Begin with the receiving application
A file format is useful only if the next player, editor, car system or phone workflow accepts it. Check that product's current documentation or import dialog first. Broad statements such as “every device opens MP3” or “phones play Opus” hide model, software-version and container differences.
| Output | PrivateMP3 processing | Choose it when |
|---|---|---|
| MP3 | New 192 kbps lossy encode | The destination explicitly accepts or requires MP3. |
| M4A | Packet copy for unchanged AAC, AC-3 or E-AC-3; otherwise new 192 kbps AAC | The destination accepts AAC in an M4A container. |
| WAV | Decoded 16-bit PCM | An editor or analysis tool requests PCM/WAV. |
PrivateMP3 does not output FLAC, Opus or Vorbis. Converting to WAV does not turn a lossy source into a higher-quality recording; it only avoids adding another lossy encoding before the next editing step.
Container and codec support are separate
M4A and MP4 are names used for files in the MP4 container family. AAC is an audio codec commonly carried inside them. An application may recognize the container but lack the decoder for the particular codec or profile, so an extension by itself is not a complete compatibility test.
Renaming .mp4 to .m4a does not remove video or rewrite the track structure. PrivateMP3 creates a new audio-only output containing the selected audio track. Likewise, renaming WebM or MKV to MP3 does not convert Opus, Vorbis or another codec into MP3.
Audacity instructions depend on version and operating system
Audacity documentation is not one universal rule. Its current desktop manual says macOS can import unprotected M4A, MP4 and MOV files as shipped, while Windows and GNU/Linux use the optional FFmpeg library for those formats. The separate Audacity 4 support page says FFmpeg is included by default on Windows and macOS and gives separate Linux instructions.
Check which Audacity generation is installed before converting only to work around an import error. Adding or enabling the documented FFmpeg component may let you retain an existing M4A instead of making an MP3. Files protected by digital rights management are a different issue; PrivateMP3 does not bypass that protection.
Playback, editing and ringtone installation are different tests
A phone being able to play an audio file does not mean its ringtone settings can install that file directly. An editor accepting WAV does not mean a car stereo will. Compatibility must be checked against the exact action, not merely the device name.
PrivateMP3 creates an audio file and triggers a normal browser download. It does not place the file in a phone's ringtone library or another application's project. For that separate workflow, use the clip task in the converter and then follow the current instructions for the receiving phone.
A short decision order
- Try the existing local audio in the receiving application if you already have it.
- If it accepts M4A and PrivateMP3 shows the copy note, choose M4A to keep the original compressed audio; the M4A section of the size guide lists the conditions.
- If the destination specifically requests MP3, make the fixed 192 kbps MP3.
- If the next step is editing or analysis and it requests PCM, choose WAV; the WAV section covers its size.
- Test the saved file in the actual destination before deleting or replacing the source.
How the format tests were run
The measured tables below were made on 6 October 2026 in Chrome 154.0.8037.98 and Edge 154.0.4258.53 on Windows 11 (Intel Core i5-13600KF, 32 GB), with files generated by FFmpeg from a 220 Hz tone and quiet pink noise. Times are from Chrome; Edge differed by at most 0.7 s on any row. Each saved file was read back with ffprobe. The method is described with the screen recording test results.
MP4: check the audio track, not the video codec
MP4 is a media container that can hold separately encoded video and audio tracks. H.264, HEVC or AV1 describes the video; AAC, MP3 or another audio codec describes the sound. Those are independent choices. PrivateMP3 lists the tracks and sends only the selected audio track through the extraction engine. It does not need to decode the video frames to make an audio file.
AAC is common in MP4, but the extension alone does not guarantee it. After the file opens, the Audio track control shows the codec, reported channel count and language when available. If the MP4 has several audio tracks, the control starts on the first one. Choose the intended track before making cuts or volume changes, because changing it later clears those edits after a confirmation.
Measured: an HEVC MP4 behaves like an H.264 one
The test file was five minutes of 1280x720 HEVC picture with 48 kHz stereo AAC at 128 kbps, 72,300,537 bytes. File check showed MP4, 1 audio track, 1 video track, AAC 128 kbps, and the editor was ready in 0.7 s.
| Output | Time | Saved size |
|---|---|---|
| MP3, 192 kbps | 3.0 s | 7,200,744 bytes |
| M4A, unchanged copy | 0.2 s | 4,869,993 bytes |
| WAV | 2.4 s | 57,600,044 bytes |
Those rates match the H.264 recordings in the screen recording test: the picture is never decoded, so its codec changes neither the time nor the output. What matters is the audio track the panel lists.
