·9 min read
Finding Music Through Emotion
My listening is primarily drawn from Tamil film music, Hindi film music, Carnatic music, Western classical music, rock and pop, jazz, New Age, and independent artists. I move between these different musical worlds, each offering something I enjoy.
There are many ways to find something in this collection. I can look for a composer, a singer, an album, or a genre. These are useful doors. But the door I most often want is missing.
Sometimes I finish listening to a song and think: I would like to hear something else that makes me feel this way.
I do not necessarily want the same singer. I may not want the same instruments, tempo, or musical tradition. I want to follow the connection I have just felt.
What would it mean to search for that?
A map of the idea
I think of this as a possible Multidimensional Music Affinity Engine. The name describes an idea I want to explore, rather than an established model. This map brings the musical questions and the listener's experience together.
Explore whole performances, individual passages, and the relationships between musical parts.
Tenderness, tension, joy, longing, release
Timbre, phrasing, ornamentation, pauses
Layering, orchestration, musical responses
Raga, melodic movement, tension and resolution
Warmth, space, clarity, dynamics
Groove, pulse, timing, intensity
Memories · Familiarity · Present mood · Comparisons between recordings · Moments I return to
Personal associations come from what I choose to share; they cannot be read from audio alone.Connections across genres, languages, performances, and passages—shaped by my listening feedback.
“Find music that might create a related emotional experience for me.”
A voice, an answering violin, and a feeling of tenderness.
Imagined examples to explain the idea, not findings about my actual preferences.What the music expresses, and what I feel
My first difficulty is understanding what emotion means in this search. I want to explore the emotion expressed by a performance, the feelings it evokes in me, and the personal connections that give those feelings meaning.
A recording may sound melancholy, yet leave me comforted. A lively song may carry a memory that makes me sad. Something quiet may hold my attention with an intensity that the word calm does not describe.
If a system calls a song sad, it has not yet told me whether I will love it, avoid it, or return to it when I need company.
That is the distinction I want to explore. I am interested in my relationship with a recording, including the feelings I cannot easily name.
The space between the instruments
Consider a bass line. Heard alone, it may do very little for me. Let it meet the drums, and a groove begins to take shape. Add a voice, and the same pattern becomes part of a different experience. Then a violin answers the singer, or the accompaniment briefly falls away, and something catches me.
Which part moved me?
I might say the violin, because its entrance was the moment I noticed. But would that entrance have mattered without the phrase before it? Would the phrase have carried the same weight without the rhythm underneath?
Knowing which instruments are present would help describe the recording. I still want to understand what they are doing together.
Perhaps the thing I respond to lives in the relationship: a voice leaning against a rhythm, an instrumental answer arriving after a pause, a melody finding its way back after a small departure. These are possibilities to investigate through listening, rather than explanations I can assume in advance.
Two singers, one raga
Another example keeps returning to me. Imagine two singers performing the same raga. I may admire both performances, yet feel deeply attached to only one.
Naming the raga does not settle the question. I would want to listen more closely to the phrasing, the movement between swaras, the gamakas, the pauses, and the way the singer and accompaniment respond to one another. Perhaps a particular turn feels unexpected to me. Perhaps the voice itself holds my attention.
I may hear these differences without being able to explain which ones matter most.
The same question follows me into other music. Two interpretations of a jazz standard or a classical work can occupy very different places in my listening life. The composition gives me one way to connect them. My attachment to a particular performance asks for a closer look.
Even the recording belongs in that question. I may be drawn to its warmth, its sense of space, or the way a voice sits among the instruments. Separating composition, performance, arrangement, and recording can help me think, but my experience arrives as a whole.
A few seconds inside a song
Sometimes my reason for returning to a recording is only a few seconds long.
A flute enters after a vocal phrase. A singer holds a note a little longer than I expect. An orchestral interlude opens out. There is a brief silence, followed by exactly the sound I was waiting for.
I listen to the whole piece, but I look forward to that passage.
This makes me cautious about describing a recording with a single mood. The beginning may be restrained, the middle unsettled, and the ending full of release. The passage I love may depend on having travelled through everything before it.
If I ask for another moment like it, how much of that journey needs to travel with the comparison? A few seconds? A phrase? An entire movement? I do not yet know.
Connections across musical borders
A Tamil film song and a jazz performance may sound very different, yet both may move me in a similar way.
In one, it might be a singer's voice followed by a violin. In the other, it might be a saxophone phrase answered by the piano. The instruments and musical styles are different, but I may feel the same tenderness or longing while listening.
I would like to discover more of these connections. If I love a passage in a film song, could it lead me to something in Carnatic music, Western classical, or jazz that creates a similar emotional experience?
The connection would be personal. Someone else might hear the same two recordings and find no link between them. I would want the system to learn from what I hear and enjoy, and let me decide whether its suggestions make sense.
Could AI learn what moves me?
There are several questions here, and I do not want to confuse them.
Describing the emotional character of a recording is one task. Finding something I personally connect with is another. Helping me choose music for my present mood adds another layer. Explaining why a passage moves me is a more demanding question still.
My starting point would be modest. Could a system learn from comparisons I can actually make?
I may be unable to explain why I love a performance, yet still say that one recording feels closer to it than another. I can point to a passage I keep replaying. I can say that two songs feel connected to me even when their sounds are very different.
Those judgments seem worth exploring. Whether they would provide enough information for useful discovery remains an open question.
There is a limit I want to keep clear. An audio model cannot recover my childhood memories from a waveform. If a song matters because of a person or a place, that connection would have to come from something I choose to share. And if a system notices a recurring musical feature in recordings I love, that would be a clue to investigate, not proof that it causes my response.
My own answers may change, too. Familiarity, attention, and the circumstances of listening all belong in the exploration. I would not want yesterday's preference to become a permanent definition of my taste.
Where I want to begin
I want to begin by listening more closely to the music I love. Which moments make me stop and listen? Which passages do I return to? Do they have something in common, even when they come from very different kinds of music?
I do not need an answer for every song. Sometimes a voice, a pause, or a few notes move me, and I cannot say why. I would like to understand a little more without losing the pleasure of simply listening.
I also want to leave room for memory. A song may matter because of what I hear in it, or because it reminds me of someone or somewhere. Both are part of my connection with the music.
My hope is to discover music I might otherwise pass by: a song, a performance, or a short passage that connects with something I already love.
That is where I want to begin. With the music, and with paying attention to what moves me.
Send the author a private note
Only they will see it — this note is never published.