Audio converter that never uploads your file

No. The converter is FFmpeg compiled to WebAssembly and it runs inside this browser tab, so your file never travels anywhere. You do not have to take our word for it: open your browser's developer tools, switch to the Network panel, and convert something. There are no requests during the conversion. Or disconnect from the internet after this page loads and convert anyway — it still finishes.

No account, no email, no credits and no watermark. There is also no per-day conversion counter, because there is no server keeping count — your own processor does the work, so it costs us nothing when you convert a hundred files. The only real ceiling is your device: we warn above about 300 MB and decline above 2 GB, which is the browser tab's limit rather than a rule we invented.

Yes, a little. Both are lossy, so the MP3 encoder has to make its guesses again on audio the AAC encoder has already thinned out. At 192 kbps or above the difference is inaudible to almost everyone, and for speech 96 kbps mono is plenty. What does not help is picking a higher bitrate than the source had — that adds bytes, not detail.

No. Renaming changes the label on the file, not the audio inside it. The contents are still AAC in an MPEG-4 container, so lenient players may open it while the strict ones you were trying to satisfy — submission forms, car stereos, DAWs — still reject it. It is the most repeated piece of bad advice on this subject and it fails exactly where it matters.

Rarely. The one measurable difference is that 320 preserves a little over a kilohertz more treble, which sits above most adults' hearing, and nobody has produced a blind test showing 320 beats the top variable-bitrate settings. It costs about 66% more storage. Choose it if you are archiving; 192 is genuinely transparent for most listening, and speech does not need either.

No. WAV is uncompressed, not restorative — the detail the MP3 encoder discarded is gone, and wrapping what remains in a bigger container will not bring it back. The file will grow roughly tenfold and sound identical. There are good reasons to do it anyway: editors, DAWs and some transcription tools want WAV, and working in WAV avoids stacking up another generation of loss.

Two kinds. If the audio is already in the codec you want, we copy it rather than re-encode it — AAC out of an .m4a into an .m4a, .aac or an .m4r ringtone comes out with the audio untouched — we measured the extracted stream and the decoded samples and both matched exactly — as does the AAC soundtrack lifted from an .mp4. And lossless to lossless — FLAC, ALAC, WAV, AIFF — can be converted as often as you like. Everything else is a fresh encode.

M4A is a container, not a codec, and that is the catch: it usually holds AAC, which is lossy, but it can hold Apple Lossless instead. The extension does not tell you which, and the two can differ in size by five times. iPhone Voice Memos writes either one depending on a setting. We detect and show you what is actually inside before you convert.

Share it out of Voice Memos first — Share, then Save to Files — because the recording is not a normal file until you do. Then convert the .m4a. Do not rename the extension; it does not transcode. And check whether the recipient actually rejects M4A, because plenty of services take it and converting would only cost you a generation of quality for nothing.

Because an .m4a holds AAC inside an MPEG-4 container, which that player was never taught to read. Microsoft's own forum answers suggest a codec pack or converting to WMA — the first is risky and the second moves you into a more obsolete format. Converting once to MP3, or opening the file in VLC or a browser, both work immediately.

No, and neither can anything else. A .cda is a shortcut of about 44 bytes that points at a track position on the disc; it contains no audio whatsoever, so there is nothing in it to convert. You have to rip the CD with something that reads the disc directly — Windows Media Player Legacy, VLC or Exact Audio Copy — and convert what comes out of that.

Because the good Opus encoder crashes this WebAssembly build, and FFmpeg's experimental one would quietly hand you a worse file at the same bitrate. We would rather say so than substitute it silently. Opus decoding works perfectly, which is the direction that actually matters — WhatsApp voice notes arrive as Opus and people need them as MP3.

Trim the dead air first, because that is free. Then drop the bitrate — 96 kbps is fine for speech, 128 for music — and fold a spoken-word recording to mono, which halves it again. Gmail's 25 MB is really about 18 MB of actual file once it is encoded for transit, so aim below that rather than at it.

You can make the file: trim to 30 seconds and export .m4r or a short MP3, and if your source is already AAC the ringtone is a straight copy of the audio with no re-encoding. Installing it is the phone's job, not the browser's. On iOS 26 you can do that from the Files app — long-press, Share, Use as Ringtone — and on older versions you still need Finder or iTunes.