Does converting MP3 to WAV improve quality?

No. High frequencies removed by MP3 are not restored by converting to WAV; the file only grows, to about 4.4 times the size of MP3 320 kbps.

How big is the difference between MP3 320 kbps and WAV?

In our 2026-09-30 measurement, MP3 320 kbps stayed within 0.2 dB of the source below 20 kHz and reduced content above 20 kHz by about 20 dB, at 23% of the WAV size. This is not a listening test, so it does not judge audibility.

Should I archive in WAV or FLAC?

Both are bit-identical to the source. FLAC saves space at 8–37% of the WAV size, while WAV opens directly in more players and editors.

MP3 vs WAV — Measured File Size and Audio Quality

WAV is a lossless format that stores the original audio at about 10.6 MB per minute, while MP3 at 320 kbps takes about 2.4 MB (23% of WAV) but is lossy and reduces content above 20 kHz.

By Chaez · Published 2026-09-30 · Last updated 2026-09-30 · Data as of 2026-09-30

How different are the file sizes?

One minute of 44.1 kHz, 16-bit stereo audio, converted with the same encoder settings as the ToneNook converter, measured 10.58 MB as WAV, 2.40 MB as MP3 320 kbps and 0.96 MB as MP3 128 kbps (measured 2026-09-30).

File size per minute (44.1 kHz, 16-bit stereo), measured 2026-09-30
Format	Size per minute	vs WAV	Identical to source?
WAV	10.58 MB	100%	Yes (bit-identical)
FLAC	0.83–3.92 MB	8–37%	Yes (bit-identical)
M4A 256 kbps	1.63–1.67 MB	15–16%	No (lossy)
MP3 320 kbps	2.40 MB	23%	No (lossy)
MP3 192 kbps	1.44 MB	14%	No (lossy)
MP3 128 kbps	0.96 MB	9%	No (lossy)

WAV and MP3 sizes are set by bitrate, not by the audio content. WAV runs at 44,100 × 16 bits × 2 channels = 1,411 kbps, and MP3 320 kbps runs at 320 kbps, about 23% of WAV. FLAC and M4A sizes depend on the content; the ranges above come from two test signals, a simple arpeggio and full-band pink noise.

What does MP3 remove?

After encoding full-band pink noise and decoding it again, MP3 128 kbps reduced everything above 19 kHz by about 70 dB, effectively removing it, and lowered 16–19 kHz by 7.5 dB. MP3 320 kbps stayed within 0.2 dB up to 20 kHz and reduced only content above 20 kHz, by about 20 dB (measured 2026-09-30).

Energy change by band vs source (dB), 60 s pink noise, measured 2026-09-30
Format	16–19 kHz	19–20 kHz	Above 20 kHz
WAV · FLAC	0	0	0
M4A 256 kbps	−0.2	−0.2	−0.2
MP3 320 kbps	0.0	−0.2	−19.7
MP3 192 kbps	−0.8	−63.4	−65.3
MP3 128 kbps	−7.5	−69.5	−69.7

The MP3 encoder (LAME) cuts from a lower frequency as the bitrate drops, spending the remaining data on more important bands. This table shows which bands are removed; it is a measurement, not a listening test of whether people can hear the difference.

How is FLAC different from WAV?

Decoding FLAC and comparing it with the source WAV sample by sample showed both test signals were bit-identical. FLAC compresses without discarding audio, so its size was 8% of WAV for the simple tone and 37% for pink noise, the hardest case to compress. Real music usually falls within this range, and the exact value varies by track.

Which format should you choose?

Recommended format by use case
Use case	Format	Why
Archiving masters · re-editing · distributor delivery	WAV or FLAC	Bit-identical to the source. Check the format your distributor requires
High-quality listening and sharing	MP3 320 kbps or M4A 256 kbps	15–23% of WAV size, within 0.2 dB below 20 kHz
Demo sharing where size matters most	MP3 128 kbps	9% of WAV size, but content above 19 kHz is removed

Converting a lossy file back to a lossless format does not restore what was removed. Keep working masters as WAV or FLAC and convert to MP3 or M4A only for distribution.

How we measured

Test signals: 60 s, 44.1 kHz, 16-bit stereo. (1) The ToneNook demo arpeggio (synthesized by us), looped — a simple signal. (2) Pink noise generated with ffmpeg — full-band and the hardest case to compress. No recordings or third-party audio were used.
Encoders: the same codecs and bitrates (-c:a, -b:a) as the ToneNook converter, run with ffmpeg 8.0.1. The converter uses ffmpeg.wasm (core 0.12.10) in the browser, so file sizes may differ slightly.
Band change: each file was decoded and the same span analyzed with an FFT; the energy ratio per band versus the source is reported in dB.
Identity: FLAC and WAV were decoded and compared with the source sample by sample.
OGG (Vorbis) was excluded because the measurement environment lacked the encoder the converter uses (libvorbis).

ffmpeg -i source.wav -c:a libmp3lame -b:a 320k out.mp3   # MP3
ffmpeg -i source.wav -c:a aac -b:a 256k -movflags +faststart out.m4a   # M4A
ffmpeg -i source.wav -c:a flac out.flac   # FLAC

Frequently asked questions

Does converting MP3 to WAV improve quality?: No. High frequencies removed by MP3 are not restored by converting to WAV; the file only grows, to about 4.4 times the size of MP3 320 kbps.
How big is the difference between MP3 320 kbps and WAV?: In our 2026-09-30 measurement, MP3 320 kbps stayed within 0.2 dB of the source below 20 kHz and reduced content above 20 kHz by about 20 dB, at 23% of the WAV size. This is not a listening test, so it does not judge audibility.
Should I archive in WAV or FLAC?: Both are bit-identical to the source. FLAC saves space at 8–37% of the WAV size, while WAV opens directly in more players and editors.