Initializing... drag & drop files here
Supports: MTS
This tool extracts the audio from an AVCHD camcorder clip and saves it as FLAC. The honest framing matters here, because it is the thing every other converter glosses over: FLAC is lossless, but converting into it does not undo what the camcorder already threw away. Most handycams record Dolby Digital AC-3, which is a lossy codec. FLAC will preserve that audio perfectly from this point forward — it will not restore detail the AC-3 encoder discarded when you pressed record.
That is still a good reason to use FLAC. It means no further generation loss for archiving, editing or mastering. It is a bad reason to expect the audio to sound better than the original clip, because it will sound exactly the same.
| Property | AVCHD audio (inside your .MTS) | FLAC output |
|---|---|---|
| Codec | Dolby Digital (AC-3) or linear PCM, per the AVCHD spec | Free Lossless Audio Codec |
| Lossy or lossless | AC-3 is lossy; LPCM is lossless | Lossless, always |
| Standard | AVCHD specification (Sony/Panasonic, 2006) | RFC 9639, published December 2024 |
| Channel layouts seen on camcorders | Stereo, or 5.1 on surround-capable models | Whatever the selected source track has, or your Mono/Stereo choice |
| Typical size | Small relative to video, but bundled inside the clip | Larger than AC-3, smaller than uncompressed WAV |
| Editable without further loss | Re-encoding AC-3 loses more each time | Yes — decompresses bit-exact every time |
| Plays on portable hardware | N/A | Widely, though some older players still refuse FLAC |
| Best for | What the camera happened to record | Archiving, editing, anything you will re-encode later |
.mts; queue several clips and they all convert with the same settings.FLAC's compression level is one of the most misread controls in audio. Every level from 1 to 12 produces bit-identical audio when decoded — that is what lossless means. What changes is how hard the encoder works to find redundancy, which costs time and buys you a smaller file. The tool's own guidance says as much: lower numbers are faster but produce larger files, higher numbers compress harder, and the setting does not affect audio quality.
| Compression level | Encoding speed | File size | Audio |
|---|---|---|---|
| 1–4 | Fastest | Largest | Bit-identical |
| 5–8 | Moderate | Smaller | Bit-identical |
| 9–12 (opens at 12) | Slowest | Smallest | Bit-identical |
There is one more behaviour worth knowing, because this page has no track picker. When a camcorder clip contains more than one audio stream, the selection follows ffmpeg's documented default: the stream with the most channels wins, and ties are broken by the lowest stream index. On a 5.1-capable camcorder that means a surround track will be chosen over a stereo one, even if stereo is what you expected. If you want a two-channel result regardless, set Audio Channel to Stereo and the surround mix will be folded down.
No. FLAC is lossless, which means it reproduces its input exactly — it cannot add back information that was never stored. If your camcorder recorded Dolby Digital AC-3, the lossy compression happened inside the camera and is permanent. What FLAC gives you is a guarantee that no further quality is lost from here on, which is exactly what you want for archiving or editing.
It depends on the recording mode. The AVCHD specification permits Dolby Digital (AC-3) and linear PCM. Consumer camcorders overwhelmingly use AC-3, which is lossy. Some higher-end models and modes record LPCM, which is already lossless — in that case converting to FLAC is a genuine lossless-to-lossless move that just makes the file smaller.
Any of them, as far as sound is concerned — the audio decodes bit-identically at every level. Choose 12 (the default) if you want the smallest file and do not mind waiting longer. Choose something in the 5–8 range if you are converting a lot of long camcorder clips and want the process to finish sooner. There is no quality trade-off to weigh.
The surround one. This page has no track selector, so stream selection follows ffmpeg's default rule: the audio stream with the most channels is chosen, with ties resolved by the lowest index. A 5.1 track therefore beats a stereo track. If you specifically want two channels out, set Audio Channel to Stereo and the multichannel source is downmixed.
Noticeably bigger, because you are moving from a lossy codec to a lossless one. AC-3 in a camcorder clip discards data to save space; FLAC has to store every sample of the decoded result. Expect a substantial jump over the AC-3 stream, while still landing well under what an uncompressed WAV of the same audio would take.
Usually not. Leaving Audio Sample Rate on the original avoids an unnecessary resampling step, and resampling can only ever approximate. Change it only when something downstream demands a specific rate — and note that upsampling adds no detail, it just makes the file larger.
Yes, using the Trim control to set a start time and duration. If you would rather cut first, the video cutter accepts .mts files directly. Note that the audio cutter does not take video files, so you cannot feed it an .MTS — trim the source as video, or convert to FLAC first and cut the FLAC afterwards.
No. This conversion produces an audio-only FLAC file; the picture is discarded. That is usually the point — people come here to pull a soundtrack, an interview or ambient sound off camcorder footage. If you need both, convert the clip to a video format instead.
This page accepts .mts only. Camcorder clips copied to a computer by software are frequently renamed .m2ts, and although the stream inside is byte-for-byte the same kind of data, the two extensions are served by separate pages here. Rename the file back to .mts, or use the M2TS equivalent page.
Your .MTS file is uploaded over an encrypted connection and the audio is extracted on our servers. Uploaded files and their results are deleted automatically after a few hours, are never shared or made public, and there is no sign-up and no watermark.