Initializing... drag & drop files here
Supports: MKV
Extracting an MKV's soundtrack to WAV gives you raw PCM: no second lossy stage, nothing thrown away beyond what the source already discarded, and a file every editor and DAW on earth opens without argument. It also gives you a file that is often larger than the video it came from. The short version: pick WAV when the audio is going into an editor, a mastering chain, or a transcription tool. Pick FLAC when you want the same fidelity at roughly half the size, and MP3 or Opus when the file has to be sent somewhere. This page is the WAV path — here is how it compares before you commit.
| WAV (this page) | FLAC | MP3 | |
|---|---|---|---|
| What it stores | Raw PCM samples in a RIFF container | The same samples, losslessly compressed | A new lossy encode |
| Adds further quality loss | No | No | Yes |
| Size for 60 minutes of 16-bit/44.1 kHz stereo | About 635 MB | Typically 300–450 MB, content-dependent | About 86 MB at 192 kbps |
| Editing and DAW support | Universal, and no decode step on load | Very good, but some hardware and older tools refuse it | Universal, though lossy edits compound |
| Metadata and tagging | Weak — RIFF tagging is patchy and inconsistently read | Strong, with a proper tag block | Strong, via ID3 |
| Hard size ceiling | Yes — about 4 GiB, see below | No practical limit | No practical limit |
| Best for | Editing, mastering, transcription, restoration | Archiving the extracted audio | Sharing, phones, car stereos |
| xconvert page | This one | MKV to FLAC | MKV to MP3 |
The Audio Codec control here does not choose a compression scheme — it chooses how each sample is stored. Two of the five entries are not what most people expect from a "WAV", so read before selecting. Sizes assume stereo, except the G.711 row.
| Audio Codec | What it is | Data rate | Time before hitting 4 GiB |
|---|---|---|---|
| PCM 16-bit Little Endian (default) | CD-standard linear PCM | 1,411 kbps at 44.1 kHz — about 10.6 MB per minute | Roughly 6.8 hours |
| PCM 24-bit Little Endian | Studio-standard linear PCM, extra headroom for gain changes | About 17.3 MB per minute at 48 kHz | Roughly 4.1 hours |
| PCM 32-bit Little Endian | 32-bit integer PCM; maximum precision, rarely necessary | About 23 MB per minute at 48 kHz | Roughly 3.1 hours |
| PCM A-law | G.711 telephony companding, 8-bit — lossy, not linear PCM | 64 kbps per channel | Not a practical concern |
| PCM mu-law | The North American / Japanese G.711 variant — also lossy | 64 kbps per channel | Not a practical concern |
There is no bitrate, quality preset, or file-size control on this page, and that is correct rather than an omission: linear PCM has no compression to tune. Size follows directly from bit depth × sample rate × channels × duration.
.mkv onto the page or click "Add Files" to browse. Queue several videos to extract them all with the same settings..wav. Files are uploaded over an encrypted connection, processed on our servers, and deleted automatically after a few hours — no sign-up, no watermark.Not necessarily the first one — this trips people up constantly. Matroska routinely carries a 5.1 mix, a stereo downmix, a director's commentary, and several language dubs in one file. With no explicit stream mapping, ffmpeg's documented automatic selection takes "the stream with the most channels" for audio and breaks ties by lowest index, so the 5.1 mix usually wins even when the stereo track appears first in the file. If you need a particular dub or commentary, isolate that track in a Matroska editor before uploading.
No, and this is the most common misconception about the format. WAV faithfully stores whatever the decoder handed it — if the MKV's track was AAC, AC-3, DTS, or Opus, the WAV preserves those existing artefacts exactly rather than removing them. Choosing 32-bit or 96 kHz makes a bigger file, not a better one. What WAV genuinely buys you is a working master that costs nothing in quality across repeated edits and saves.
Because you traded compression for simplicity. A 16-bit/44.1 kHz stereo PCM stream runs at a constant 1,411 kbps, while the lossy audio inside a typical MKV sits somewhere between 128 and 640 kbps — and the video, which you just discarded, was itself compressed dozens of times more aggressively. Dropping the picture and still ending up with a larger file is normal here. FLAC holds the same samples in roughly half the space if the size bothers you.
Yes, and it is structural. WAV is built on RIFF, whose headers store chunk and file lengths in 32-bit fields, capping a conventional .wav at 4,294,967,296 bytes. The RF64 and Wave64 extensions exist to get past that, but our output is written as plain RIFF, so a recording long enough to exceed the cap produces a file whose header cannot describe it — some editors will still read the audio, many players will not. At the 16-bit/44.1 kHz default that ceiling is about 6.8 hours; at 32-bit it is closer to three. Split long sources before uploading.
Yes. G.711 A-law and mu-law are companded 8-bit encodings designed for telephone networks: they squeeze a wider dynamic range into eight bits using a logarithmic curve, which discards resolution permanently. They sit under the Audio Codec list because WAV can legitimately wrap them, and they are the right pick when you are feeding a telephony or IVR system that expects G.711 at 64 kbps. For anything musical or archival they are the wrong choice — stay on the linear PCM entries.
There is no fidelity reason to. Extra bit depth cannot recover detail that AAC or AC-3 already threw away, so a 24-bit extraction of a lossy track is a bigger file holding the same information. The one legitimate case is workflow: if the WAV is going into a chain that will apply gain, EQ, or heavy processing, the extra headroom keeps intermediate rounding out of the audible range. Extract at 16-bit if the file is going straight to transcription or a simple edit.
Yes, two ways. Use the Trim control on this page to set a start time and duration, which extracts only that window. Or, if the source is a large film and you want to shorten the upload itself, cut the video first with the video cutter — note that the audio cutter takes audio files rather than video, so it is the tool for trimming the WAV afterwards, not the MKV beforehand.
Your MKV is uploaded over an encrypted connection, converted on our servers, and both the upload and the extracted WAV are deleted automatically after a few hours. There is no sign-up, no watermark, and files are never shared or made public. Since PCM output is large, upload and download time is usually the slow part rather than the conversion — if you need a compact copy afterwards, WAV to MP3 handles that in one more step.