How to apply the incremental changes

May 25, 2026 · View on GitHub

From here on out, I'm planning to provide incremental updates, downloading, extracting etc costs everyone time and effort

Not only do you need to maybe dump tables or something to push back in the new data, but extracting, waiting maybe a few hours just inserting the data

So Feb 22 2026 dataset is perhaps the last full-dataset (for now)

You can download the incremental changes here: https://mega.nz/folder/QqQWkJKI#LVtUdSYU8hxEuD_XsylYRA

How to apply the incremental changes

for f in $(ls changes_*.sql.gz | sort); do
    echo "Restoring $f"
    gunzip -c "$f" | psql -d minimedia -h 192.168.1.2 -U postgres -q
done

Dataset incremental 2026-05-25

Service	Type	Amount
Deezer	Artist	9.4mil
Deezer	Albums	41.6mil
Deezer	Tracks	206.1mil
Tidal	Artist	8.1mil
Tidal	Albums	29.9mil
Tidal	Tracks	117.9mil
Spotify	Artist	1.2mil
Spotify	Albums	3.7mil
Spotify	Tracks	18mil
MusicBrainz	Artist	2.7mil
MusicBrainz	Albums	5.5mil
MusicBrainz	Tracks	54.3mil
SoundCloud	Artist	3.08mil
SoundCloud	Albums	2.8mil
SoundCloud	Tracks	10.3mil
Last.Fm	Artist	5k
Last.Fm	Albums	9k
Last.Fm	Tracks	19k

Note: Spotify pulling is discontinued due API limitations/payment requirement

SoundCloud pulling started at 2026-04-04, be patient till pull is complete

Last.Fm pulling started at 2026-05-25, be patient till pull is complete

Dataset of Feb 22 2026

Datasets of MusicBrainz, Tidal, Spotify, Deezer

You can use these tables for MiniMedia's database, it will save you huge amounts of time, instead calling the API's yourself

Tidal, Spotify, Deezer datasets were obtained through their API, took months of calling their API's 24/7

MusicBrainz came mostly through their own published dataset + API

Note for Deezer dataset: The Preview Url (to listen to the first x seconds of a song) and TrackToken (for playback) fields will be empty, it took too much space to store all of this for me

Packed: CSV-Format 10.2GB, SQL 21.7GB

Unpacked CSV-Format 178GB, SQL 149GB

Loving the work I do? buy me a coffee https://buymeacoffee.com/musicmovearr

MusicBrainz

You can officially download it here: https://metabrainz.org/datasets/postgres-dumps#musicbrainz

But the official dataset is huge, a lot larger then I'm sharing, this is because I saved it more efficiently

Type	Name
long	ArtistId
string	ArtistName
int	ArtistNbAlbum
int	ArtistNbFan
bool	ArtistRadio
string	ArtistType
string	ArtistHref
string	ArtistImageHref
long	AlbumId
long	AlbumArtistId
string	AlbumName
string	AlbumMd5Image
int	AlbumGenreId
long	AlbumFans
string	AlbumReleaseDate
string	AlbumRecordType
bool	AlbumExplicitLyrics
int	AlbumExplicitContentLyrics
int	AlbumExplicitContentCover
string	AlbumType
string	AlbumUPC
string	Label
long	AlbumNbTracks
TimeSpan	AlbumDuration
bool	AlbumAvailable
string	AlbumHref
string	AlbumGenreName
string	AlbumGenrePicture
long	TrackId
bool	TrackReadable
string	TrackTitle
string	TrackTitleShort
string	TrackTitleVersion
string	TrackISRC
TimeSpan	TrackDuration
int	TrackPosition
int	TrackDiscNumber
long	TrackRank
string	TrackReleaseDate
bool	TrackExplicitLyrics
int	TrackExplicitContentLyrics
int	TrackExplicitContentCover
double	TrackBPM
double	TrackGain
string	TrackMd5Image
long	TrackArtistId
long	TrackAlbumId
string	TrackType
string	TrackHref