus/download_art.log.

The mpplus -D command downloads album cover art for each album in your music library and places the album cover art in the file cover.jpg in the album folder. The mpplus -D command downloads images of size 600x600.

If you wish to name album cover art differently or if you wish to create a different sized album cover art, then execute the sacad_r command directly rather than with mpplus -D as follows:

sacad_r <MUSIC_DIRECTORY> <SIZE> <FILENAME>

Where is the full pathname to your music library, is the desired size of the image, and is the desired filename.

For example, to download album art for a music library located in /u/audio/music with an image size of 500x500, in PNG image format, and album art filenames album_cover.png execute the command:

sacad_r /u/audio/music 500 album_cover.png

Learn more about the Smart Automatic Cover Art Downloader sacad at https://github.com/desbma/sacad

Downloading album cover art with the fetchart Beets plugin

MusicPlayerPlus includes the fetchart Beets plugin. This plugin is pre-installed, pre-configured, and enabled. The fetchart Beets plugin can be used as an alternative method of downloading album cover art and provides significant flexibility and options not available to sacad.

Our experience here at MusicPlayerPlus labs has found that mpplus -D finds and downloads a higher number of album cover art than the fetchart plugin. The fetchart plugin is useful for augmenting an initial cover art download with mpplus -D. The fetchart plugin also has many configurable options allowing the user to specify download parameters and sources. The choice of which method to use to download cover art is left to the user, both methods being supported and useful.

The fetchart plugin can also be used to selectively download album cover art by specifying a Beets query limiting the download to those albums that do not already have album cover art. This can be useful as a way to augment a previous cover art download with mpplus -D.

To identify albums without cover art, issue the command:

/usr/share/musicplayerplus/scripts/find-no-covers.sh

After identifying music library folders without album cover art using the above command, the fetchart Beets plugin, can be used to download cover art for those albums with no cover art by specifying a Beets query. For example, if the artist “John Doe” has no album cover art, issue the command:

beet fetchart 'John Doe'

Playlist creation

MusicPlayerPlus has configured the smartplaylist Beets plugin to create several MPD playlists automatically. In addition to these smart playlists which get automatically created and updated when the music library changes, MusicPlayerPlus includes the create_playlist command. This command can be used to generate new playlists based on Beets queries. Any valid Beets query can be supplied as a command-line argument and all songs matching the query will be added to a newly created playlist.

Usage: create_playlist [-n playlist-name] [-q beets-query] [-ou]
Where:
	-n 'playlist-name' specifies the playlist name
	-q 'beets-query' specifies the Beets query to use
        multiple '-q query' arguments may be provided
	-o indicates overwrite any pre-existing playlist of same name
	-u displays this usage message and exits

For example, to create a playlist named “LoveList” of all songs in the music library with “love” in the song title, artist name, or album name, issue the command create_playlist -n LoveList -q love. To create a playlist named “Dance Party” with all songs whose metadata attributes include both “danceable” and “party” values, issue the command:

create_playlist -n "Dance Party" -q mood_party:party -q danceable:danceable

[NOTE:] The flexible tags that Essentia adds based on the acoustic analysis it performs are interesting but not necessarily accurate. Tags like gender, danceable, mood_party, and others reflect an aspect of the acoustic profile but may not reflect the tag name. For example, the gender tag can have the value male or female but it does not mean that the vocalist is actually male or female, just that the acoustics in the song are more that of a male or female. Many of the songs by progressive rock band “Uriah Heep” are tagged as gender:female because the vocalist for that band, a male, has a high pitched voice. Think of these flexible tags based on acoustic analysis as representing various aspects of the acoustics rather than strictly how they are named.

Fetching lyrics

The Beets media library management system can fetch lyrics using the lyrics plugin. The MusicPlayerPlus default Beets configuration does not enable lyrics fetching using the lyrics plugin during Beets import. Albums and songs imported using Bandcamp as a metadata source will automatically fetch lyrics during import. That is, the default MusicPlayerPlus Beets import will only fetch lyrics for Bandcamp sourced items.

