How the Audio Merger Works
AnyTool joins audio entirely in the browser using two engines. A Web Audio preview decodes each file with decodeAudioData, draws per-file waveforms and plays the full merged sequence — with per-file trim, gaps and crossfades applied via scheduled AudioBufferSourceNodes and GainNode curves — at zero encoding cost. The final export uses a self-hosted single-threaded ffmpeg.wasm engine (bundled core, no CDN): inputs are resampled to a common base, then concatenated (or blended with acrossfade), with optional EBU R128 loudnorm level-matching, and encoded to MP3 (libmp3lame), WAV (pcm_s16le), M4A (AAC), OGG (libvorbis) or FLAC. A WAV-only export can be rendered directly via an OfflineAudioContext for a sub-second result.
- Runs client-side — tracks are never uploaded
- Self-hosted ffmpeg core (bundled) — no CDN
- Crossfade, gap/gapless and per-file trim with drag reorder
- EBU R128 loudness level-matching across tracks
- Live Web Audio preview of the merged result before export
- MP3/WAV/M4A/OGG/FLAC with bitrate control
