How to concatenate audio files without uploading — browser tool
- Step 1Open the tool (optionally go offline after) — Load the page. Once it's loaded, you can disconnect from the network if you want to be certain nothing transmits — the in-browser FFmpeg engine handles everything locally. Drop 2+ files into the dropzone.
- Step 2Add your files in order — Accepts
MP3, WAV, FLAC, M4A, OGG, Opusand more. Files concatenate top-to-bottom in list order. Each shows duration and channel info so you can confirm the sequence of, say, multi-part testimony. - Step 3Leave crossfade at 0 — For an exact concatenation of statements/segments, keep
crossfadeS: 0so no audio overlaps — important when timing or completeness of a record matters. - Step 4Choose the output format — WAV or FLAC for a lossless confidential master; MP3 (192 kbps) for a smaller copy to share under controlled conditions. M4A for Apple workflows.
- Step 5Run the concatenation locally — FFmpeg-WASM decodes, resamples to 48 kHz stereo, joins, and encodes — entirely on your machine. No progress data or file content is sent anywhere.
- Step 6Download merged.<ext> — The combined file saves directly to your device. Verify the total duration equals the sum of the parts (a hard concat adds no time). The only server interaction is an optional anonymous run counter for signed-in dashboard stats.
Where the audio goes (privacy model)
Concatenation is fully client-side. This table maps each step to whether any data leaves your device.
| Step | Runs where | Leaves device? |
|---|---|---|
| File decode + resample | Browser (FFmpeg-WASM) | No |
| Concat + encode | Browser (FFmpeg-WASM) | No |
| Download merged file | Browser -> local disk | No |
| Anonymous run counter (signed-in) | Optional ping, no file content | Count only, no audio |
Output format for confidential workflows
Pick lossless for a master record, lossy for a controlled distributable.
| Format | Encoder | Use for |
|---|---|---|
| WAV | pcm_s16le (lossless) | Master record / evidentiary copy (largest) |
| FLAC | flac (lossless) | Lossless archive at smaller size than WAV |
| MP3 | libmp3lame @ 192k | Smaller copy to share under controlled access |
| M4A | aac @ 192k | Apple ecosystem distribution |
Tier limits (audio family)
Real caps. Long depositions/calls usually need Pro or higher for the duration and batch headroom.
| Tier | Max file size | Max duration / file | Files per batch |
|---|---|---|---|
| Free | 50 MB | 30 min | 1 |
| Pro | 200 MB | 120 min | 10 |
| Pro-Media | 100 GB | Unlimited | 100 |
| Developer | 100 GB | Unlimited | Unlimited |
Cookbook
Real confidential-merge scenarios. All processing is local — no audio is uploaded in any of these.
Concatenate multi-part deposition audio
A deposition recorded across several files (breaks, lunch) joined into one continuous record, kept lossless.
Inputs (in order): depo-am.wav 48000 Hz 2ch 1:52:00 depo-pm-1.wav 48000 Hz 2ch 0:58:00 depo-pm-2.wav 48000 Hz 2ch 0:41:00 Options: crossfade 0, format WAV Output: merged.wav 3:31:00 (lossless, exact concat) (requires Pro-Media for the 100GB / unlimited duration)
Join phone-statement segments without uploading
Witness statements recorded on a phone (M4A) joined into one file, all on-device.
Inputs: statement-1.m4a 44100 Hz 1ch 6:20 statement-2.m4a 44100 Hz 1ch 4:05 Options: crossfade 0, format WAV Output: merged.wav 48000 Hz 2ch 10:25 (mono upmixed to stereo)
Combine then redact spoken PII
Join the segments first, then remove named PII from the combined recording before sharing.
Step 1 concatenate-audio-files-no-upload part1.wav + part2.wav -> merged.wav Step 2 pii-redactor merged.wav -> redact name/address ranges Both steps run locally; nothing uploads.
Offline merge to prove no transmission
For maximum assurance, load the page, disconnect the network, then merge.
1. Open the concatenate tool, let it load (caches WASM) 2. Turn off Wi-Fi / disconnect ethernet 3. Drop files, run merge -> merged.wav downloads 4. The job completes offline = no upload occurred
Make a lossless master + a shareable copy
Concatenate once to FLAC for the archive, once to MP3 for a controlled-access copy.
Run A format FLAC -> merged.flac (lossless master) Run B format MP3 -> merged.mp3 (192k, smaller copy) Keep the FLAC as the record; distribute the MP3.
Edge cases and what actually happens
Worried audio might be uploaded
By designIt isn't. The concatenation runs in your browser on FFmpeg-WASM; file content never reaches a server. You can confirm by disconnecting the network after the page loads — the merge still completes. The only optional server contact is an anonymous run counter (count only, no audio) for signed-in dashboard stats.
Only one file provided
RejectedConcatenation needs 2+ files (Audio merger needs at least 2 files.). To split a single confidential recording into parts, use audio-splitter; to redact PII from one file, use pii-redactor.
Concatenation re-encodes the audio
By designThe merge always decodes and re-encodes (there is no stream copy), and normalises to 48 kHz stereo. For an evidentiary master you want lossless, so output WAV or FLAC — MP3/M4A apply a lossy re-encode. The join itself is exact (no audio added or dropped).