To view a list of music library items that do not have lyrics metadata, issue the command:

beet list -f '$artist: $album - $title' lyrics::^$

To fetch lyrics for all songs in the music library that do not already have lyrics. after the Beets import process completes issue the command:

mpplus -L

MusicPlayerPlus has chosen to provide a default Beets configuration that disables auto-fetching of lyrics during import in order to reduce the import time which can be quite lengthy. To enable auto-fetching of lyrics during import, prior to performing the music library Beets import, change the auto: no setting to auto: yes in the lyrics: section of $HOME/.config/beets/config.yaml

The lyrics plugin also has a configuration option to specify the lyrics sources to query when fetching lyrics. Some of these online lyrics services have blocked the Beets User-Agent thereby disabling those services for Beets users. Google does not block Beets lyrics queries but the Google lyrics service requires additional configuration - providing an API key.

For these reasons, MusicPlayerPlus distributes a Beets lyrics configuration that auto-fetches from Bandcamp when that is used as a metadata source and only queries the Polish lyrics service at https://www.tekstowo.pl for lyrics plugin requests. You may wish to add to or replace this service. To modify the list of lyrics services queried by the Beets lyrics plugin, change the sources: tekstowo setting in the lyrics: section of $HOME/.config/beets/config.yaml to one or more of google musixmatch genius tekstowo.

See the Beets lyrics plugin documentation for more information on configuring and using the Beets lyrics plugin.

Finding duplicate tracks

The Beets duplicates plugin can be used to identify duplicate tracks in the music library. Why and how I ended up with so many duplicate tracks in my library is a mystery but Beets helped me find and delete them, saving quite a bit of disk space and confusion. MusicPlayerPlus includes a default configuration for the duplicates plugin that uses checksums to identify duplicate files in the music library. The duplicates plugin has several methods available including comparing MusicBrainz IDs, any checksum program you wish to configure, and any set of attributes you prefer to use.

MusicPlayerPlus utilizes checksums calculated by ffmpeg as the preferred method for detecting duplicates. Edit the duplicates configuration section of $HOME/.config/beets/config.yaml to enable an alternate method.

In addition to configuring the method for identifying duplicates, you may wish to modify how the duplicates plugin decides which of the duplicates to keep and which to delete. By default, the tie-breaking procedure favors the most complete metadata attribute set. If you would like to consider the lower bitrates as duplicates, set tiebreak: items: [bitrate] in the duplicates section of the Beets config.yaml.

The recommended workflow for identifying and removing duplicate tracks from a music library consists of two passes. First, simply list any duplicates with the command beet duplicates. Review these files manually to confirm they are not wanted. The listing of duplicates in the first pass made no changes to the music library. To delete the duplicates from the Beets library and from disk, run a second pass with the command beet duplicates -d.

After deleting duplicates, some filenames may have an inappropriate suffix that was added when they were initially created. For example, there may have been duplicate files Some Artist/Some Album/Some Track.mp3 and Some Artist/Some Album/Some Track.2.mp3. The duplicates plugin may have deleted Some Artist/Some Album/Some Track.mp3 and left Some Artist/Some Album/Some Track.2.mp3 in the library. To rename these kinds of files use the beet move command. Again, recommended workflow is two passes. First perform a dry-run with the command beet move -p. This will tell you what it would do but not do anything. If that looks good then perform the file renaming with the command beet move.

Automated audio analysis and audio-based information retrieval

MusicPlayerPlus includes a pre-compiled essentia_streaming_extractor_music binary along with pre-installed and configured xtractor Beets plugin. This enables Beets to analyze audio files, extract information on a wide variety of audio parameters, and write metadata to the Beets library. This process can provide information on audio in a music library that can be used to filter and select songs by their detected audio qualities.

[NOTE on acousticbrainz] The acousticbrainz plugin performs a similar function and even uses the same Essentia technology to analyze audio data. However, acousticbrainz does not perform an analysis on all songs and, worse, the AcousticBrainz service is being retired in 2023. For these reasons, MusicPlayerPlus provides the built-in capability to perform acoustic audio analysis using Essentia and the xtractor plugin.

To perform an audio analysis and audio-based information retrieval using MusicPlayerPlus, issue the command mpplus -X 'query' where ‘query’ is a Beets media library query specifying which songs or albums or artists to match. A special query of ‘all’ indicates the entire music library.