The picture can fail to preview while audio still works
The Play control asks the browser's built-in media element to open the original MP4. The waveform and conversion use the selected audio track through a separate path. If Play is unavailable but the waveform reaches Ready, the audio track has still passed the checks needed for conversion. Conversely, seeing the picture or hearing native playback does not prove that the selected track can be decoded by the extraction path.
The saved MP3, M4A or WAV contains the selected audio only. Video and subtitle tracks are not copied into an audio output. Cuts are made against the audio timeline shown in the waveform.
MOV and iPhone video: the audio codec decides, not HEVC
MOV is the QuickTime movie container. It can hold separate picture and sound tracks with their own codecs. An iPhone Camera recording commonly uses H.264 or HEVC for the picture and AAC for sound, but AAC is not guaranteed in every MOV made by a camera or editor.
PrivateMP3 does not decode H.264 or HEVC frames when making an audio file. Once the MOV structure is open, it selects an audio track and ignores the pictures. That means extraction can work even when the browser cannot preview the video. It does not mean every HEVC MOV will work: damaged container metadata or an unsupported audio codec can still stop the job.
AAC decides whether M4A can be copied
An unchanged copy into M4A needs a AAC, AC-3 or E-AC-3 track; the conditions are the same for every container.
This distinction matters with camera files that use PCM, ALAC or another audio codec. The video may be perfectly ordinary while the M4A copy path is unavailable. Check the audio track, not the H.264 or HEVC label, when deciding which output preserves the source.
The iPhone file picker can be the slow part
Choosing a video from Photos may make iOS prepare or convert a browser-compatible copy before the file input returns. PrivateMP3 has not received the file yet, so it cannot show progress for that wait. Retrieving an original from cloud storage adds another possible delay.
If the video already exists in Files, choose Browse and select it there. That avoids the Photos export step, although a cloud-only Files item still has to download. Once the filename appears in the converter, the hand-off is complete and PrivateMP3 starts scanning the local file.
If the MOV has pictures but no usable audio
Play the original and confirm that it contains sound. If several audio tracks exist, choose the intended language or mix from the track picker. A recording interrupted before it was finalized may lack the metadata needed to locate its tracks. If the container opens but the waveform does not appear, the browser may lack a decoder for the audio codec. A current desktop browser is a useful second check because it removes the iPhone Photos hand-off from the test.
WebM: Opus or Vorbis, never a copy
WebM is a restricted form of the Matroska container. Its audio track can be Opus or Vorbis, and a video WebM can pair that audio with VP8 or VP9 pictures. Browser recorders often choose Opus; older files may use Vorbis. A player needs support for both the container and the codec inside it. Recognizing the .webm extension does not prove that a particular audio track can be decoded.
There is no packet-copy route to MP3
Opus and Vorbis packets cannot be placed unchanged into an MP3 file. PrivateMP3 decodes them and writes a new 192 kbps MP3. Because both WebM audio codecs are lossy, that is another lossy generation. A larger MP3 bitrate would not restore information removed when the WebM was recorded.
The other output buttons do different work, but neither preserves the original packets. M4A means a new AAC encode because a standards-compliant WebM does not contain AAC. WAV means decoded 16-bit PCM: useful for editing, much larger, and no better than the decoded source. Keeping the original WebM is the only way to keep its Opus or Vorbis bitstream untouched.
Voice recording and music
Speech can usually use a lower bitrate than music, especially for one mono voice. PrivateMP3 does not offer that choice; its MP3 output is fixed at 192 kbps. That setting favors compatibility, not the smallest possible long-form voice file. Music is more likely to expose damage from another lossy pass. If the destination accepts WebM, keep the original. If the music will be edited, use WAV as the working file and make one delivery encode after editing.
Measured: Vorbis and Opus WebM
| File, 48 kHz stereo | File check said | MP3 | M4A (new AAC) | WAV |
|---|---|---|---|---|
| 3:00, VP8 + Vorbis, 24,754,081 bytes | Vorbis, 79 kbps | 1.7 s, 4,321,320 bytes | 1.2 s, 4,356,280 bytes | 1.6 s, 34,560,620 bytes |
| 10:00, VP8 + Opus, 81,204,433 bytes | Opus, 69 kbps | 10.2 s, 14,401,320 bytes | 4.5 s, 14,518,706 bytes | 5.7 s, 115,201,580 bytes |
Both codecs decoded in both browsers. The Opus audio was asked for at 96 kbps and measured 69 kbps, because Opus lowers its rate when the signal is simple. The 192 kbps MP3 made from it is 2.8 times the size of the source audio and carries nothing the 69 kbps stream did not. M4A from WebM is never a copy: the panel says the source is Opus or Vorbis, not a codec it can copy, before Convert is pressed.
