Our Taxonomy
Cyanite’s taxonomy is constantly being refined and improved. Our internal music and engineering team leads our research, development, and training to continuously improve the AI’s capabilities.
Overview
BPM
The BPM classifier provides you the BPM of the track.
“value”: 30 – 285
“confidence“: 0 – 1
“bpmRangeAdjusted”: 60 – 180
Key
The Key classifier provides you with the predicted key.
“aMinor”,
“eMinor”,
“bMinor”,
“fsMinor”,
“csMinor”,
“gsMinor”,
“dsMinor”,
“bbMinor”,
“fMinor”,
“cMinor”,
“gMinor”,
“dMinor”,
“cMajor”,
“gMajor”,
“dMajor”,
“aMajor”,
“eMajor”,
“bMajor”,
“fsMajor”,
“dbMajor”,
“abMajor”,
“ebMajor”,
“bbMajor”,
“fMajor”
“confidence“: 0 – 1
Time Signature
The Time Signature classifier provides you with timeSignature.
“3/4”,
“4/4”
Mood
The mood multi-label classifier provides mood, moodTags, moodAdvanced, moodAdvancedTags, moodMaxTimes.
(The mood can be retrieved both averaged over the whole track and segment-wise over time with 15s temporal resolution)
“aggressive”: 0 – 1,
“calm”: 0 – 1,
“chilled”: 0 – 1,
“dark”: 0 – 1,
“energetic”: 0 – 1,
“epic”: 0 – 1,
“happy”: 0 – 1,
“romantic”: 0 – 1,
“sad”: 0 – 1,
“scary”: 0 – 1,
“sexy”: 0 – 1,
“ethereal”: 0 – 1,
“uplifting”: 0 – 1
“aggressive”,
“calm”,
“chilled”,
“dark”,
“energetic”,
“epic”,
“happy”,
“romantic”,
“sad”,
“scary”,
“sexy”,
“ethereal”,
“uplifting”
(every possible mood as potential moodTags)
“mood”: “agressive”,
“start”: 15,
“end”: 45
“mood”: “calm”,
“start”: 15,
“end”: 45
“mood”: “chilled”,
“start”: 15,
“end”: 45
…
(every possible mood as moodMaxTimes)
“anxious”: 0 – 1,
“barren”: 0 – 1,
“cold”: 0 – 1,
“creepy”: 0 – 1,
“dark”: 0 – 1,
“disturbing”: 0 – 1,
“eerie”: 0 – 1,
“evil”: 0 – 1,
“fearful”: 0 – 1,
“mysterious”: 0 – 1,
“nervous”: 0 – 1,
“restless”: 0 – 1,
“spooky”: 0 – 1,
“strange”: 0 – 1,
“supernatural”: 0 – 1,
“suspenseful”: 0 – 1,
“tense”: 0 – 1,
“weird”: 0 – 1,
“aggressive”: 0 – 1,
“agitated”: 0 – 1,
“angry”: 0 – 1,
“dangerous”: 0 – 1,
“fiery”: 0 – 1,
“intense”: 0 – 1,
“passionate”: 0 – 1,
“ponderous”: 0 – 1,
“violent”: 0 – 1,
“comedic”: 0 – 1,
“eccentric”: 0 – 1,
“funny”: 0 – 1,
“mischievous”: 0 – 1,
“quirky”: 0 – 1,
“whimsical”: 0 – 1,
“boisterous”: 0 – 1,
“boingy”: 0 – 1,
“bright”: 0 – 1,
“celebratory”: 0 – 1,
“cheerful”: 0 – 1,
“excited”: 0 – 1,
“feelGood”: 0 – 1,
“fun”: 0 – 1,
“happy”: 0 – 1,
“joyous”: 0 – 1,
“lighthearted”: 0 – 1,
“perky”: 0 – 1,
“playful”: 0 – 1,
“rollicking”: 0 – 1,
“upbeat”: 0 – 1,
“calm”: 0 – 1,
“contented”: 0 – 1,
“dreamy”: 0 – 1,
“introspective”: 0 – 1,
“laidBack”: 0 – 1,
“leisurely”: 0 – 1,
“lyrical”: 0 – 1,
“peaceful”: 0 – 1,
