Workflow guide | 9 min read
How to Batch Convert the Sample Rate of Multiple Audio Files
A recording day can leave you with dozens or hundreds of WAV files at a sample rate that does not match the edit or the delivery spec. Converting them one by one is repetitive, and every manual pass is another chance to pick the wrong setting on one file.
Batch conversion applies one set of settings to a whole queue while keeping every file separate. This guide covers when it makes sense, what happens to channels and production metadata, and how to run it.
Batch convert audio files →What is batch sample rate conversion?
The sample rate is the number of samples that represent each second of audio: 48,000 per second at 48 kHz, 96,000 at 96 kHz. Sample rate conversion rebuilds the audio on a new grid of samples so the same sound, at the same duration and pitch, is described at a different rate.
Batch sample rate conversion applies the selected target to several files in one run. Each source file produces its own output file. It is not the same as:
- merging files into one longer file,
- combining separate tracks into a polyphonic WAV,
- downmixing channels, for example 5.1 to stereo.
Ten files in, ten files out, each with the same channels it started with.
When should you batch convert audio sample rates?
| Situation | Typical target |
|---|---|
| 96 kHz field recordings going into a video edit | 48 kHz / 24-bit WAV |
| A session with mixed 44.1, 48 and 96 kHz sources | One common rate for the whole session |
| Sound library preparation | The rate your library or client standardises on |
| Delivery to a post house or broadcaster | The rate and bit depth in the delivery spec |
| Large multi-day recording sessions | The project's working rate |
| Working copies made from a high-rate archive | The workflow's rate, originals kept untouched |
Do not resample for the sake of it. If files already match the working rate or the delivery specification, leave them as they are. Each conversion is an extra processing step, and converting a 48 kHz file up to 96 kHz adds size without adding detail. Keep your originals and convert once, from the original.
Batch converting 96 kHz audio to 48 kHz
96 kHz is common on set and in sound design. The extra headroom above the audible range is useful when a sound is going to be pitched down or heavily processed, and many recordists capture effects and ambiences at that rate by default.
48 kHz is the standard working rate for most video, film and television post-production. Editing software can often play a 96 kHz file in a 48 kHz timeline, but it does so by converting on the fly, which costs processing and can differ from one application to the next. Preparing the files at 48 kHz before the edit gives everyone the same material. For background, see why 48 kHz is the video standard and 48 kHz vs 96 kHz.
Halving the rate is not just a matter of dropping every other sample. Content above the new Nyquist limit (24 kHz for a 48 kHz file) has to be filtered out first; otherwise it folds back into the audible range as aliasing. Sample rate converters are not all equivalent in how well they do this.
Audio Tool Kit uses a high-quality resampling process designed to avoid aliasing during conversion, the same one used by the single-file Sample Rate Converter. The batch tool changes the workflow, not the conversion quality. More detail on the process is in sample rate conversion explained.
How to batch convert multiple audio files
The steps below follow the Batch Sample Rate Converter as it works today.
1. Add your audio files
Drop several files at once, or add more later. Each file becomes its own line in the queue, with its own status and progress, and each one is converted to its own output file.
2. Choose the output settings
Pick a target sample rate (48 kHz for most video work, or Source to keep each file's own rate), a bit depth for WAV or AIFF (24-bit is the usual post-production choice), and the output format. The same settings apply to every file in the queue.
3. Set the output filename
The original base filename is always kept. An editable suffix is added, filled in automatically from your settings. You can type your own suffix instead or clear it entirely.
Original ambience.wav Default, 48 kHz / 24-bit ambience_48k_24bit.wav Custom suffix ambience_MASTER.wav 4. Choose an output folder
In browsers that support it (Chrome and Edge on desktop today), choose an output folder and each file is written there as soon as it is ready. Existing files are never overwritten. Without a folder, each finished file is downloaded by your browser as soon as it is ready; the browser may ask once to allow multiple downloads.
5. Convert the batch
Files are processed one after another, not all at once. Each completed file is saved before the next one starts, with progress shown per file. On a long queue of large recordings this keeps memory use to one file at a time and means finished files are already on disk if you stop halfway.
What happens to multichannel WAV files?
Changing the sample rate should never change the channel structure on the side. For WAV output, Audio Tool Kit keeps the channel count and the channel order:
- Mono → mono
- Stereo → stereo
- 5-channel → 5-channel
- 5.0 → 5.0 when the source declares a known, compatible speaker layout
- 5.1 → 5.1 when the source declares a known, compatible speaker layout
- 8-channel → 8-channel
Channel count and speaker layout are different things. A channel count says how many tracks the file holds. A speaker layout (the channel mask in a WAVE_FORMAT_EXTENSIBLE header) says which speaker each track feeds. When a WAV carries an explicit layout that matches its channel count, that layout is written to the output. When the source only states a number of channels, as many polyphonic recorder files do, no layout is invented: the channels are kept in order and left unassigned. For more on these files, see PolyWAV vs mono WAV.
Production example
A 5.0 ambience recorded at 96 kHz holds five channels declared as L, R, C, Ls, Rs. Converted to 48 kHz WAV, the output still holds five channels declared as L, R, C, Ls, Rs, in the same order, with the same duration.