Mono statement becomes stereo
ExpectedMono inputs (typical of phone recordings) are upmixed to stereo on the common timeline, so the output is stereo. The content is unchanged; both channels carry the same audio. There's no mono output option.
Long deposition exceeds Free caps
BlockedFree caps each file at 30 minutes / 50 MB and allows only 1 file per batch. Multi-hour, multi-part recordings need Pro (120 min / 200 MB / 10 files) or Pro-Media/Developer for unlimited duration and 100 GB files.
Free tier batch is 1 file
Tier limitConcatenation needs at least 2 files, but the Free audio batch limit is 1. Practical merging starts at Pro (10 files). The local-only privacy model is identical at every tier.
Mixed formats across statements
By designA WAV recording and an M4A phone statement concatenate cleanly because both are normalised to 48 kHz stereo first. You don't have to convert them to a common format beforehand.
Timestamps / metadata not preserved
DroppedRe-encoding produces a fresh stream, so embedded recording timestamps, ID3 tags, or device metadata from the inputs are not carried into the merged file. If chain-of-custody metadata matters, document it separately; add tags with id3-editor afterward.
Large lossless output uses lots of memory
ExpectedA multi-hour WAV concat decodes everything to PCM in browser memory and writes a very large file (an hour of 48 kHz stereo WAV is ~660 MB). Use FLAC to cut output size, and ensure enough free RAM for big jobs.
Want a tamper-evident record
Not providedThe merger joins audio but does not sign, hash, or watermark the output for tamper-evidence. If you need an integrity guarantee for evidentiary use, hash the merged file yourself (e.g. SHA-256) and record the hash through your own chain-of-custody process.
Frequently asked questions
Is my audio uploaded to a server?
No. The entire concatenation runs in your browser on FFmpeg 8.1 (WebAssembly). File content is never transmitted — there is no server-side processing path for your audio. You can verify this by disconnecting the network after the page loads; the merge still completes. This is why it's suitable for privileged, regulated, or embargoed recordings.
Can I use this fully offline?
Yes, after the initial load. Once the page and the WASM engine are cached, you can disconnect from the network and the concatenation will still run and download locally. Going offline is also the simplest way to prove to yourself (or a compliance reviewer) that nothing is being uploaded.
Does anything at all get sent to JAD's servers?
Only an optional anonymous run counter for signed-in users' dashboard stats — a count, with no file content, file names, or audio. If you're not signed in, or you go offline, nothing is sent. Your recordings never leave the device either way.
What output format should I use for a master record?
WAV or FLAC — both are lossless, so the concatenated record carries no re-encode loss. WAV is uncompressed (largest); FLAC is lossless-compressed (smaller). Use MP3/M4A (192 kbps) only for a smaller copy to distribute under controlled access.
Does the merge keep the original audio exactly?
The concatenation is exact — no audio is added or dropped at the seams (with crossfade 0). However, the tool always re-encodes and resamples to 48 kHz stereo, so output WAV/FLAC for a lossless result; MP3/M4A apply a (small, lossy) re-encode. Mono inputs are upmixed to stereo.
Can I merge multi-hour depositions or calls?
Yes, with the right tier. Free caps each file at 30 minutes / 50 MB; Pro at 120 minutes / 200 MB / 10 files; Pro-Media and Developer remove the duration limit and allow 100 GB files (100 / unlimited per batch). Multi-hour records typically need Pro-Media.
Do I need an account to merge sensitive files?
No account is required to run a merge and download the result. You don't have to register confidential material against a profile to process it. Accounts only affect tier limits and dashboard stats, not the local processing.
Will recording timestamps or metadata be preserved?
No. Re-encoding writes a fresh stream, so embedded timestamps, device metadata, and ID3 tags from the inputs aren't carried over. For chain-of-custody, document metadata separately and, if needed, add tags afterward with id3-editor.
How many files can I concatenate at once?
At least 2. The batch ceiling is by tier: Pro 10, Pro-Media 100, Developer unlimited. Free's audio batch limit is 1 file, so multi-file concatenation needs Pro or higher.
Can I join files of different formats and sample rates?
Yes. Every input is normalised to 48 kHz stereo before concatenation, so a phone M4A, a 48 kHz WAV, and a 44.1 kHz MP3 all join cleanly without you pre-converting them. Mono files are upmixed to stereo.
Can the tool prove the file wasn't tampered with?
No — it joins audio but doesn't sign, hash, or watermark the output. If you need tamper-evidence for evidentiary use, compute a hash (e.g. SHA-256) of the merged file yourself and record it in your own chain-of-custody documentation.
What should I do after concatenating a confidential recording?
Common local-only next steps: redact spoken PII with pii-redactor before sharing, split the record into parts with audio-splitter, or level it with loudness-normalizer. All run in-browser, so the audio stays on your device throughout.
Privacy first
Every JAD Audio tool runs entirely in your browser via FFmpeg (WebAssembly) and RNNoise. Your audio files never leave your device — verified by zero outbound network requests during processing.