Audio Speed Changer

FFmpeg's atempo filter uses time-stretching, not simple sample rate manipulation. When you change playback speed by resampling (playing at a different sample rate), the pitch shifts proportionally — which is the classic chipmunk-fast or slow-monster effect. Time-stretching analyzes overlapping windows of audio and shrinks or expands the time axis while keeping frequency content (pitch) constant.

The atempo filter's valid range per pass is 0.5× to 2.0×. For extremes like 0.25×, we chain two passes: atempo=0.5 applied twice produces 0.25× speed. The result at all supported presets is a properly stretched audio file at the same pitch as the original.

At 0.25×: dramatically slowed, useful for identifying individual notes or phonemes. Speech becomes very slow but remains intelligible for careful listening.

At 0.5×: half speed. The standard for musician practice — all notes are present, rhythm is clearly audible, passages that blur together at full speed become separable. Good for transcription of dense speech too.

At 0.75×: the sweet spot for language learning and slower transcription. Barely perceptible slowing to an untrained ear; significantly easier to follow for a non-native speaker.

At 1.25×: slightly faster. Useful for long content where you want modest time savings without the effort of concentrating at 1.5×.

At 1.5×: the standard commute-mode speed for podcasts and lectures. Most well-articulated speakers are fully comprehensible. Background music and dense technical content are harder.

At 1.75×: aggressive. Works for familiar speakers you know well and straightforward content. Not great for dense technical material.

At 2×: double speed. The maximum available. Works for re-listening to familiar content, fast-scanning a recording you already know.

The speed changer always outputs MP3 at 192 kbps, regardless of the input format. This ensures broad compatibility — the result plays on every device without any special format support. The atempo filter operates on decoded PCM audio before encoding, so no quality is lost in the speed transformation itself; the only quality reduction relative to a lossless source is the MP3 encoding at 192 kbps, which is audibly transparent for most material on typical playback gear.

Speed changers that work server-side upload your audio file, process it, and stream back the result — which means your audio passes through a third-party server. This tool uses FFmpeg compiled to WebAssembly. All processing happens in browser memory. Nothing is uploaded. The WASM bundle loads once and is cached for future visits. Disconnect from the internet after page load and the tool still works — proof that there's no server in the loop.