Audio formats explained for editing and sharing

How to use this workflow

1. Check the destination

   Find out which formats the receiving editor, device or service accepts. There is no single format that is ideal for every destination.

2. Keep a source for further edits

   Retain the best available original. Use PCM or a suitable lossless format between processing steps when the next tool supports it.

3. Make a delivery copy

   Use the converter to choose a supported format and settings. Listen to the exported file and keep the editable source separately.

What the converter’s choices mean

A codec describes how sound data is encoded. A container holds encoded data and associated information. M4A can hold AAC or ALAC; its extension alone does not tell you whether the audio is lossy. The choices below describe this converter’s outputs.

Swipe or scroll sideways to read all columns.

Bitrate, sample rate and bit depth are different

Bitrate is data per second, commonly shown in kbps for compressed outputs. Sample rate is samples per second per channel, shown in Hz or kHz. PCM bit depth describes the number of bits used for each sample value. They are not interchangeable quality scores.

For uncompressed stereo PCM, 44,100 samples/second × 16 bits × 2 channels equals 1,411,200 bits/second. A sixty-second example contains 10,584,000 bytes of sample data before container overhead. This is arithmetic for those settings, not a claim about every WAV file.

Input support is not determined by the extension alone

Most waveform tools rely on the browser’s audio decoder, while the converter and extractor use a locally loaded FFmpeg engine. A file that opens in one path may fail in another because codec support differs.

Decoding in Web Audio can resample to the audio context’s sample rate. Do not assume every processing path preserves the original encoded sample rate or metadata. If tags, artwork or exact stream preservation matter, verify them in the exported file.

Turn a destination requirement into a format choice

Ask the receiver for a format, codec and sample rate before converting a full recording. A request for an M4A file leaves an important question unanswered: this converter can put either AAC or ALAC in that container. The table below uses hypothetical receiving requirements, not claims that every editor or service requires these settings.

Select the named Output format and then check the controls that appear. MP3 and AAC offer 44.1 or 48 kHz. WAV, FLAC, AIFF and ALAC expose 16- or 24-bit depth; FLAC also offers compression levels 0–8. A FLAC compression setting changes the encoding tradeoff, not the decoded sample values.

Swipe or scroll sideways to read all columns.

Verify one delivery before processing the rest

For a new destination, convert one representative file first. Open the download in the actual receiving application, play its beginning and ending, and inspect the displayed codec, sample rate, duration and channel count where available. A successful download alone does not establish that the destination can use it.

Record the selected settings beside your source filename and inspect any required tags or artwork separately. The converter selects the first audio stream; it is not an archive of every stream in a media file. Its queue holds three files and uses the current conversion settings for queued jobs. Group files by the requirement they share, or process different requirements separately. Keep the original until the receiver has accepted the output.

Separate a converter engine failure from a source failure

Use the exact status message to choose the next test. “FFmpeg could not load” occurs before the queued conversion can run. An individual queue item marked Failed, or a nonzero “FFmpeg exited” message, occurs during processing. A processing exit alone does not prove that the codec is unsupported: a damaged file or another processing problem can produce the same broad symptom.

Keep the failed original. Remove completed or failed queue items as needed, then add the small Original phrase WAV available in the reverb guide and convert it with the same output settings. If that file converts while yours fails, investigate the source. If both stop at engine loading, investigate the runtime download, connection and browser environment. This is a diagnostic procedure, not a report of deliberately induced failures.

Swipe or scroll sideways to read all columns.

Keep useful evidence when a source still fails

Check that the source plays in a trusted local player, then record its codec, duration and channel count if available. Playback in that player establishes only that its decoder can read the file. It does not guarantee the browser decoder or this FFmpeg build supports the same input.

If another editor can decode it, create a short PCM WAV copy and try that copy here. Renaming the original extension is not a conversion. Avoid sharing a private recording merely to demonstrate the error; an exact error message and a non-sensitive sample can be enough for a technical handoff. Keep the original until a converted download opens and includes the material you need.

Estimate decoded memory from duration, not the extension

A Web Audio buffer stores one 32-bit floating-point value per sample per channel. At a stated 48 kHz stereo decoding rate, its sample storage is duration in seconds × 48000 × 2 × 4 bytes. A ten-minute recording therefore needs about 230.4 MB for one decoded sample buffer, even if its compressed input is much smaller. These are calculated sample-storage figures, not a measurement of the browser’s total memory.