For example, to analyze and retrieve audio information on all songs with filename or pathname that include the string “Love Party”, issue the command:

mpplus -X 'love party'

To analyze and retrieve audio information on the entire music library:

mpplus -X all

The process of analyzing and retrieving audio information can be very time consuming. For this reason, the process is run in the background and a terminal window is opened to monitor the extraction process.

Due to the long extraction times needed to perform an audio analysis, it is often desirable to identify only those songs that do not already have the desired metadata and perform the analysis selectively on those songs using an appropriate Beets query string. For example, to identify all songs in the music library that do not have a ‘mood_party’ metadata tag, issue the command:

beet list -f '$artist: $album - $title' mood_party::^$

Other songs for which the xtractor plugin may prove useful are those for which acousticbrainz mistakenly added a bpm value of 0 or those for which no bpm attribute has been set. To find songs with a bpm value of 0, issue the command:

beet list -f '$artist: $album - $title' bpm:0

To find songs for which no bpm attribute has been set, issue the command:

beet list -f '$artist: $album - $title' bpm::^$

To analyze and retrieve information for all songs in the music library with a bpm value of 0 or no bpm attribute, issue the command:

mpplus -X bpm:0 bpm::^$

If audio analysis and extraction of metadata on a large music library is too time consuming then it may be preferable to use the acousticbrainz plugin to download metadata from MusicBrainz followed by selective use of the Beets xtractor plugin. Using acousticbrainz is faster because no audio analysis is needed, the MusicBrainz database already has the metadata for songs recognized by their MusicBrainz ID. Unfortunately, the MusicBrainz metadata is not always correct and the AcousticBrainz service is being retired. But until that day (sometime in 2023) it may be preferable to enable the acousticbrainz Beets plugin and perform a two-pass metadata update of the music library:

Uncomment the acousticbrainz plugin in $HOME/.config/beets/config.yaml
Run beet acousticbrainz
Repair those songs for which acousticbrainz set bpm to 0
- beet xt bpm:0

This two-pass solution will not repair all of the mistakes that AcousticBrainz has made, only the beets-per-minute metadata, but it is a lot faster than a full analysis and extraction using beet xt.

It is preferable to perform the time consuming full analysis and extraction using beet xt or mpplus -X all as this will extract metadata with a much higher accuracy and it will continue to be supported after AcousticBrainz is no longer available.

MusicPlayerPlus custom xtractor metadata configuration

By default, the following paths are set in the xtractor section of the Beets configuration

  
xtractor:
    auto: no
    threads: 2
    ...
    essentia_extractor: /usr/bin/essentia_streaming_extractor_music
    extractor_profile:
        outputFormat: json
        ...
        highlevel:
            compute: 1
            svm_models:
                - /usr/share/mpplus-essentia/svm_models/danceability.history
                - /usr/share/mpplus-essentia/svm_models/gender.history
#               - /usr/share/mpplus-essentia/svm_models/genre_dortmund.history
#               - /usr/share/mpplus-essentia/svm_models/genre_electronic.history
#               - /usr/share/mpplus-essentia/svm_models/genre_tzanetakis.history
                - /usr/share/mpplus-essentia/svm_models/genre_rosamerica.history
                - /usr/share/mpplus-essentia/svm_models/ismir04_rhythm.history
                - /usr/share/mpplus-essentia/svm_models/mood_acoustic.history
                - /usr/share/mpplus-essentia/svm_models/mood_aggressive.history
                - /usr/share/mpplus-essentia/svm_models/mood_electronic.history
                - /usr/share/mpplus-essentia/svm_models/mood_happy.history
                - /usr/share/mpplus-essentia/svm_models/mood_party.history
                - /usr/share/mpplus-essentia/svm_models/mood_relaxed.history
                - /usr/share/mpplus-essentia/svm_models/mood_sad.history
                - /usr/share/mpplus-essentia/svm_models/moods_mirex.history
                - /usr/share/mpplus-essentia/svm_models/timbre.history
                - /usr/share/mpplus-essentia/svm_models/tonal_atonal.history
                - /usr/share/mpplus-essentia/svm_models/voice_instrumental.history

Determine how many threads your system will support with a command like:

lscpu | egrep 'Model name|Socket|Thread|NUMA|CPU\(s\)'

and set the threads configuration value accordingly.

