- Pitchfork's 50 Best Albums of 2016 (source)
- Pitchfork's Honorable Mentions of 2015 (source)
- Pitchfork's 50 Best Ambient Albums of All Time (source)
- Rate Your Music's Top Albums of 2016 (source)
- Rate Your Music's Top Dream Pop Albums (source)
- Rate Your Music's Top Shoegaze Albums (source)
- Rolling Stone's 500 Greatest Songs of All Time (source)
Spotify collaborates with the Echo Nest Project and provides algorithmic analysis of all tracks. These "audio features" can be used to filter existing playlists. For example, to create a melodic playlist out of "Pitchfork's Best New Tracks":
#order by instrumentalness
https://open.spotify.com/user/pitchforkmedia/playlist/5ItokEl1f0bbHeXWFiisrm
To create a hip-hop playlist:
#order by instrumentalness
#reverse
https://open.spotify.com/user/pitchforkmedia/playlist/5ItokEl1f0bbHeXWFiisrm
To create a workout playlist:
#order by energy
https://open.spotify.com/user/pitchforkmedia/playlist/5ItokEl1f0bbHeXWFiisrm
To create a dancing playlist:
#order by danceability
https://open.spotify.com/user/pitchforkmedia/playlist/5ItokEl1f0bbHeXWFiisrm
To create a playlist of a user's recently played tracks:
http://www.last.fm/user/username/library
To create a playlist of tracks a user has listened to on a particular day:
http://www.last.fm/user/username/library?from=2012-07-21&to=2013-07-21
To create a playlist of a user's favorite tracks for a given year:
http://www.last.fm/user/username/library/tracks?from=2012-01-01&to=2013-01-01
To create a playlist of a user's favorite tracks of all time:
http://www.last.fm/user/username/library/tracks
To create a playlist of a user's loved tracks:
http://www.last.fm/user/username/loved
To create a playlist of a user's favorite albums:
http://www.last.fm/user/username/library/albums
To create a playlist of a user's favorite artists:
http://www.last.fm/user/username/library/artists
To create a playlist of an artist's top tracks:
http://www.last.fm/music/Artist+Name/+tracks
To create a playlist of artists similar to an artist:
http://www.last.fm/music/Artist+Name/+similar
Last.fm's webpages are a goldmine of recommendations. For example, if one has a high compatibility with another user, that user's "top tracks" often makes for an interesting playlist.
Similarly, one's own "top tracks" are also a good source of favorites, especially in combination with a time criteria. The "top tracks" chart for an earlier year is a good indication of one's listening habits in that time period, and a rich source of "lost gems" which one may have forgotten about in the meantime.
Last.fm also provides information about an artist's top tracks and similar artists. Such functionality is built into Spotify as well, but Last.fm's recommendations are a good supplement.
#alternate by artist
http://www.last.fm/music/Artist+Name/+similar
#order by lastfm username
#artist Beach House
#order by popularity
#group by entry
#top ...
#top ...
#top ...
...
#order by lastfm
#group by entry
#top ...
#top ...
#top ...
...
#order by lastfm
#group by entry
#album ...
#album ...
#album ...
...
#alternate by album
#album ...
#album ...
#album ...
...
#order by popularity
#alternate by artist
#similar ...
#similar ...
#similar ...
...
#order by lastfm
#alternate by artist
#similar ...
#similar ...
#similar ...
...
Converting multiple playlists at once can easily be done in the Bash shell:
for f in *.txt; do spotgen "$f" "${f%.txt}.spotify.txt"; done
This converts playlist1.txt, playlist2.txt, playlist3.txt to playlist1.spotify.txt, playlist2.spotify.txt, playlist3.spotify.txt.
Furthermore, if the shell supports globbing, then one can recursively convert all playlists in a directory with the **/*.txt pattern:
for f in **/*.txt; do spotgen "$f" "${f%.txt}.spotify.txt"; done
One can easily import extended M3U playlists containing EXTM3U metadata. If the playlist does not contain metadata (or if the files are badly tagged), however, then one can attempt to infer the title and artist from the file paths instead. To do this, open the playlist in a text editor and use regular expressions to transform it. For example, the regular expression
%s/\(^[^\/]+\).*\/[-0-9]*[-. ]*\(.*\)\..*/\2 - \1/g
transforms the playlist
Beach House/Teen Dream/04 Walk in the Park.mp3
Beach House/Bloom/10 Irene.mp3
Beach House/Bloom/04 Other People.mp3
Beach House/Bloom/06 Troublemaker.mp3
Beach House/Bloom/08 Wishes.mp3
to the compatible text file
Walk in the Park - Beach House
Irene - Beach House
Other People - Beach House
Troublemaker - Beach House
Wishes - Beach House
Since file name standards vary, the transformation requires some personal judgment; that is why it is not built into the script. (See this blog post for more examples.)
Note that other playlist formats, like ASX, PLS, WPL and XSPF, are difficult to work with directly. Therefore, it is recommended to save to M3U format and proceed as outlined above.
Single tracks should be on the form Artist - Track. However, Spotify isn't terribly strict about this; Track - Artist also works well, as does a Spotify search. One can even add field filters.