This site’s waveform decoding path also copies channel arrays, and processing can allocate output buffers. Engine code, waveforms, temporary objects and a longer slowed output add other allocations. The table is a lower-level budget illustration, not a safe-device limit. The FFmpeg converter uses a different execution path, so these exact AudioBuffer allocations should not be assumed for every tool.

Swipe or scroll sideways to read all columns.

Keep the task while reducing the amount decoded

If a long source fails in a waveform tool, first save any completed output and close unused processing tabs. If your goal is one verse or a short diagnostic sample, create a shorter PCM excerpt in a capable external editor and load that excerpt here. Keep the complete original separately. Merely renaming MP3 as WAV changes neither duration nor decoded sample requirements.

Test the smallest passage that still contains the needed lead-in, event and ending. If you must process the full recording and it still exceeds the device’s resources, use a local editor designed for that workload. A file below the 250 MB local-file limit can still need substantial working memory; lowering the source MP3 bitrate does not shorten the decoded recording.

Diagnose silence without treating a waveform as audible proof

A waveform shows sample activity, while audible playback also depends on the selected passage and your output path. First move to a section with visible activity and start playback using the page control. Observe whether the playback position advances. Then check the browser tab mute state, system output device and speaker or headphone level. Use a comfortable listening level; do not turn everything to maximum to hunt for a signal.

Test a short recording you know contains sound through the same browser and output device. If that is also silent, investigate the playback path before converting the original. If it plays, open the problem source in a trusted local player and compare the same passage. A quiet beginning, a source with sound on only one channel or an unsuitable selected output can produce different next steps.

When reporting the problem, record whether the position moved, which passage you tried and whether the known-good file played. A moving cursor alone does not prove sound reached the speaker, and a waveform alone does not prove the audible content is the material you intended. Keep the original; renaming an extension or repeatedly re-encoding is not a diagnosis.

Swipe or scroll sideways to read all columns.

Use PCM arithmetic to spot an incomplete WAV delivery

For an uncompressed PCM WAV, sample rate, channel count, bit depth and duration determine the sample-data size. A hypothetical 90-second, 48 kHz, stereo, 24-bit file contains a calculated 90 × 48000 × 2 × 3 = 25,920,000 bytes of samples, or 25.92 decimal MB. Container headers and metadata add other bytes. Use the complete calculation rather than comparing filenames or rounded download sizes.

If a delivery expected to contain that full passage is only around 8.64 MB, check duration and format before sending it. Thirty seconds at those PCM settings accounts for 8.64 MB of sample data, but that size alone does not prove truncation: a compressed codec or fewer channels could also change the result. Open the file in the receiving editor and inspect its actual codec, duration, sample rate, depth and channels, then listen to the intended final words.

Use this receipt only when the stream is confirmed to be uncompressed PCM. It does not predict the size of every WAV container, FLAC, MP3 or a video containing audio. Record whether a size is decimal MB or binary MiB when comparing displays; 25,920,000 bytes is about 24.72 MiB. If the ending is absent, return to the original source and export the required passage again. A larger destination file cannot recover material that was omitted from the loaded clip.

Swipe or scroll sideways to read all columns.

Check which audio stream the converter will keep

A container can hold more than one audio stream, such as two language versions or a main mix and commentary. This converter selects the first audio stream; it does not provide a stream-selection menu. A successful output therefore establishes that one selected stream was converted, not that every soundtrack or alternate language was preserved. The first stream can also differ from the default track chosen by a separate media player.

Inspect the source in an external application that lists streams before conversion when alternate audio matters. For an invented two-stream file, note “audio stream 1: commentary; audio stream 2: main mix.” Converting that source here would target the first audio stream, so do not promise delivery of the main mix merely because it was the track you heard in another player. Extract the intended stream in a capable external editor and use that saved audio file as the conversion input.

After download, listen to an unmistakable spoken line or musical cue near the start and another later in the file. Confirm language, content and channel requirements with the receiver. Keep the original container if alternate streams are part of the archive. If a file contains no audio stream, changing the output extension or increasing bitrate cannot create the missing soundtrack.

Swipe or scroll sideways to read all columns.

Common questions

Continue with your audio

• Compare output settings →
• Understand repeated encoding →
• Create an MP3 copy →
• Prepare PCM WAV →
• Check for an audio stream before extraction →
• Find out whether an M4A file contains AAC or ALAC →
• Prepare audio at the sample rate your video project needs →