# Change Audio Pitch
Free online pitch changer — shift any audio up or down by semitones without changing its speed. Transpose songs, fix tuning, deepen voices in your browser.
## Run it
- **CLI:** `gizza tool audio-pitch-shift 'url=https://example.com/input' 'semitones=3' 'format=mp3'`
- **Web:** https://gizza.ai/tools/audio-pitch-shift/
## Inputs
- `audio` — Audio _(file; accept: audio/*)_
- `semitones` — Semitones (-24 to 24) _(field)_
- `format` — Format _(field)_
## Output
- Pitch-shifted audio (audio)
## Query parameters
Open the tool pre-filled and auto-run via URL:
- `semitones` — Semitones (-24 to 24)
- `format` — Format
- `url` — fetch the input file from a public URL (CORS-permitting)
Example: `https://gizza.ai/tools/audio-pitch-shift/?url=https://example.com/input&semitones=3&format=mp3`
---
## Shift audio pitch in your browser
Pick an audio file and how many semitones to move it: positive values raise
the pitch, negative values lower it, and the speed stays exactly the same. Use
it to transpose a backing track into a singable key, deepen a voice recording,
or make the classic chipmunk / slowed-and-deep effects without touching the
tempo. Drag the slider for whole semitones, type a decimal for cent-level
fine-tuning, or hit one of the preset chips. The shift runs entirely in your
browser with ffmpeg compiled to WebAssembly, so your audio is never uploaded
to a server.
### Worked example
Transpose a backing track up a whole step (D to E): upload the track, set
**Semitones** to `2`, leave **Format** on `mp3`, and run. The result plays in
the new key at the original tempo and downloads as the original name with a
`-pitch-shifted.mp3` suffix. Some more common shifts:
- `12` — one octave up (a 440 Hz A becomes an 880 Hz A).
- `-12` — one octave down.
- `-3` — lower a song three semitones into a more comfortable vocal range
(a common karaoke trick when the original is too high).
- `1` — raise a recording made in E♭ tuning back to standard pitch, so you
can play along without retuning your guitar. `-1` does the reverse.
- `0.5` — a quarter-tone nudge (50 cents) for tuning mismatches; one cent is
`0.01`, so tuning an A=432 Hz recording up to A=440 Hz is about `0.32`
semitones (and `-0.32` converts 440 Hz music down to 432 Hz).
### Semitone cheat sheet
Count piano keys, black keys included — each key is one semitone. Moving from
C to D is `2`, C to E is `4`, down a perfect fourth is `-5`, and any octave is
`12`. So a song in G that you want in E goes down `-3`; a song in C that you
want in D goes up `2`. Producers use the same counting to layer harmonies: a
copy of a vocal shifted `3`, `4`, `5`, or `7` semitones lands on the classic
harmony intervals.
### How it works
The tool uses the classic resample method, fully offline: the audio is
resampled to a new rate (which shifts pitch and speed together, like changing
tape speed), then time-stretched back to the original duration with ffmpeg's
`atempo` filter. Each semitone is a frequency change of about 5.95% (the
twelfth root of 2), and fractional shifts are computed with the same formula.
Pitch accuracy is within a fraction of a cent, and the output duration
matches the input.
### Formats
- **MP3** — lossy at 192 kbps; small and playable everywhere (the default).
- **WAV** — lossless 16-bit PCM; largest, ideal for further editing.
- **OGG** — lossy Vorbis at 192 kbps; open format, good quality per byte.
- **FLAC** — lossless and compressed; smaller than WAV, still a perfect copy.
- **M4A** — AAC in an mp4 container; good quality at small sizes.
### Limits and edge cases
- Input files up to 10 MiB; anything ffmpeg can decode works (mp3, wav, flac,
m4a/aac, ogg, opus, and most video containers' audio tracks).
- Shifts are limited to ±24 semitones (two octaves). Large shifts sound
increasingly processed — time-stretching artifacts (a slight metallic or
fluttery quality) grow with the shift size; within ±5 semitones they are
hard to hear on most material.
- `0` semitones is rejected as a no-op rather than silently re-encoding your
file — use the audio-convert tool if you only want a format change.
- The output is always re-encoded (the filter chain requires it) and resampled
to 44.1 kHz, the CD-standard rate most music already uses.
- Everything in the mix shifts together — this transposes the whole track, it
cannot re-tune one instrument or correct a single off-key note.
## FAQ
Will changing the pitch also change the speed of my audio?
No — that is the point of this tool. The pitch moves by the semitones you
choose while the tempo and total duration stay the same. If you want the
tape-speed effect where pitch and speed change together (or a speed change
with the pitch preserved), use the change-speed tool instead.
Is a pitch changer the same as a pitch shifter?
Yes — "pitch changer", "pitch shifter", "key changer", and "transpose tool"
all describe the same operation: moving every note in a recording up or down
by a fixed interval while keeping the speed unchanged. This page does exactly
that, measured in semitones.
How many semitones do I need to change key?
Count the steps between the keys: each semitone is one step on a piano
(including black keys). C to D is `2`, C to E is `4`, down a fourth is `-5`.
An octave is `12`. Fractional values work too — `0.5` is a quarter tone, and
one cent is `0.01`, so you can fix small tuning offsets precisely.
Can I play along with a song recorded in E♭ (half-step-down) tuning?
Yes — shift the recording up `1` semitone and it lands in standard tuning, so
you can play along without retuning your instrument. The reverse works too:
shift a standard-tuning song down `-1` to match a guitar already tuned to
E♭, or down `-2` for D (whole-step-down) tunings.
Can I shift a voice without it sounding robotic?
Small shifts (roughly ±4 semitones) keep voices natural-sounding. Bigger
shifts change the character noticeably — up sounds chipmunk-like, down sounds
giant-like — because the formants (the resonances that make a voice sound
human) shift along with the pitch. That's fun for effects, but there's no
formant-preserving mode in this tool.
Is my audio uploaded anywhere?
No. The page downloads an ffmpeg WebAssembly build once and then processes
your file locally in the browser tab — the audio never leaves your device.