Découpeur Audio

The audio trimmer exports WAV by default because WAV is a straightforward uncompressed PCM format that browsers can encode natively without requiring a third-party codec library. When your audio is decoded and trimmed in the browser using the Web Audio API's AudioContext and AudioBuffer interfaces, the resulting audio data is an array of 32-bit floating-point PCM samples. Writing those samples directly to a WAV file is computationally trivial: the WAV format uses a simple header structure followed by the raw sample data, and the entire encoding can be done in a few lines of JavaScript with no external dependency. MP3 encoding, by contrast, requires a psychoacoustic compression algorithm — specifically the MPEG-1 Audio Layer III encoder — which is a complex DSP operation not natively available in browsers. Generating MP3 in the browser requires shipping a WebAssembly-compiled encoder such as LAME, which adds significant bundle weight and encoding time, particularly for long files. Additionally, MP3 is a lossy format, meaning encoding to MP3 after the trim introduces a generation loss — your audio goes through at least one lossy encode cycle, which subtly degrades quality relative to the source. WAV preserves every sample bit-perfectly from the trim operation. File size is larger than MP3, but for short clips this is rarely an issue. All processing runs entirely in your browser using the Web Audio API — no data leaves your device. As a practical tip, if you need a smaller file for sharing, import the exported WAV into any free audio app such as Audacity or VLC and encode it to MP3 at your desired bitrate without any further quality loss.

The audio trimmer decodes files using the browser's built-in Web Audio API, specifically the AudioContext.decodeAudioData() method, which delegates decoding to the browser's native codec stack. The formats supported therefore depend on the browser and operating system, but in practice all modern browsers universally support MP3 (MPEG-1 Audio Layer III), AAC (used in M4A and MP4 containers), WAV (PCM, uncompressed), FLAC (lossless), and Ogg Vorbis. WebM with Opus or Vorbis audio is also widely supported, particularly in Chrome and Firefox. Format support varies slightly by browser: Safari has excellent AAC and ALAC support due to macOS's Core Audio framework, while Firefox and Chrome offer strong Opus and Vorbis support. Formats unlikely to be supported natively include WMA (Windows Media Audio), AIFF on non-Apple browsers, and older formats like RealAudio or AMR. If you upload an unsupported format, the decodeAudioData() call will throw a DOMException with an EncodingError, and the tool will display an error message rather than crashing silently. For maximum compatibility, if you have audio in an unusual format, convert it to MP3 or WAV using a free tool before uploading. All decoding runs entirely in your browser — no data leaves your device, which is particularly important if your audio contains private recordings or confidential meetings. As a practical tip, if you are unsure whether your format will be accepted, simply try uploading it — the error feedback is immediate, and converting to a standard format takes only seconds with tools like VLC or FFmpeg.

There is no hard-coded length limit imposed by the tool itself, but practical limits arise from browser memory constraints. When you upload an audio file, the browser decodes the entire file into an AudioBuffer in memory. An AudioBuffer stores audio as 32-bit floating-point samples: a 10-minute stereo file at 44,100 Hz requires approximately 10 × 60 × 44,100 × 2 channels × 4 bytes = about 212 MB of RAM just for the decoded audio data, before the browser's own overhead. Most modern desktop browsers can handle this comfortably. A 30-minute stereo file would require approximately 635 MB of decoded audio RAM, which is feasible on machines with 8 GB or more of system memory but may cause the browser tab to become slow or, on memory-constrained devices, crash. Mobile browsers, which typically operate with stricter memory limits per tab — often capped at 256–512 MB for the entire tab — are more likely to struggle with files longer than 15–20 minutes. The file size on disk is a rough indicator: a compressed MP3 file that is 50 MB on disk may expand to 10× or more in uncompressed PCM form during decoding. The tool provides a waveform visualization rendered via the Canvas API using the decoded sample data, and drawing a high-resolution waveform for a very long file also adds GPU memory overhead. All processing runs entirely in your browser — no data leaves your device. As a practical tip, if you need to trim a very long recording, first rough-cut it into shorter segments using a desktop audio tool like Audacity, then use the browser trimmer for precision fine-tuning of each segment.

Not natively within the current version of the tool. The browser's Web Audio API provides decoding support for many compressed formats through AudioContext.decodeAudioData(), but it does not expose a corresponding encode path for lossy codecs like MP3. The only audio format the browser can generate natively from raw PCM sample data is WAV, which is why the export defaults to that format. Producing an MP3 from the browser requires a JavaScript or WebAssembly implementation of the LAME encoder, which is a substantial library addition. Some tools implement this using lamejs or a WebAssembly build of libmp3lame; these work but add a noticeable encoding delay for longer files and introduce the quality trade-off of lossy encoding. The current WikiPlus audio trimmer prioritizes simplicity, speed, and zero-dependency local processing, so MP3 export is not included. For users who need MP3 output, the recommended workflow is: trim and download the WAV from this tool, then convert the WAV to MP3 using any free application. VLC Media Player (File > Convert/Stream) can do this with a visual interface and no command line needed. FFmpeg users can run a single command: ffmpeg -i trimmed.wav -codec:a libmp3lame -q:a 2 trimmed.mp3. Audacity also exports MP3 directly after installing the LAME plugin. All audio processing in this tool runs entirely in your browser — no data leaves your device. As a practical tip, if small file size is critical, use the WAV export for archiving and keep the MP3 conversion as a last step to avoid accumulating generation losses across multiple edits.