MP3 vs WAV: compressed delivery vs uncompressed master
By Andrew Butson · July 2, 2026 · 7 min read
Use WAV while you are recording, editing or archiving. It holds the exact waveform with nothing thrown away, so repeated edits and re-saves never compound quality loss. Use MP3 for everything downstream of that: sharing, streaming, and everyday listening, where a file roughly a tenth the size sounds indistinguishable from the original to most people on most gear. WAV is the master; MP3 is what you actually hand someone.
WAV and MP3 answer different questions. WAV asks "how do I store this recording with nothing lost?" and answers it in the simplest way possible: write down the exact amplitude of the waveform, over and over, many thousands of times a second, with no compression at all. MP3 asks "how small can I make this and still have it sound the same to a listener?" and answers it by throwing away the parts of the sound a person is least likely to notice.
Neither format is wrong for what it does. The mismatch happens when the wrong one ends up in the wrong place: a WAV clogging up someone’s phone storage, or an MP3 being used as a mixing session’s working file, quietly degrading with every re-export.
| MP3 | WAV | |
|---|---|---|
| Compression | Lossy (perceptual encoding) | None (raw PCM samples) |
| Typical size (per minute, CD quality) | ~1 MB at 128 kbps | ~10 MB |
| Sound quality | Near-transparent at high bitrates, audibly worse at low ones | Bit-perfect, exact waveform |
| Editing and re-saving | Each re-encode compounds the loss | No generation loss across edits |
| Playback support | Universal and lightweight (the default everywhere) | Universal, but a much larger download |
| Typical use | Streaming, downloads, everyday listening | Recording, mixing, mastering, archiving |
| Metadata | ID3 tags widely supported | Minimal; extensions exist but are inconsistent |
What PCM actually is
A WAV file is, underneath its small header, pulse-code modulation (PCM): the literal amplitude of the sound wave, measured and written down at a fixed rate. CD-quality audio samples 44,100 times a second, with each sample stored as a 16-bit number, doubled up for stereo. There is no model of hearing involved and no attempt to guess what matters: every single measurement is kept, in order, exactly as captured.
That simplicity is precisely why WAV is the default in recording and editing software: decoding it means reading numbers off disk, with no computation and no possibility of introducing a decoding artefact. It is also exactly why the files are so large. The format spends the same number of bits on every instant of audio, whether that instant is a cymbal crash or three seconds of silence.
The size maths
The numbers are worth having in your head. A minute of CD-quality (16-bit, 44.1 kHz, stereo) WAV audio comes to roughly 10 MB — 44,100 samples per second × 2 bytes per sample × 2 channels × 60 seconds. The same minute as a 128 kbps MP3 comes to roughly 1 MB. That is a tenfold difference from one file, and it multiplies fast: a three-minute song is about 30 MB as WAV and about 3 MB as MP3; an hour-long podcast recording is around 600 MB as WAV against maybe 55 MB as MP3.
Higher MP3 bitrates close some of that gap (320 kbps runs to roughly 2.4 MB a minute), but even at the top of MP3’s usual range it is still four times smaller than the uncompressed original, which is the whole reason lossy formats exist for distribution.
How MP3 gets there
MP3 shrinks a file by modelling human hearing rather than the sound itself. Quiet sounds masked by louder ones nearby, and frequencies at the very edges of what people can hear, are reduced or dropped outright, because most listeners will not consciously miss them. The bitrate you choose controls how aggressively this happens: a high bitrate like 256–320 kbps makes conservative cuts that are very hard to notice; a low bitrate like 96–128 kbps cuts much deeper, and on complex material such as cymbals, massed strings, and applause, the shortcuts start to become audible as a thin or swishy quality.
This is a one-way process. Once the psychoacoustic model has decided a piece of the signal is expendable and removed it, no later step can put it back, which is why decoding an MP3 back to WAV, however tempting, never recovers the original recording.
When WAV is required, and when it's waste
WAV earns its size in a handful of specific jobs: recording a take, mixing and mastering in a DAW, or handing a finished master to a client, label or broadcaster who expects uncompressed PCM. In all of these, the file is going to be read, processed and re-saved repeatedly, and PCM is the one format where none of that compounds any loss.
Outside those jobs, WAV is usually just wasted space. A voice memo you are sending a friend, a podcast episode going up for streaming, a song loading onto a phone: none of these need the full uncompressed waveform, and carrying it around costs bandwidth and storage for no audible benefit. If you do need a lossless archive without WAV’s bulk, FLAC gets you the same bit-perfect audio at roughly half the size — see our FLAC vs MP3 guide for that trade-off.
Converting between them
WAV to MP3 is the standard last step of a project: pick a bitrate suited to the destination (128 kbps is fine for spoken word, 256–320 kbps is the safer choice for music), and Morphr encodes the WAV down for sharing or upload.
MP3 to WAV runs the other way but does not upgrade anything: it decodes the compressed file back into raw PCM samples, which is occasionally useful when a tool insists on WAV input, but the audio inside is still exactly what the MP3 already contained, missing detail and all. It is a container swap, not a quality restoration.
Reach for MP3 when…
- Sharing a track online, by email, or anywhere storage is limited.
- The playback device is a phone, earbuds or car stereo, not a mixing desk.
- Bandwidth or storage is the real constraint, not archival integrity.
- 128–320 kbps is genuinely enough quality for how it will be heard.
Reach for WAV when…
- Recording or tracking audio in a DAW.
- Mixing or mastering, where every edit and re-save must not compound loss.
- Delivering a master to a client, label or broadcaster who expects PCM.
- Archiving a recording you might still need at full quality — though FLAC often archives just as safely at half the size.
Shrink an uncompressed recording down to a shareable MP3, entirely in your browser; nothing is uploaded.
Runs entirely in your browser — nothing is uploaded.
Frequently asked questions
- Does MP3 sound worse than WAV?
- It depends on the bitrate. At 256–320 kbps, most listeners cannot reliably tell an MP3 from the original WAV in a blind test on typical playback gear. At 128 kbps or lower, the difference becomes audible, especially on complex or percussive material.
- Can I get WAV quality back by converting an MP3 to WAV?
- No. Decoding an MP3 to WAV only unpacks the samples the MP3 already stores. It cannot restore the detail the MP3 encoder discarded when it was first compressed.
- Why are my WAV files so big?
- WAV stores every sample of the waveform with no compression, which means CD-quality audio runs to roughly 10 MB per minute. That is the direct cost of keeping the recording bit-perfect rather than modelling what a listener would notice missing.
- What MP3 bitrate should I use?
- Around 128 kbps is generally fine for spoken word and podcasts, where the material is simple. For music, 256–320 kbps is the safer choice: it keeps the compromises close to inaudible for most listeners and most gear.