Slow it down.
Write it down.

How this tool slows music — and what that means for your ears

Tempo and pitch are linked in this tool: slow the playback and the pitch falls with it, exactly like a record turned at the wrong speed. For transcription, that linkage is not a flaw to route around — used at the right rate, it is the method.

At most rates, the drop is awkward. 75% speed puts the track about five semitones low — near a real interval but not exactly on it, so every note you try to name sits slightly off a piano key. Transcribing against that is guessing with extra steps.

At 50%, the linkage becomes clean. Half speed is exactly one octave down, and an octave preserves every note name. If you prefer to transcribe at, say, 70% with the pitch held at the original — no octave arithmetic at all — that is unlinked tempo and pitch, which is an app feature: the Y2K Player app does it and saves the setting per song. This page's tool is the quick, linked version, and at 50% it is genuinely enough to get real work done.

Half speed = one octave down: the old trick, still working

Players have transcribed at half speed for decades. A record cut for 33 1/3 rpm and played at 16 2/3 runs at exactly half speed; tape machines switched between speeds did the same thing. Long before software, people learned solos this way — and the octave relationship is why it worked.

Halving the speed lowers everything by exactly twelve semitones. Twelve semitones is one octave, and an octave changes no note names: a G is still a G, a minor chord is still the same minor chord. So the rule of the method is simple — write what you hear, one octave up. Hear a G3 in the slowed playback, write a G4. Nothing else about the music needs translating, because rhythm scales evenly too: a note on beat 3 is still on beat 3, just twice as far away in time.

Why slower hearing is better hearing

Fast lines blur into gestures; the ear catches the outline and invents the interior. At half speed each note has twice the duration, ornaments separate from the notes they decorate, and grace notes stop hiding inside the beat. You are not making the music simpler — you are giving your ear time to be accurate about it.

The working loop: small, repeated, checked

Transcription is a loop, in both senses. The workflow that survives contact with a hard solo is short phrases, repeated until they give in.

Set the loop around two to four beats — not bars, beats, when the line is dense. Let it repeat while you sing or hum the fragment back; if you can sing it, you can find it, and if you cannot sing it yet, more repetitions are cheaper than wrong guesses. Write it down, then move the loop forward so the new loop overlaps the last note you wrote. The overlap catches the errors that live on phrase boundaries.

The waveform view earns its place here: note attacks show as visible edges, so you can place loop points between notes instead of on top of them, and you can see how many attacks a blurred run actually contains before you try to name them.

Rhythm first, pitch second

When a phrase resists, split the job. Do one pass where you only write the rhythm — attacks and durations, no note names — at 75% or 85%, where timing is easy to feel. Then drop to 50% and fill in the pitches on the rhythm skeleton you already trust. Trying to catch both at once is how transcriptions stall; either job alone is usually easy.

A note on listening levels

Transcribing means hours of repeated listening, often on headphones. Keep the volume at the low end of comfortable — accuracy comes from repetition, not loudness, and a quiet detail you miss today will still be there tomorrow. Turning it up to catch a buried note works once; doing it for a whole session leaves your ears too tired to judge pitch well.

Start to finish: taking a solo off the record

1. Drop the song into the tool above. This page starts at 50% — exactly one octave down.
2. Play the passage once at 100% for context, then return to half speed.
3. Set the loop around the first two to four beats of the phrase, using the waveform's note attacks as edges.
4. Sing the fragment back until it is yours, find it on your instrument, and write it one octave up from what you hear.
5. Move the loop forward with a small overlap, and repeat until the phrase is done.
6. Check the whole phrase at 100% against what you wrote — full speed exposes rhythm errors that half speed forgives.
7. If you want the half-speed copy to keep working from, export it as MP3 or WAV — free, whole song, with a short spoken note after the end that you can crop off.

For the long transcriptions

You love taking music apart by ear? There is a player for that.

Y2K Player is a music player for iPhone and iPad, and it removes the two frictions this web tool leaves in place. First, unlinked tempo and pitch: transcribe at 70% or 80% in the original key, no octave arithmetic. Second, memory: it saves each song's settings, so a transcription you left mid-solo on Tuesday reopens at the same speed with the same loop on Thursday.

Loop and trim sit on a proper waveform, your transcription projects live in playlists next to your library, and everything works offline with no account. When a finished transcription becomes a cover, it exports video at 1080p over visualizers. Free to use. Pro is one payment — forever.