“quiet”: 0 – 1,
“relaxed”: 0 – 1,
“serene”: 0 – 1,
“soothing”: 0 – 1,
“spiritual”: 0 – 1,
“tranquil”: 0 – 1,
“bittersweet”: 0 – 1,
“blue”: 0 – 1,
“depressing”: 0 – 1,
“gloomy”: 0 – 1,
“heavy”: 0 – 1,
“lonely”: 0 – 1,
“melancholic”: 0 – 1,
“mournful”: 0 – 1,
“poignant”: 0 – 1,
“sad”: 0 – 1,
“frightening”: 0 – 1,
“horror”: 0 – 1,
“menacing”: 0 – 1,
“nightmarish”: 0 – 1,
“ominous”: 0 – 1,
“panicStricken”: 0 – 1,
“scary”: 0 – 1,
“concerned”: 0 – 1,
“determined”: 0 – 1,
“dignified”: 0 – 1,
“emotional”: 0 – 1,
“noble”: 0 – 1,
“serious”: 0 – 1,
“solemn”: 0 – 1,
“thoughtful”: 0 – 1,
“cool”: 0 – 1,
“seductive”: 0 – 1,
“sexy”: 0 – 1,
“adventurous”: 0 – 1,
“confident”: 0 – 1,
“courageous”: 0 – 1,
“resolute”: 0 – 1,
“energetic”: 0 – 1,
“epic”: 0 – 1,
“exciting”: 0 – 1,
“exhilarating”: 0 – 1,
“heroic”: 0 – 1,
“majestic”: 0 – 1,
“powerful”: 0 – 1,
“prestigious”: 0 – 1,
“relentless”: 0 – 1,
“strong”: 0 – 1,
“triumphant”: 0 – 1,
“victorious”: 0 – 1,
“delicate”: 0 – 1,
“graceful”: 0 – 1,
“hopeful”: 0 – 1,
“innocent”: 0 – 1,
“intimate”: 0 – 1,
“kind”: 0 – 1,
“light”: 0 – 1,
“loving”: 0 – 1,
“nostalgic”: 0 – 1,
“reflective”: 0 – 1,
“romantic”: 0 – 1,
“sentimental”: 0 – 1,
“soft”: 0 – 1,
“sweet”: 0 – 1,
“tender”: 0 – 1,
“warm”: 0 – 1,
“anthemic”: 0 – 1,
“aweInspiring”: 0 – 1,
“euphoric”: 0 – 1,
“inspirational”: 0 – 1,
“motivational”: 0 – 1,
“optimistic”: 0 – 1,
“positive”: 0 – 1,
“proud”: 0 – 1,
“soaring”: 0 – 1,
“uplifting”: 0 – 1
“anxious”,
“barren”,
“cold”,
“creepy”,
“dark”,
“disturbing”,
“eerie”,
“evil”,
“fearful”,
“mysterious”,
“nervous”,
“restless”,
…
(every possible moodAdvanced as potential moodAdvancedTags)
Emotion
The Emotion classifier provides emotionalProfile and emotionalDynamics.
(The Emotion can be retrieved averaged over the whole track)
“variable”,
“negative”,
“balanced”,
“positive”
“low”,
“medium”,
“high”
Genre
The genre multi-label classifier provides genre, genreTags.
(The genre can be retrieved both averaged over the whole track and segment-wise over time with 15s temporal resolution)
“ambient”: 0 – 1,
“blues”: 0 – 1,
“classical”: 0 – 1,
“electronicDance”: 0 – 1,
“folkCountry”: 0 – 1,
“funkSoul”: 0 – 1,
“jazz”: 0 – 1,
“latin”: 0 – 1,
“metal”: 0 – 1,
“pop”: 0 – 1,
“rapHipHop”: 0 – 1,
“reggae”: 0 – 1,
“rnb”: 0 – 1,
“rock”: 0 – 1,
“singerSongwriter”: 0 – 1
“ambient”,
“blues”,
“classical”,
“electronicDance”,
“folkCountry”,
“funkSoul”,
“jazz”,
“latin”,
“metal”,
“pop”,
“rapHipHop”,
“reggae”,
“rnb”,
“rock”,
“singerSongwriter”
(every possible genre as potential genreTags)
Sub-Genre
The sub-genre multi-label classifier provides subgenre, subgenreTags.
