如何不上传视频就把 MP4 里的音频提取为 MP3

Bottom line: MP4 is a container, not a single kind of media. Inside most MP4 files you will find H.264, HEVC or AV1 video plus AAC audio. When you “convert MP4 to MP3”, you are actually demuxing that audio track, decoding it if needed, and re-encoding it into MPEG Audio Layer III. That distinction matters because audio quality is controlled by the source and the MP3 bitrate, not by the MP4 extension itself.

AIMP6 runs this workflow on the client for normal-sized files. The browser loads FFmpeg.wasm only after you start a job, reads the file from local memory or the Origin Private File System, writes an MP3, and hands the result back to your device. The video never needs to be uploaded to a conversion server, which is especially useful for interviews, lectures, private voice recordings and draft edits.

Start by choosing the highest-quality source available. A low-bitrate webinar recording cannot become a detailed master simply by selecting 320 kbps MP3. For voice, 96–128 kbps is often enough; for music, 192 kbps is a sensible default; 320 kbps should be reserved for cases where you need maximum compatibility rather than efficient file size.

After extraction, listen to the beginning and end of the track. Container timestamps occasionally cause a short delay or clipped ending. If the result is unexpectedly quiet, the source probably uses unusual dialnorm metadata or was already encoded at a low volume. Re-encoding cannot recover detail that was removed earlier.

Browser memory is the main limit. Long multi-gigabyte files can strain a tab because WASM operates inside a memory envelope. Keep local processing for clips and everyday files; use a cloud job for long recordings. In both cases, preserve the original until you have verified the output.