96 kHz · L R C Ls Rs → 48 kHz · L R C Ls Rs
The conversion changes the sample rate, not the creative channel structure. If you do need a stereo version, make it as a separate, deliberate step with the Surround to Stereo tool.
What happens to BWF, iXML and timecode metadata?
Production WAV files carry more than audio. The BEXT chunk holds the originator, the recording date and the time reference that places the file on the timecode line. iXML carries the scene, take and track names your editor and sound editor rely on. Lose them and a conform or an autosync can fall apart. See BWF metadata explained for the details.
For WAV to WAV conversion, supported production metadata is preserved, including:
- the BEXT chunk (description, originator, date),
- iXML,
- the time reference used for timecode,
- track information, such as track names,
- cue marker positions,
- the existing recorder or DAW coding history.
Several of these are stored as sample counts, so they must be recalculated when the rate changes. A marker at exactly one second sits at sample 96,000 in a 96 kHz file and at sample 48,000 after conversion to 48 kHz: the number changes so the position in time does not. The same applies to the time reference, so the file still lands at the right timecode.
The sample rate and bit depth information in the metadata is updated where appropriate, and a line describing the conversion is added to the coding history after the existing recorder entries. Not every proprietary chunk from every recorder or DAW can be guaranteed; the supported chunks above are the ones handled deliberately.
Should you export WAV, AIFF or MP3?
| Format | Use it when | Keep in mind |
|---|---|---|
| WAV | Production metadata and multichannel structure matter. | The format that carries the BWF, iXML and timecode described above. |
| AIFF | A destination asks for AIFF. | Uncompressed and multichannel, but the WAV production metadata is not carried over. |
| MP3 | Listening copies, previews and reference delivery. | Compressed, mono or stereo only, and has no PCM bit depth. |
MP3 can only hold mono or stereo. A file with more than two channels is refused for MP3 instead of being silently downmixed, and the rest of the queue carries on. Choose WAV or AIFF to keep every channel. The differences between the formats are covered in WAV file format explained.
Batch conversion vs converting files individually
Queuing files does not change the quality of the conversion: the same resampling and metadata handling run for each file. The difference is workflow.
Batch
- one set of settings
- many source files
- sequential processing
- automatic output naming
- output-folder workflow
- progress for each file
Individual
Better when each file needs different settings, or when you want to listen to and inspect a file before and after converting it. The single-file Sample Rate Converter is built for that.
Are the audio files uploaded?
Your audio files are processed locally in your browser. For large production recordings this also means there is no upload to wait for before the conversion starts, and no download of converted files from a server afterwards.
Working with large batches
- Choose an output folder when your browser offers it, so each file is written as soon as it is ready.
- Files are processed one at a time, so a long queue does not load every file into memory at once.
- Completed files are saved individually; if you stop the batch, finished files are already saved.
- Keep the browser tab open until the queue has finished.
- Long multichannel or high sample rate files need much more memory than short stereo files. On devices with limited memory, split very large jobs into smaller batches.
Tip Spot check one converted file per batch in the Audio Inspector: sample rate, bit depth, channel count and duration should read as expected.
Convert multiple audio files →Frequently asked questions
- Can I batch convert 96 kHz WAV files to 48 kHz?
- Yes. Add the 96 kHz files to the queue, choose 48 kHz as the target sample rate, pick a bit depth and a format, and start the batch. Each file is filtered and resampled, then saved as its own 48 kHz file.
- Does changing sample rate change the number of channels?
- No. For WAV output, a mono file stays mono, a stereo file stays stereo and a 5 or 8 channel file keeps all of its channels in the same order. Sample rate conversion does not merge, split or downmix channels.
- Can I batch convert 5.1 WAV files?
- Yes, as WAV or AIFF. All six channels are kept in their original order. When the source WAV declares a compatible speaker layout, that layout is written to the WAV output. MP3 cannot hold more than two channels, so 5.1 files are refused for MP3 rather than folded down to stereo.
- Are BWF and iXML metadata preserved?
- For WAV to WAV conversion, supported production metadata is kept: the BEXT chunk, iXML, the time reference, track information, cue markers and the existing coding history. Sample-based positions are recalculated for the new rate. Unknown proprietary chunks from specific recorders or DAWs are not guaranteed to survive.
- Can I customize the output filenames?
- Yes. The original base filename is always kept and an editable suffix is added. By default the suffix describes the output, for example ambience_48k_24bit.wav. You can replace it with your own text, such as _MASTER, or clear it.
- Can I batch convert files to MP3?
- Yes, for mono and stereo sources. MP3 has no PCM bit depth, so the automatic suffix only carries the sample rate, for example ambience_48k.mp3. Files with more than two channels are refused for MP3 so their channel structure is never changed silently.
- Are my audio files uploaded?
- No. Your audio files are processed locally in your browser.
- What happens if one file in the batch fails?
- The failed file is marked with its error and the queue moves on to the next file. Failed or cancelled files can be retried once the batch has finished.