(The subgenre can be retrieved both averaged over the whole track and segment-wise over time with 15s temporal resolution)
“bluesRock”: null – 1,
“folkRock”: null – 1,
“hardRock”: null – 1,
“indieAlternative”: null – 1,
“psychedelicProgressiveRock”: null – 1,
“punk”: null – 1,
“rockAndRoll”: null – 1,
“popSoftRock”: null – 1,
“abstractIDMLeftfield”: null – 1,
“breakbeatDnB”: null – 1,
“deepHouse”: null – 1,
“electro”: null – 1,
“house”: null – 1,
“minimal”: null – 1,
“synthPop”: null – 1,
“techHouse”: null – 1,
“techno”: null – 1,
“trance”: null – 1,
“contemporaryRnB”: null – 1,
“gangsta”: null – 1,
“jazzyHipHop”: null – 1,
“popRap”: null – 1,
“trap”: null – 1,
“blackMetal”: null – 1,
“deathMetal”: null – 1,
“doomMetal”: null – 1,
“heavyMetal”: null – 1,
“metalcore”: null – 1,
“nuMetal”: null – 1,
“disco”: null – 1,
“funk”: null – 1,
“gospel”: null – 1,
“neoSoul”: null – 1,
“soul”: null – 1,
“bigBandSwing”: null – 1,
“bebop”: null – 1,
“contemporaryJazz”: null – 1,
“easyListening”: null – 1,
“fusion”: null – 1,
“latinJazz”: null – 1,
“smoothJazz”: null – 1,
“country”: null – 1,
“folk”: null – 1
“bluesRock”,
“folkRock”,
“hardRock”,
“indieAlternative”,
…
(every possible subGenre as potential subGenreTags)
Classical Epoch
The Classical Epoch multi-label classifier provides classicalEpoch, classicalEpochTags.
(The Classical Epoch can be retrieved both averaged over the whole track and segment-wise over time with 15s temporal resolution – it is triggered once the Classical main genre is tagged)
“middleAge”: 0 – 1,
“renaissance”: 0 – 1,
“baroque”: 0 – 1,
“classical”: 0 – 1,
“romantic”: 0 – 1,
“contemporary”: 0 – 1
“middleAge”,
“renaissance”,
“baroque”,
“classical”,
“romantic”,
“contemporary”
(every possible classicalEpoch as potential classicalEpochTags)
Voice
The voice classifiers provides voice, voiceTags, voicePresenceProfile, and predominantVoiceGender.
(The voice classifier results can be retrieved both averaged over the whole track and segment-wise over time with 15s temporal resolution)
“female”: 0 – 1,
“male”: 0 – 1,
“instrumental”: 0 – 1
“female”,
“male”,
“instrumental”
“none”,
“low”,
“medium”,
“high”
“none”,
“female”,
“male”
Instruments
The instruments classifiers provides instrumentsPresence, instrumentTags, and instrumentsSegments.
(instrumentsPresence and instrumentTags results can be retrieved averaged over the whole track. instrumentsSegments can be retrieved segment-wise over time with 15s temporal resolution)
available instruments:
“percussion”,
“synth”,
“piano”,
“acousticGuitar”,
“electricGuitar”,
“strings”,
“bass“,
“bassGuitar”,
“brassWoodwinds”
possible values:
“absent”,
“partially”,
“frequently”,
“throughout”
“percussion”,
“synth”,
“piano”,
“acousticGuitar”,
“electricGuitar”,
“strings”,
“bass“,
“bassGuitar”,
“brassWoodwinds”
(every possible instrument as potential instrumentTags)
“percussion”: 0 – 1,
“synth”: 0 – 1,
“piano”: 0 – 1,
“acousticGuitar”: 0 – 1,
“electricGuitar”: 0 – 1,
“strings”: 0 – 1,
“bass“: 0 – 1,
“bassGuitar”: 0 – 1,
“brassWoodwinds”: 0 – 1
Transformer Caption
The transformer caption classifier provides transformerCaption.