The default MusicPlayerPlus Beets configuration includes a customized xtractor plugin configuration providing many flexible metadata tags. These include metadata tags for Beets per Minute, Key and Scale, Replay Gain, Integrated Loudness, Loudness Range, Timbre, Danceability, and Moods. The metadata names for these custom flexible tags are

bpm
danceability
beats_count
average_loudness
initial_key
key_key
key_scale
chords_key
chords_scale
replay_gain
integrated_loudness
loudness_range
timbre
danceable
genre_rosamerica
is_danceable
mood_acoustic
mood_aggressive
mood_electronic
mood_happy
mood_sad
mood_party
mood_relaxed

If the music library has been analyzed with Essentia using the xtractor plugin (e.g. mpplus -X all or mppinit metadata) then these custom tags can be used in Beets queries.

For example, to list all songs which have been tagged as being in the key of G:

beet list key_key:G key_scale:major

To list all songs in the key of G minor:

beet list key_key:G key_scale:minor

The customized entry for these metadata targets in the xtractor section of the Beets configuration $HOME/.config/beets/config.yaml is:

    low_level_targets:
      bpm:
        path: "rhythm.bpm"
        type: integer
        required: yes
      danceability:
        path: "rhythm.danceability"
        type: float
      beats_count:
        path: "rhythm.beats_count"
        type: integer
      average_loudness:
        path: "lowlevel.average_loudness"
        type: float
        required: yes
      initial_key:
        path: "tonal.initial_key"
        type: string
    high_level_targets:
      key_key:
        path: "tonal.key_edma.key"
        type: string
      key_scale:
        path: "tonal.key_edma.scale"
        type: string
      chords_key:
        path: "tonal.chords_key"
        type: string
      chords_scale:
        path: "tonal.chords_scale"
        type: string
      replay_gain:
        path: "metadata.audio_properties.replay_gain"
        type: float
      integrated_loudness:
        path: "lowlevel.loudness_ebu128.integrated"
        type: float
      loudness_range:
        path: "lowlevel.loudness_ebu128.loudness_range"
        type: float
      timbre:
        path: "highlevel.timbre.value"
        type: string
        required: yes
      danceable:
        path: "highlevel.danceability.value"
        type: string
        required: yes
      is_danceable:
        path: "highlevel.danceability.all.danceable"
        type: float
      genre_rosamerica:
        path: "highlevel.genre_rosamerica.value"
        type: string
      mood_acoustic:
        path: "highlevel.mood_acoustic.value"
        type: string
        required: yes
      mood_aggressive:
        path: "highlevel.mood_aggressive.value"
        type: string
        required: yes
      mood_electronic:
        path: "highlevel.mood_electronic.value"
        type: string
        required: yes
      mood_happy:
        path: "highlevel.mood_happy.value"
        type: string
        required: yes
      mood_sad:
        path: "highlevel.mood_sad.value"
        type: string
        required: yes
      mood_party:
        path: "highlevel.mood_party.value"
        type: string
        required: yes
      mood_relaxed:
        path: "highlevel.mood_relaxed.value"
        type: string
        required: yes

The MusicPlayerPlus Beets xtractor plugin configuration can be modified to suit the use case. However, if the xtractor has been run prior to modifying its configuration, it may be necessary to re-analyze the digital audio in order to update the metadata to reflect the revised configuration. To force re-analysis and metadata update using the xtractor plugin, use the -f flag:

beet xtractor -f -w [QUERY]

Previously analyzed xtractro output files are stored in $HOME/.config/beets/xtraction_data/ as JSON format files named either by the audio’s MusicBrainz track ID or a randomly generated UUID.

To manually examine the acoustic properties calculated using Essentia of a particular song, identify that song’s MusicBrainz track ID then look for the JSON file with that ID in the above directory. To identify a song’s MusicBrainz track ID, use a Beets query that uniquely identifies the song in a command like:

beet info -l [QUERY] | grep mb_trackid

MusicPlayerPlus acousticbrainz metadata configuration

MusicPlayerPlus supports two different methods for retrieving audio-based track metadata. The previous section described the configuration of the Beets Xtractor plugin which utilizes Essentia to analyze and retrieve audio-based information. The Xtractor process can be extremely time consuming, especially for large music libraries, as it entails a complex analysis of each audio track. For this reason, an alternative method of retrieving audio-based information is provided.