Was the recording finalized?
Play the original through to the end before converting. A recorder that crashed, a tab that closed before the recording stopped, or a partially copied file may be missing audio packets or duration information. No converter can recreate packets that were never saved. Also check that the capture contains an audio track; microphone or system audio may not have been enabled. A file merely renamed .webm may contain a codec outside the WebM rules, and a browser may parse the container while lacking a decoder for its audio.
M4A files: AAC or ALAC
M4A is an audio container in the MPEG-4 family, not a codec name. AAC is the common payload and is lossy: some information was discarded when the recording was encoded. An M4A can also contain ALAC, the Apple Lossless Audio Codec. That difference changes the quality history of the conversion. AAC to MP3 is a second lossy encode. ALAC to MP3 would start from lossless audio and still end lossy, but in current Chrome and Edge it does not start at all: neither browser can decode ALAC, as the test below shows.
Decide whether you need MP3 at all
If the original opens in the software or device you plan to use, keep it. Converting AAC to MP3 cannot improve it, and raising the MP3 bitrate cannot restore what AAC removed. MP3 is useful when the destination explicitly requires .mp3 or rejects M4A.
What each output does with an M4A
| Source and choice | Actual processing |
|---|---|
| AAC → MP3 | Decode AAC and encode a new 192 kbps MP3. |
| ALAC → MP3 or WAV | Needs an ALAC decoder in the browser. Chrome 154 and Edge 154 have none, so there is no waveform and Convert stays off. |
| Unedited AAC → M4A | Copy the AAC packets without re-encoding. The converter shows a copy note. |
| Edited AAC → M4A | Decode and encode new AAC. ALAC is neither copied nor, in those browsers, decodable. |
| Decodable audio → WAV | Decode to a larger 16-bit PCM working file. |
A cut or volume change rules out packet copy because it changes the timeline or sample values. Simply choosing M4A is therefore not a promise of lossless output.
Measured: one AAC M4A and one ALAC M4A
| File, 3:00, 44.1 kHz stereo | File check said | MP3 | M4A | WAV |
|---|---|---|---|---|
| AAC, 4,964,185 bytes | AAC, 219 kbps | 1.5 s, 4,321,018 bytes | 0.09 s, 4,965,348 bytes, unchanged copy | 1.2 s, 31,752,044 bytes |
| ALAC, 24,316,106 bytes | ALAC, 1080 kbps, no decoder in this browser | Convert disabled in both browsers | ||
The AAC file was requested from FFmpeg at 256 kbps; the panel measured 219 kbps from its packets, and the unchanged copy kept exactly that rate. The MP3 came out smaller than the source because 192 kbps is below 219 kbps. That is a lower data rate after a second lossy encode, not a cleaner file.
The ALAC file opened and listed its track, but WebCodecs reported no support for ALAC in either browser, so nothing could be decoded. On a Mac, Apple's Music app can make a converted copy: in Music, choose Settings, Files, Import Settings, pick the format in Import Using, then select the song and choose File > Convert > Create [format] Version. Apple notes that the original and the converted version both stay in the library. (Apple: convert music file formats in Music on Mac) A WAV or AAC version made that way opens in the converter normally.
When there is no waveform
Check that the original still plays in the application that created it. The browser may understand the M4A container but lack a decoder for the codec inside. MP3 and WAV then cannot run. A complete, unedited AAC track may still use the decoder-free M4A copy path; ALAC cannot. A damaged file or misleading extension needs to be re-exported from its original application. Keep the source until you have listened through the result.
When the waveform never appears or Convert stays grey
Use the message and controls on the page to find the stage that stopped, then choose the next action that still works.
Read the screen as a sequence
PrivateMP3 identifies the container, lists its tracks, describes the audio tracks, calculates duration, and then decodes samples for the waveform. Conversion is a separate job. The last visible change shows which earlier checks passed.
| What you see | What passed | Where it stopped | What to check |
|---|---|---|---|
That file type cannot be opened here | The local file reached the worker. | Container identification. | Open the original in its creating application; renaming it is not a repair. |
The file opened, but its track list could not be read | The container was identified. | Track enumeration. | Keep the source and make a fresh export in the creating application. |
This file has no audio track | The container and track list opened. | No track was classified as audio. | Confirm that the source export included sound. |
| The editor shows 0:00 and Convert is disabled | An audio track was found. | No usable duration was obtained. | Check whether the creating application reports a usable duration; re-export if it does not. |
| No waveform and decoder text | The selected audio track was found. | Its decoder support check. | Try another listed track, or an unchanged copy into M4A if the track is AAC, AC-3 or E-AC-3. |
| The waveform shows a duration or decoding error | The scan stage was reached. | Duration lookup or sample decoding. | Follow the specific message shown for that stage. |
| Convert stays disabled while the waveform draws | The file and selected track opened. | The decoder check is still running. | Wait for Ready; if this track has no usable decoder and is AAC, AC-3 or E-AC-3, an unchanged copy into M4A is the exception. |
1. Separate a container failure from a missing audio track
The unsupported-file message comes from failure to identify the container. The no-audio message is later: the container opened and its tracks were listed, but none was audio. Changing MP3 or M4A output cannot help when no source audio track exists.
