8 Music & Lyrics

Artists create and perform as authors performers music which often has lyrics.

More about ► artists and their performer groups, ensembles, orchestras, collectives, ► bands.

8.1 Music contents

8.1.1 Voice

is_instrumental: If the recording appears to be instrumental accoring to Essentia. speachiness: How much speach is heard in the recording according to the Spotify Echo Nest engine. voice_gender: F for female, M for male, I for instrumental. The voice_gender_probability shows how certain the Essentia engine was about the gender of the singer.

The voice_gender property is different from the artist_gender property. The voice_gender is applicable for bands, ensembles because it tries to guess the performer’s gender and not that of the composer.

8.1.2 Mood

mood_acoustics: The probablity and the value of the Essentia perceived accoustiness measurement. mood_aggressive: The probablity and the value of the Essentia aggressiveness measurement. mood_electronic:The probablity and the value of the Essentia perceived ‘electronic’ measurement. mood_happy: The probablity and the value of the Essentia happiness measurement. mood_sad: The probablity and the value of the Essentia sadness measurement. mood_relaxed: The probablity and the value of the Essentia relaxed mood measurement. mood_party: The probablity and the value of the Essentia party mood measurement. timbre: The probablity and the value of the Essentia ‘timbre’ value, which has a robust connection with the impression of “brightness” of a sound.

8.1.3 Genre

Genres are taxonomies to organize music. When no consistent method is used to attach a genre label to a music, we talk about folksonomies. Genres has a consistent set of criteria.

The Essentia measures as based on how close the song is to the sound of reference genre datasets. While they do not map the song to a precise taxonomy, their values are very useful to find similar songs to a recording, because they attach a value to 8 dimensions of a recording, such as how much it is similar to blues songs, alternative songs, etc.

genre_dortmund: The percieved genre by Essentia on the basis of the TU Dortmund Music Audio Benchmark Data Set, values alternative, blues, electronic, folkcountry, funksoulrnb, jazz, raphiphop, rock. genre_dortmund_blues: how much the song resembles the reference blues songs in the TU Dortmund Music Audio Benchmark Data Set.

The Spotify genres follow the taxonomy of the Spotify recommender sytem. A song may have zero, one, two or more genres attached to it, which will be used to find similar songs or suitable audiences for a song on the Spotify platform.

The Bandcamp folksonomies are self-attached genre tags by the artists or their labels. They do not follow a consistent taxonomy.

8.1.4 Rhythm

bpm: Beats per minute. beats_count: Total beats counted by Essentia in the recording. beat_onset: The onset beat of the recording according to Essentia.

8.1.5 Danceability

danceability_spotify: The level of danceability according to the Spotify’s Echo Nest engine. danceability_essentia: The level of danceability according to the Essentia model.

8.1.6 Tonality

is_atonal: If the song is atonal (1) or tonal (0). key_essentia: The key of the song according to the Essentia engine. key_spotify: The key of the song according to the Spotify Echo Nest engine. tuning_frequency: The apparant tuning frequency for the normal A sound on the recording. chords_changes_rate