How to convert mp3 to wav and change sample rate
- Step 1Open the converter — Open the mp3-to-wav tool. The Sample rate dropdown is where the rate change happens during conversion.
- Step 2Drop your MP3 — Add one
.mp3. It's decoded in-browser by FFmpeg WASM — no upload. One file at a time. - Step 3Choose the target sample rate — Pick the rate your destination requires: 48 kHz for a video session, 16 kHz for transcription, 44.1 kHz for music, 96 kHz for a high-rate project. This is the rate the WAV will be written at.
- Step 4Choose channels — Stereo by default; Mono for speech/transcription or single-mic recordings. Mono + 16 kHz is the common ASR target.
- Step 5Convert and resample — Run it. FFmpeg decodes the MP3 and resamples to your chosen rate in the same pass, then writes 16-bit PCM WAV.
- Step 6Verify the rate and use it — The result panel shows the output sample rate and channels. Confirm they match your destination (e.g. exactly 16000 Hz mono for an ASR API) before importing or uploading to that system.
Target rate by destination
Pick the rate the destination wants. Downsampling (e.g. 44.1 → 16 kHz) discards content above the new Nyquist; upsampling adds no detail.
| Destination | Target rate | Channels | Why |
|---|---|---|---|
| Video/film DAW session | 48 kHz | Stereo | Matches 48 kHz video timelines; avoids drift vs picture |
| Speech-to-text / ASR API | 16 kHz | Mono | What most speech engines expect; smaller payloads |
| Music project / CD | 44.1 kHz | Stereo | Standard music rate; usually matches the MP3 source |
| High-rate / mastering | 96 kHz | Stereo | Match a 96 kHz session — but this upsamples a 44.1 kHz source |
| Telephone-band analysis | 16 kHz | Mono | Nyquist 8 kHz covers voice band |
Up vs down sampling — what to expect
Resampling math. The MP3's actual content ceiling (often a ~16 kHz lowpass) is the real limit, not the WAV rate.
| Direction | Example | Effect |
|---|---|---|
| Downsample | 44.1 kHz → 16 kHz | Discards content above 8 kHz (new Nyquist); much smaller file |
| Same rate | 44.1 kHz → 44.1 kHz | Pure decode, no resample — cleanest for fidelity |
| Upsample | 44.1 kHz → 96 kHz | Bigger file, no new frequencies — the source has none above ~22 kHz |
Real controls and limits
Only sample rate and channels are adjustable; bit depth is fixed 16-bit. Tier caps apply to the MP3 input.
| Item | Value / choices | Note |
|---|---|---|
| Sample rate | 44.1 / 48 / 96 / 16 kHz | -ar, default 44.1 kHz |
| Channels | Stereo / Mono | -ac, default Stereo |
| Bit depth | 16-bit (fixed) | pcm_s16le |
| Free tier | 50 MB / 30 min / 1 file | Both size and duration enforced |
| Pro tier | 200 MB / 120 min / 10 files | Higher caps |
| Pro-media / Developer | 100 GB / unlimited | 100 / unlimited files |
Cookbook
Rate-matching recipes. Decode + resample happen together; output is 16-bit PCM WAV at the target rate.
44.1 kHz MP3 → 48 kHz WAV for a video session
Your video edit runs at 48 kHz. The music bed is a 44.1 kHz MP3. Convert directly to 48 kHz so the NLE/DAW doesn't resample on import.
Input: bed.mp3 (192 kbps, 44.1 kHz, stereo) Options: Sample rate = 48 kHz · Channels = Stereo (-ar 48000) Output: bed.wav (16-bit PCM, 48 kHz, stereo) One resample, here — not a hidden one inside the editor.
Podcast MP3 → 16 kHz mono WAV for transcription
A transcription API wants 16 kHz mono PCM. Decode and downsample in one step to hit that exact shape and shrink the payload.
Input: episode.mp3 (128 kbps, 44.1 kHz, stereo, 35:00) Options: Sample rate = 16 kHz · Channels = Mono (-ar 16000 -ac 1) Output: episode.wav (16-bit PCM, 16 kHz, mono) 35 min > Free's 30 min cap -> needs Pro or higher.
Keep 44.1 kHz to avoid any resample
The MP3 and your destination are both 44.1 kHz. Keep the rate so the conversion is a pure decode — the most faithful path.
Input: song.mp3 (320 kbps, 44.1 kHz, stereo) Options: Sample rate = 44.1 kHz · Channels = Stereo Output: song.wav (16-bit PCM, 44.1 kHz, stereo) No resampling step — only bitstream -> PCM.
Upsample to 96 kHz to match a session (size warning)
Everything in your master session is 96 kHz, so the WAV must match — even though the MP3 source is 44.1 kHz and gains no real bandwidth.
Input: vox.mp3 (256 kbps, 44.1 kHz, mono) Options: Sample rate = 96 kHz · Channels = Mono (-ar 96000) Output: vox.wav (16-bit PCM, 96 kHz, mono) ~2.2× the bytes of a 44.1 kHz WAV; no frequencies above the source's ~16-22 kHz.
