Video to MP3 Converter – Extract the Soundtrack Without Uploading the Video
This video to MP3 converter pulls the sound out of a video you already have and saves it as an MP3, a WAV, or the video's original audio without re-encoding. Open an MP4, MOV, M4V, WebM, MKV or MPEG-TS file, optionally pick the part you want on a waveform of the soundtrack, and download the result. Everything happens inside your browser tab, so a lecture recording, a family clip or a meeting never leaves your device, and there is no length limit in browsers that support WebCodecs audio.
Containers, codecs and why most videos hold AAC
A video file is a container (MP4, MOV, WebM, MKV) holding separate tracks, and the audio track has its own codec. Phones, cameras and most websites store AAC; WebM files carry Opus or Vorbis; MKV and broadcast recordings sometimes hold AC-3, E-AC-3 or MP3. The converter reads the container, shows you the codec, bitrate and channel layout of every audio track, and lets you choose a track when there are several (for example a commentary or a second language).
Copying is lossless and instant; re-encoding is not
Because the audio is already compressed, the fastest and cleanest option is to copy those packets into an audio-only file: AAC becomes .m4a, Opus becomes .opus, and an MP3 soundtrack is copied straight to .mp3. Nothing is decoded, so quality is identical and it takes about a second even for a long film. Converting to MP3 instead decodes the AAC and encodes it again with LAME. Each lossy encode throws away a little more detail, an effect called generation loss, which is why the tool never offers bitrates as a way to "add" quality the video never had.
How MP3 bitrate turns into file size
MP3 here is constant bitrate, so the size is simple arithmetic: bytes ≈ seconds × kbps × 125. A 30-minute podcast at 64 kbps mono is about 14 MB; the same half hour at 192 kbps stereo is about 41 MB. The bitrate chart draws these sizes for your selection and marks the row that roughly matches the video's own audio. As a rule of thumb MP3 needs about 1.5 times the bitrate of AAC or Opus for similar quality, so a 128 kbps AAC soundtrack maps to 192 kbps MP3. To compare bitrates and sizes without a file, try the Audio Bitrate Converter.
Loudness versus peak: LUFS, K-weighting and gating
Phone and meeting recordings are often quiet. Simply raising the volume until the loudest sample hits the ceiling (peak normalisation) fails when one clap or cough is much louder than the speech around it. The Volume option instead measures integrated loudness in LUFS using ITU-R BS.1770: the audio is passed through a K-weighting filter that mimics how we hear, averaged in overlapping 400 ms blocks, and silent or very quiet blocks are gated out so pauses don't drag the number down. The tool then applies one gain so the selection reaches the target, while keeping sample peaks under−1 dBFS and never adding more than 20 dB.
- −16 LUFS: the usual target for speech and podcasts.
- −14 LUFS: roughly where Spotify and YouTube normalise music playback.
- −23 LUFS: the EBU R 128 broadcast target, shown on the gauge for reference.
Mono, stereo and multichannel sources
Mono halves the bitrate needed for the same quality and is ideal for a single voice. Stereo keeps music and ambience wide. Surround soundtracks (5.1, 7.1) are mixed down to stereo with the standard speaker rules before encoding, and the loudness and clipping checks are made on that mix, so what you measure is what you get. If a loud 5.1 mix would clip once folded to stereo, the tool warns you and suggests a Volume option.
Trimming, fades and tags
Drag the handles on the waveform, type exact times, or set the start and end at the playhead with the [ and ] keys. Short linear fades and a 5 ms anti-click ramp smooth hard cuts. Title, artist, album, year, a comment and cover art (a frame of the video, the file's own artwork or an image you choose) are written as ID3v2.4 tags so music and podcast apps show them. Only the fields you see are written; other metadata in the video, such as location, is left out. For sample-accurate editing of the result, open it in the Audio Trimmer; to cut the video itself, use the Video Trimmer.
Credits
Reading, decoding and writing audio uses Mediabunny (MPL-2.0, source). MP3 encoding uses LAME (LGPL), compiled to WebAssembly and loaded only when you convert.