If the creating application cannot play the original with sound, keep the file and use that application's recovery or export function. If it can, make a new local export with audio enabled. See the checks above for a WebM recording or an iPhone or QuickTime MOV.
2. Check the selected track and the duration
One audio track is selected automatically. With several, an Audio track selector appears and starts on the first. An option can show the track name, language, codec, mono/stereo or channel count, purpose flags such as commentary, and no decoder
. Those labels depend on metadata in the file; the File check panel under the editor lists the same details for every track along with sample rate and measured bitrate.
Changing the track after making a selection or edit opens a warning. Continuing clears cuts, volume edits, the selected range and undo history before starting the new track's waveform; cancelling keeps the current work. Choose the track before editing when possible. For language, commentary or separate mixes in Matroska, use the MKV audio track guide.
Duration is checked separately. If its calculation fails, the editor can remain open at zero. Convert then stays disabled because the output is not longer than 0.01 seconds. There is no special unfinished container
message for this state.
3. A blank waveform and a disabled button are different tests
The Play control is a third, independent check. PrivateMP3 attaches the local file to the browser's built-in audio element. If that preview cannot attach, Play is disabled with the tooltip Preview is not available for this container
. The worker may still read and decode the selected audio track, so Play alone does not diagnose the container or decoder.
The waveform starts only if the selected track reports that it can be decoded. Decoded channels are combined into one visual trace, and peaks appear progressively. If duration lookup or audio decoding fails, the waveform area names that stage instead of showing one generic error.
A re-encoded MP3, WAV or M4A waits for the Ready status. Convert requires an open file, an output longer than 0.01 seconds, no conversion already running and a successful waveform scan. If the selected track is AAC, AC-3 or E-AC-3 and has no usable decoder, an unchanged copy into M4A is the exception because it does not decode samples.
The page first checks whether AudioDecoder exists; the waveform then checks the exact selected track. The WebCodecs specification allows support to differ by codec configuration. If that track cannot be decoded, the page names the decoder problem and keeps Convert disabled unless the unchanged copy route for AAC, AC-3 or E-AC-3 is available.
4. Use the unchanged copy only when the page says it will copy
The decoder-free route requires M4A output, an AAC, AC-3 or E-AC-3 track, and the complete unchanged timeline: no cuts, selected-output clip, fade or volume edits. PrivateMP3 confirms the route with Copied without re-encoding — same audio, no quality loss.
Without AudioDecoder, MP3 and WAV are disabled. Convert is available only when those copy conditions and a positive duration are both present. ALAC, Opus, Vorbis, DTS, an unidentified codec or an edited track cannot use the exception. Chrome and Edge have no decoder of their own for AC-3 or E-AC-3, so the page loads a separate one when a file has such a track, and MP3 and WAV then work. If that load fails, the page selects M4A itself and says the track will be copied unchanged; the MKV audio tracks guide has the tests. The AAC and ALAC table above lists each output path.
5. What an incomplete file can look like here
There is no single damaged-file detector or repair step. Inconsistent structure can fail during container identification, track enumeration, duration calculation, waveform decoding, packet reading or output finalization. PrivateMP3 names these stages separately, but a stage-specific message still cannot prove why the source failed there.
The MediaRecorder standard describes normal completion with stop and a dataavailable event. PrivateMP3 cannot tell whether a particular recorder finished its own writing process. If the recorder, camera or tab ended unexpectedly, check the original first; the page has not diagnosed it as incomplete.
A recovery order that preserves the original
- Keep the source file; do not replace it with a failed export.
- Play it in the creating application and confirm sound through the end.
- Note the exact symptom: unsupported container, no audio, 0:00, decoder text or scan failure.
- Choose the intended audio track before making edits.
- If the M4A copy note appears, try that unchanged output.
- For 0:00 or a scan failure, make a fresh export in the source application. Changing only the extension keeps the same structure.
- Use a second current browser only to compare decoder support. A different result suggests a browser difference; an identical result does not prove damage.