(It contains a string of max. 30 words describing the track in one or few sentences)
“transformerCaption”
Musical Era
The musical era classifier describes the era the audio was likely produced in, or which the sound of production suggests.
It includes musicalEraTag.
“late 1940s / early 1950s”,
“early 1950s / mid 1950s”,
“mid 1950s / late 1950s”,
“late 1950s / early 1960s”,
“early 1960s / mid 1960s”,
“mid 1960s / late 1960s”,
“late 1960s / early 1970s”,
“early 1970s / mid 1970s”,
“mid 1970s / late 1970s”,
“late 1970s / early 1980s”,
“early 1980s / mid 1980s”,
“mid 1980s / late 1980s”,
“late 1980s / early 1990s”,
“early 1990s / mid 1990s”,
“mid 1990s / late 1990s”,
“late 1990s / early 2000s”,
“early 2000s / mid 2000s”,
“mid 2000s / late 2000s”,
“late 2000s / contemporary”,
“contemporary”
“early 1950s / mid 1950s”,
“mid 1950s / late 1950s”,
“late 1950s / early 1960s”,
“early 1960s / mid 1960s”,
“mid 1960s / late 1960s”,
“late 1960s / early 1970s”,
“early 1970s / mid 1970s”,
“mid 1970s / late 1970s”,
“late 1970s / early 1980s”,
“early 1980s / mid 1980s”,
“mid 1980s / late 1980s”,
“late 1980s / early 1990s”,
“early 1990s / mid 1990s”,
“mid 1990s / late 1990s”,
“late 1990s / early 2000s”,
“early 2000s / mid 2000s”,
“mid 2000s / late 2000s”,
“late 2000s / contemporary”,
“contemporary”
Movement
The Movement classifier provides movement, movementTags.
(The Movement classifier results can be retrieved both averaged over the whole track and segment-wise over time with 15s temporal resolution)
“bouncy”: 0 – 1,
“driving”: 0 – 1,
“flowing”: 0 – 1,
“groovy”: 0 – 1,
“nonrhythmic”: 0 – 1,
“pulsing”: 0 – 1,
“robotic”: 0 – 1,
“running”: 0 – 1,
“steady”: 0 – 1,
“stomping”: 0 – 1
“bouncy”,
“driving”,
“flowing”,
“groovy”,
“nonrhythmic”,
“pulsing”,
“robotic”,
“running”,
“steady”,
“stomping”
(every possible movement as potential movementTags)
Character
The Character classifier provides character, characterTags.
(The Movement classifier results can be retrieved both averaged over the whole track and segment-wise over time with 15s temporal resolution)
“bold”: 0 – 1,
“cool”: 0 – 1,
“epic”: 0 – 1,
“ethereal”: 0 – 1,
“heroic”: 0 – 1,
“luxurious”: 0 – 1,
“magical”: 0 – 1,
“mysterious”: 0 – 1,
“playful”: 0 – 1,
“powerful”: 0 – 1,
“retro”: 0 – 1,
“sophisticated”: 0 – 1,
“sparkling”: 0 – 1,
“sparse”: 0 – 1,
“unpolished”: 0 – 1,
“warm”: 0 – 1
“bold”,
“cool”,
“epic”,
“ethereal”,
“heroic”,
“luxurious”,
“magical”,
“mysterious”,
“playful”,
“powerful”,
“retro”,
“sophisticated”,
“sparkling”,
“sparse”,
“unpolished”,
“warm”
(every possible character as potential characterTags)
Valence / Arousal
The Valence / Arousal classifier provides valence, arousal.
(The Valence / Arousal classifier results can be retrieved both averaged over the whole track and segment-wise over time with 15s temporal resolution)
“valence”: 0 – 1
“arousal”: 0 – 1
Energy
The Energy classifier provides energyLevel and energyDynamics.
(The Energy can be retrieved averaged over the whole track)
“energyLevel”
possible values:
“variable”,
“medium”,
“high”,
“low”
“energyDynamics”
possible values:
“low”,
“medium”,
“high”
Product
Resources
Follow Us
About
