HOW IT WORKS
How your type is worked out
Listen to 24 songs and answer 12 scenarios to shape your music spirit. Your choices determine the result, not a random draw.
01 / Three axes
How do you listen? Three questions.
12 scenarios, 4 per question. The side you pick more often wins. The three answers are the first three letters of your code.
DISCOVERY · 新歌探索
Example: 「You have time to listen. You choose the next song.」A Replay a familiar favorite → F；B Try an artist I do not know → D。
IMMERSION · 聆聽沉浸
Example: 「A song catches your attention. Your other task can wait.」A Let it accompany another activity → R；B Put things down and listen closely → I。
SOCIAL · 音樂交流
Example: 「On a road trip, it is your turn to play music.」A Play the sequence I arranged → P；B Let everyone add a song → S。
Tied 2 to 2? Your liked songs decide: more new songs lean D, more detail-heavy songs lean I, more shareable songs lean S.
02 / 16 types
Three listening dimensions form eight combinations. Focused or broad genre evidence makes sixteen portraits. The animals help you remember them; they do not diagnose personality or rank taste.
MEET THE MUSIC SPIRITS
Which spirit sounds like you?
Sixteen listening portraits, shaped by your scenario choices and the genres you like.
Motion follows your reduced-motion preference.
Familiar favorites fit right into your day.
Familiar sounds · Everyday company · Personal picks · Focused genres
Familiar sounds from different genres keep you company.
Familiar sounds · Everyday company · Personal picks · Broad genres
A familiar chorus is your invitation to listen together.
Familiar sounds · Everyday company · Shared listening · Focused genres
So many familiar favorites—and someone to share them with.
Familiar sounds · Everyday company · Shared listening · Broad genres
A favorite song still has another detail waiting for you.
Familiar sounds · Focused listening · Personal picks · Focused genres
Across familiar genres, there is always a detail worth a pause.
Familiar sounds · Focused listening · Personal picks · Broad genres
Wait—this is the part we need to hear together.
Familiar sounds · Focused listening · Shared listening · Focused genres
Every familiar genre has a passage worth sharing.
Familiar sounds · Focused listening · Shared listening · Broad genres
A new sound in your favorite corner gives the day a fresh route.
New sounds · Everyday company · Personal picks · Focused genres
Headphones on. Another sound world to drift through.
New sounds · Everyday company · Personal picks · Broad genres
Found a new sound I like. Sending you the link.
New sounds · Everyday company · Shared listening · Focused genres
This sounds new. Want to give it a listen together?
New sounds · Everyday company · Shared listening · Broad genres
One unfamiliar detail is enough to play that passage again.
New sounds · Focused listening · Personal picks · Focused genres
Dive across genres and listen to every layer.
New sounds · Focused listening · Personal picks · Broad genres
Found a sound worth listening closely to. Come explore it with me.
New sounds · Focused listening · Shared listening · Focused genres
An unfamiliar sound can start a whole new conversation.
New sounds · Focused listening · Shared listening · Broad genres
These are examples of possible results. Balanced or incomplete evidence keeps its open listening portrait; no type is forced.
Unresolved balanced evidence stays a mix of listening styles; incomplete or zero-favorite evidence stays exploring your taste. No animal type is forced, and your songs and coordinates remain available.
03 / Broad and focused
Focused or broad: how your liked genres are distributed.
Sort your liked songs into genre piles. It is not how many piles, but how even they are. One pile scores 1; four equal piles score 4. Four or more is B, less is F.
The formula, if you want it: 1 ÷ Σ(each pile’s share²).
5 of 8 are R&B; that scores about 2. Focused.
8 songs across 6 genres score about 5. Everything goes.
04 / How songs are picked
24 songs for you, drawn from 218.
12 genres, 1680 to 2026, 27 in Mandarin.
Two songs per genre, so nobody misses jazz.
12 are the same for everyone so friends can compare; 12 change every time so it stays fresh.
Classics and new releases. People who like more new songs are usually D.
The tags on every song
Why not draw at random
Fully random, you and a friend might share no songs at all. So one song per genre, 12 in all, is the same for everyone, and the other 12 are redrawn every time. Two per genre keeps every map cell from going blank.
What this sample cannot tell you
Two songs per genre is a sketch, not a measurement: "9 genres liked" means you liked at least one of the two songs in nine cells, and you may simply dislike those two classical pieces. The songs are chosen by us, not sampled from streaming data; 27 are in Mandarin so far, and more are coming. Like, Neutral, Dislike and Unable to judge are recorded separately. Neutral is a midpoint; Unable to judge is excluded from similarity.
05 / How genres are grouped
How are genres grouped? Following MusicBrainz.
Genres follow MusicBrainz genres and community (folksonomy) tags, which we simplify into these 12 families. Each song's small tags, the subgenres, and their aliases match its vocabulary and are grouped into those 12 cells. Three layers, one job each.
Genres
12 cells that never change. The map, share cards, and bars all use this layer.
Subgenres
Each belongs to one cell. They answer “why is this song here?”
- new jack swing → Funk → R&B
- amapiano → Afrobeats → Global
- kayōkyoku → J-pop → Global
Small tags
Two or three per song, written by hand. They do not score; they help you know the song.
06 / A note
This test measures your relationship with music.
After 24 songs you will know which genres you like, how you usually listen, and which friend hears most like you. That is the fun of listening, not a personality analysis.
The questions draw on music psychology research, but this is a sketch of listening preferences, not a formal psychological scale.
Another day, another set of songs, and the result may differ. Taste moves. A genre you did not pick is not one you dislike; it just did not land this time.
References: Gold-MSI · BMRQ · Schäfer 2013 · Rentfrow 2011 · MusicBrainz