The Beets acousticbrainz plugin can be used to query the AcousticBrainz database and retrieve already analyzed audio-based information for tracks in that database. This process is much faster than the Xtractro process as the analysis has previously been performed by the AcousticBrainz service.

If the music library metadata has been updated using the acousticbrainz plugin (e.g. mpplus -x all or mppinit -a metadata) then these custom tags can be used in Beets queries.

average_loudness
bpm
chords_changes_rate
chords_key
chords_number_rate
chords_scale
danceable
gender
genre_rosamerica
initial_key
key_strength
mood_acoustic
mood_aggressive
mood_electronic
mood_happy
mood_party
mood_relaxed
mood_sad
moods_mirex
rhythm
timbre
tonal
voice_instrumental

Unfortunately, the AcousticBrainz service is deprecated, no longer updated, and will be disabled in 2023. Further, it does not provide accurate information on some metadata. Locally analyzed acoustic information using Essentia is of higher quality and reliable. The AcousticBrainz service provides lower quality information and is not reliable. However, the significant advantage to using AcousticBrainz is the overhead in time is much much lower than that with the Xtractor plugin.

Configuring the Discogs metadata source

Beets uses MusicBrainz as the default source for metadata. This is sufficient for many Beets users as MusicBrainz has a vast amount of data and requires no registration or API key to utilize. However, MusicBrainz does not have every artist, album, and song in its database and there are several Beets plugins that extend metadata search to additional sources.

MusicPlayerPlus provides a pre-configured Beets bandcamp plugin to enable Bandcamp as a metadata source. Many artists have chosen to only distribute their works on Bandcamp in order to avoid the miniscule royalty payments made to artists by streaming services. Bandcamp is also a frequent home for artists who have not been signed to a label or are otherwise not well known. The default Beets configuration that MusicPlayerPlus provides will automatically use Bandcamp as a metadata source thereby finding matches for artists, albums, and songs in the music library that were downloaded from Bandcamp.

The default MusicPlayerPlus Beets configuration, using both MusicBrainz and Bandcamp as metadata sources, should satisfy most use cases. However, some users may wish to enable additional metadata sources. One of the most complete sources of music metadata is Discogs. To enable Discogs as a metadata source, it will be necessary to register for a Discogs account, provide authorization and credentials, and configure the discogs plugin. This process is not too difficult but MusicPlayerPlus cannot perform this automatically. It must be performed by the Beets user. Follow these steps to enable Discogs as a metadata source:

Register for a free Discogs account by visiting https://www.discogs.com
No need to install python3-discogs-client as this was done during the MusicPlayerPlus initialization with mppinit
Login to Discogs to generate a user token
- Visit the Developer settings page at https://www.discogs.com/settings/developers
- Press the Generate new token button
- Copy the generated token to your clipboard (select the token and press Ctrl-c)
Add the generated token in the discogs section of the Beets configuration
- Edit $HOME/.config/beets/config.yaml
- Uncomment the discogs entry in the plugins section
- Uncomment the discogs: plugin configuration section
- Set the user_token: configuration option in the discogs: section
  - user_token: Your_Discogs_User_Token

After editing config.yaml the discogs plugin configuration should look something like the following:

  
discogs:
    source_weight: 0.0
    index_tracks: yes
    user_token: DruiHPid_I_Just_Made_This_Up_LSAPTWUCnpRJ

After completing the above steps Discogs will be used as a metadata source by Beets. The user_token setting will alleviate the need to authenticate or provide credentials when accessing Discogs with Beets.

For additional information on the Beets discogs plugin, visit https://beets.readthedocs.io/en/stable/plugins/discogs.html

MusicPlayerPlus Beets plugins

Beets includes an extensive set of plugins that can be used to enhance and extend the functionality of the media library management Beets provides. Many Beets plugins are installed and configured automatically by MusicPlayerPlus.

MusicPlayerPlus 2.0.1 and later include automated installation and configuration for the following Beets plugins:

See the MusicPlayerPlus article on Beets plugins for an introduction to the installed and configured MusicPlayerPlus Beets plugins.

Introduction to Beets Media Library Management