How to Transcribe Music by Ear
Transcribing (lifting notes and chords straight off a recording) is where ear training finally pays off in real music. From the outside it looks like a talent. It's actually a workflow, and here it is, step by step.
Build the skills freeNo account needed, and it works on your phone or desktop.
Why transcribe at all?
Every hour you spend transcribing is compound ear training: you learn the song, and your ear learns the patterns the song is made of. Players who transcribe regularly stop needing sheet music for everything. They pick up tunes at jam sessions in real time. They soak up the vocabulary of the styles they love. It's the closest thing music has to learning a language by living where it's spoken.
The workflow
Step 1: listen through, twice, doing nothing
Resist the urge to grab your instrument. Listen once for the shape of the song (sections, repeats, where the energy moves), then once more while humming along with the part you want to transcribe. Here's the thing: if you can't sing a phrase from memory, you're not ready to find its notes. Singing it is how you carry it to the instrument.
Step 2: find the key
Hum the note the song feels like it wants to end on. That's almost always the tonic. Find it on your instrument, then decide major or minor by the overall color (the same bright-or-dark judgment you build in chord recognition). Knowing the key shrinks every later search from twelve candidates down to mostly seven.
Step 3: follow the bass
Loop four or eight bars and track only the lowest voice. Bass notes are usually chord roots, so this step hands you the song's harmonic skeleton. Listen for the intervals the bass moves by: down a 5th, up a 4th, down a step. Those moves you've been drilling in the interval trainer? They're exactly what root motion is made of.
Step 4: name the chord qualities
Once you've got the roots, each chord only needs a quality, and major, minor, or dominant 7th covers most popular music. Use the color method from recognizing chords by ear: is it bright, dark, or bright but pushing? And write the progression down as you go. Paper is part of the workflow.
Step 5: chunk the melody
Take the melody two to four notes at a time, about what working memory holds comfortably. Sing the chunk, find it, check it against the recording, then join it to the previous chunk. First identify the starting note relative to the chord underneath it (it's the root, 3rd, or 5th most of the time), then hear each move as an interval. And never hunt note-by-note with the recording playing. Sing first, then find.
Step 6: slow down what resists
Fast lines? Use any player that slows audio down without changing the pitch (most streaming and practice apps do 50 to 75%). This isn't cheating. It's how everyone learns fast passages, and the speed comes back on its own once the notes are known.
Step 7: play along, then write it down
You're done when you can play the passage along with the recording and it disappears into the track. Then write it out (notation, tab, or a chord chart). That locks in the learning and builds a personal library of vocabulary that's yours.
The skills underneath, and where to train them
- Pitch memory, holding a note in your head long enough to find it: note recognition.
- Interval recognition, hearing melodic moves and bass motion: the interval trainer, plus anchor songs for quick reference.
- Chord quality, that major/minor/dominant call at every bar: the chord trainer.
- Consistency, because ten minutes a day beats marathon sessions: the daily routine and the daily challenge.
Starting from zero? Work through the beginner's guide to ear training first. Transcription gets dramatically easier just a few weeks in.
Frequently asked questions
What should I transcribe first?
Simple, slow melodies you already know well: nursery rhymes, hymn tunes, folk songs, then pop vocal lines. Familiarity lets your memory carry part of the load while your ear learns the mechanics. Save the solos and full chord charts for later.
Why start with the bass line?
Bass notes are usually chord roots, so the bass reveals the song's harmonic skeleton and turns every chord into a single quality decision. It's also easier to pick out by ear than the inner voices.
Is slowing the recording down cheating?
No, it's standard practice at every level. The skill you're training is pitch recognition, not reaction time, and speed comes back naturally once the notes are known.
How does ear training speed this up?
Transcription is really just interval recognition, chord recognition, and pitch memory applied in sequence. Every sound you can name instantly is a search you skip.