What the words mean
Lossless audio is bit-for-bit identical to the source after decoding. Examples: WAV, FLAC, ALAC.
Lossy audio discards some of the original signal during encoding. The result is much smaller but slightly less detailed. Examples: MP3, AAC, OGG Vorbis, Opus.
How they sound
Lossless and a well-encoded high-bitrate lossy file sound nearly identical on consumer headphones and speakers.
Differences emerge at low bitrates, on high-end studio monitors, or with trained ears focused on subtle cymbal sustain, reverb tails, and vocal sibilance.
The trade-offs
Lossless: perfect quality, but 5–10× larger files. Best for editing, archiving, mastering, and audiophile listening.
Lossy: smaller files, broader compatibility, slight quality cost. Best for streaming, podcasts, web playback, and casual listening.
How to pick
If the file will be edited or processed further, keep it lossless.
If the file will be played once and discarded (a podcast, voice memo, demo), lossy at 128–192 kbps is plenty.
When in doubt, store the master lossless and export a lossy copy when you need to share.