Downsample for a legacy 16 kHz system
An older voice system ingests 16 kHz WAV only. Decode the MP3 and downsample to meet the fixed requirement.
Input: prompt.mp3 (96 kbps, 44.1 kHz, mono) Options: Sample rate = 16 kHz · Channels = Mono Output: prompt.wav (16-bit PCM, 16 kHz, mono) Content above 8 kHz is discarded by design (new Nyquist = 8 kHz).
Edge cases and what actually happens
Upsampling expected to add quality
No new detailGoing from 44.1 kHz to 96 kHz inflates the file but adds no frequencies — the MP3 has nothing above its content ceiling (often a ~16 kHz lowpass, never above ~22 kHz). Upsample only to match a session, not to 'improve' the audio.
Downsampling discards high content
ExpectedResampling to 16 kHz removes everything above 8 kHz (the new Nyquist). For speech that's fine; for music it dulls the sound. Choose the lowest rate your destination actually needs.
The exact rate you need isn't offered
Limited choicesThe UI exposes 44.1 / 48 / 96 / 16 kHz. For 22.05 kHz, 8 kHz, or 88.2 kHz, this tool can't target them directly — convert to a close rate and resample in your own code, or use the dedicated sample-rate-converter within a format.
DAW still resamples after import
AvoidableIf the WAV's rate doesn't match the session rate, the DAW resamples again. Set the Sample rate dropdown to the session rate so your conversion is the only resample.
Need 24-bit at the new rate
16-bit onlyOutput is always 16-bit PCM regardless of rate. There's no bit-depth control; the resample doesn't change that.
Long file exceeds the duration cap
rejectedA 35-minute episode exceeds Free's 30-minute cap even if under 50 MB — duration and size are checked independently. Use Pro (120 min) or higher, or split with audio-trimmer.
Resampled WAV bigger than expected
By designHigher rates mean more samples per second and larger files; 96 kHz is ~2.2× a 44.1 kHz WAV. Downsampling to 16 kHz mono is the smallest option here.
Aliasing concerns when downsampling
HandledFFmpeg applies an anti-aliasing lowpass during downsampling, so frequencies above the new Nyquist are filtered before decimation — no aliasing artifacts from the rate change itself.
Multiple files needing the same rate
single fileThe tool is single-file. Apply the same Sample rate setting to each file one at a time. For combining afterward, use audio-merger (which can also re-encode to a common rate).
Expecting fidelity gain from the conversion
ExpectedNo rate change recovers MP3's lost data. The point is matching the destination rate cleanly, not improving the source. WAV stores the existing audio uncompressed.
Frequently asked questions
Can I change the sample rate while converting MP3 to WAV?
Yes. The Sample rate dropdown sets the output rate, and FFmpeg decodes + resamples in one pass. Choices are 44.1, 48, 96, and 16 kHz.
Which rate should I pick for transcription?
16 kHz mono is the common target for speech-to-text engines. Set Sample rate = 16 kHz and Channels = Mono to produce exactly that shape.
Does upsampling to 96 kHz improve the audio?
No. It only makes the file bigger. The MP3 has no frequency content above its encoder's ceiling, so a higher rate adds zeros' worth of new information, not detail.
What happens to high frequencies when I downsample to 16 kHz?
Everything above 8 kHz (the new Nyquist) is removed, with an anti-aliasing filter applied first. That's expected and fine for speech; it dulls music.
Why not just import the MP3 and let the DAW resample?
The DAW's resampler quality and timing are out of your hands, and it happens on every playback for some hosts. Resampling once, here, to the session rate gives a controlled, stable result.
Can I get 22.05 kHz or 8 kHz?
Not in this UI — it offers 44.1/48/96/16 kHz. For other rates, resample in your own pipeline, or use the sample-rate-converter to change rate within a format.
Is the output 24-bit at higher rates?
No. Bit depth is always 16-bit PCM regardless of the sample rate you choose. The tool has no bit-depth control.
Does changing rate make the file upload anywhere?
No. The decode and resample run in-browser via FFmpeg WASM. Nothing uploads; after the page loads the conversion works offline.
Will a 40-minute MP3 convert on the free tier?
No — Free caps at 30 minutes per file (and 50 MB). A 40-minute file needs Pro (120 min) or Pro-media/Developer (unlimited). Both duration and size are enforced.
How big will the WAV be at each rate?
Roughly: 16 kHz mono ≈ 1.9 MB/min; 44.1 kHz stereo ≈ 10.1 MB/min; 96 kHz stereo ≈ 22 MB/min. PCM scales with rate × channels × 2 bytes × seconds.
Does the converter avoid aliasing when downsampling?
Yes. FFmpeg's resampler applies an anti-aliasing lowpass before decimation, so the rate change itself doesn't introduce aliasing.
What if I need to resample an existing WAV, not an MP3?
Use the sample-rate-converter, which changes rate within a format. This tool is specifically MP3 → WAV with a rate change folded into the decode.
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.