Lossy vs lossless audio
What each actually does to the sound, when the difference is audible, and why converting between them in the wrong order wastes the file size you paid for.
In short
Lossless formats keep every sample; lossy formats discard sound you are unlikely to hear. For listening, a good lossy file is transparent to most people. For editing or archiving, keep lossless — and never expect converting lossy back to lossless to restore anything.
Every audio format falls into one of two camps, and almost every practical decision about audio files follows from which camp you are in.
Lossless keeps everything
WAV stores the raw samples with no compression at all — it is the digital equivalent of writing down every measurement. FLAC compresses those same samples the way a ZIP file compresses text: it finds patterns and stores them more briefly, and decompressing gives back the identical waveform, sample for sample.
So FLAC is typically 40–60% the size of the equivalent WAV with no quality difference whatsoever. If you want lossless and size matters at all, FLAC is strictly the better choice.
Lossy throws sound away
MP3, AAC and Vorbis use a psychoacoustic model — a model of what human hearing actually resolves. A quiet sound immediately after a loud one is masked and cannot be heard, so it is not stored. Frequencies above most people's hearing range are removed. Precision is spent where the ear is sensitive and withheld where it is not.
Done well at a reasonable bitrate, this is genuinely hard to distinguish from the source on normal equipment. Done aggressively, you hear it: cymbals turn watery, reverb tails swirl, and the top end sounds smeared.
| Format | Type | Encoder used | Bitrate setting |
|---|---|---|---|
| WAV | Lossless, uncompressed | pcm_s16le | Ignored — there is nothing to set |
| FLAC | Lossless, compressed | flac | Ignored |
| MP3 | Lossy | libmp3lame | Applies |
| AAC (.m4a) | Lossy | aac | Applies |
| OGG | Lossy | libvorbis | Applies |
The rule that matters most
Lossy compression is not reversible, and it compounds. Converting an MP3 to WAV does not restore anything — you get a large file containing exactly the same damage. Converting that WAV back to MP3 discards a second time, on top of the first.
Which to use
- Recording, editing or archiving — lossless. WAV while working, FLAC for storage.
- Sending, uploading or listening — lossy. MP3 for maximum compatibility, AAC or OGG for slightly better quality at the same size.
- Sending to someone who will edit it — lossless, even though it is bigger. You are handing them a working file, not a listening copy.
Mistakes worth avoiding
- Converting MP3 to WAV or FLAC to "improve quality". It only improves file size, in the wrong direction.
- Setting a bitrate for a WAV or FLAC export and expecting it to do something. Those encoders ignore it.
- Re-encoding the same file several times while trying different settings. Go back to the original each time.
Tools used in this guide
Related guides
MP3 bitrate explained
What each bitrate actually sounds like, the sizes they produce, and why speech and music want completely different numbers.
2 min readaudioWhy WAV files are so large
The exact arithmetic behind 10MB a minute, why that is the point rather than a flaw, and what to convert to when you no longer need it.
2 min read