Speed & Pitch Changer

About the Speed & Pitch Changer

Transpose any track to a target musical key and time-stretch it independently, right in your browser. The continuous pitch control covers −12 to +12 semitones plus ±50 cents of fine tuning, and the speed slider sweeps 0.5× to 2.0×. A pure pitch shift never changes the duration, and the output keeps your source sample rate (44.1k, 48k or 96k), so it is safe for broadcast and DAW material. Built for musicians matching a vocal to a karaoke key, producers detuning samples, DJs nudging BPM, and learners slowing speech for transcription.

How does the key/tempo transposer work?

Pitch is shifted with FFmpeg asetrate (raise the declared sample rate by 2^(semitones/12 + cents/1200) to go up, lower it to go down) and the tempo change that introduces is cancelled with an atempo WSOLA time-stretch. Set speed to 1.0× and the durations match exactly, so a pure transposition never changes length.

Why does a pure pitch shift keep the same duration?

Transposing a song to another key should not make it longer or shorter. When speed is 1.0× the tempo compensation is exactly 1/pitchFactor, which divides out the duration change asetrate would otherwise cause. The readout shows the same original and new duration to confirm it.

What pitch range and resolution do I get?

−12 to +12 semitones (a full octave each way) in 1-semitone steps, plus a separate cents slider for ±50 cents. Use semitones to transpose to a key and cents to match a reference pitch (for example A=440) or to subtly detune a layered sample.

Can I speed up or slow down without changing pitch?

Yes. Leave pitch and cents at 0 and move the speed slider between 0.5× and 2.0×. The atempo time-stretch keeps the original pitch, which is ideal for transcription, practice slow-downs, or speeding up a podcast.

Does it preserve sample rate, bit depth and channels?

The tool reads the decoded source sample rate (44.1k/48k/96k) and writes the output at that same rate, so 48 kHz broadcast or 96 kHz DAW material stays at its native clock. WAV exports as 16-bit PCM; stereo and mono inputs keep their channel count.

How do cents help with tuning?

100 cents equal one semitone. Cents let you fine-tune below a semitone — nudge +10 cents to match an instrument tuned slightly sharp, or detune two copies of a sample by a few cents for a wider, chorused sound.

What about BPM and key matching for DJ sets?

The speed slider is also a tempo ratio: 1.10× raises a 120 BPM track to roughly 132 BPM. The readout shows the speed/BPM ratio and the resulting duration so you can beat-match or fit a track to a fixed slot.

Are there limits to natural-sounding shifts?

This is a sample-rate (resampling) pitch shift, not a formant-preserving one, so large vocal shifts will sound chipmunked or deepened. For natural voices stay within about ±5 semitones; instruments and sound-design detuning tolerate the full ±12 range.

What formats can I load and export?

Load MP3, WAV, OGG, FLAC, OPUS, M4A and common video containers (MP4, WEBM, MOV) for audio extraction. Export to the same format or convert to MP3, WAV, OGG, or AAC (muxed into an M4A container so it plays everywhere).

Is my audio uploaded anywhere?

No. Decoding, pitch/tempo processing and encoding all run locally via Web Audio API and FFmpeg.wasm. Your file never leaves the browser, which matters for unreleased songs or private interviews. Maximum file size is 50MB.