Understanding audio formats
MP3, AAC, FLAC, OGG — and what they actually trade.
The four audio formats you'll meet, the lossy/lossless divide, and the bitrate numbers that matter for each.
Lossy vs lossless.
The big split. Lossy formats (MP3, AAC, OGG Vorbis, Opus) discard audio information the human ear is least likely to notice — quiet sounds masked by loud ones, high-frequency content above 16 kHz — and compress aggressively. Lossless formats (FLAC, ALAC, WAV) preserve every sample exactly, with smaller compression ratios. The same 5-minute song is roughly 50MB as WAV, 25-30MB as FLAC, 5MB as MP3 at 128kbps, 7MB as MP3 at 192kbps.
Bitrate is the dial.
For lossy formats, bitrate (kbps — kilobits per second) determines how much information per second of audio is kept. 128 kbps MP3 is "okay for podcasts, not great for music"; 192 kbps is "good for casual music listening"; 256 kbps and 320 kbps are the practical ceiling for MP3 — going higher gives diminishing returns because the format has structural limits. AAC at the same bitrate sounds better than MP3 because the codec is newer; Opus at 128 kbps beats MP3 at 192.
The four formats to know.
MP3 (1993): the format your parents recognise. Patents expired in 2017. Most compatible, slightly inferior codec quality. AAC (1997): Apple's preferred default; better than MP3 at the same bitrate. The default for YouTube, Spotify, podcasts on Apple. FLAC: lossless, free, audiophile-friendly. Use for archival or critical listening. Opus: the modern winner — sounds great at any bitrate, used by every VoIP system (WhatsApp voice notes, Zoom audio). Open source, no patents.
A worked re-encode.
Convert a 5-minute WAV (50MB) to common formats. MP3 at 192kbps: 7MB, 7× reduction, near-CD-quality for casual listening. AAC at 192kbps: 7MB, audibly better than MP3 at the same size. Opus at 128kbps: 4.6MB, comparable quality to 192kbps MP3. FLAC: 25-30MB, lossless, every original sample preserved. Each is right for a different use case.
One song, four formats
5 min WAV (50 MB) re-encoded
Same source; different bytes-per-quality trade.
MP3 192 → 7MB ; AAC 192 → 7MB ; Opus 128 → 4.6MB ; FLAC → ~28MB
= Pick by use case
Re-encoding is lossy.
Converting MP3 to MP3 (at a different bitrate) is a quality loss — the source was already lossy, and re-encoding compounds the artifacts. Converting MP3 to FLAC doesn't recover lost quality; FLAC of a lossy source is a lossless storage of a lossy original. For archival, always start from a lossless source (CD rip, WAV, FLAC). The "convert my MP3s to FLAC" move is a common mistake — it makes the files bigger without making them better.
Sample rate and bit depth.
Below the format choice. CD-quality audio is 44.1 kHz / 16-bit; "studio quality" is 48 kHz / 24-bit (and higher). For voice content, 22 kHz / 16-bit is indistinguishable to most listeners and halves the file. Downsampling to 22 kHz loses everything above 11 kHz — fine for speech, not for music with high-frequency detail like cymbals. The converter should preserve the source rate by default and let you opt into downsampling explicitly.
FFmpeg in your tab.
Browser audio conversion uses ffmpeg.wasm — the same FFmpeg binary that powers half of the internet's media pipelines, compiled to WebAssembly. First load is slow (~20 MB), subsequent operations are fast. The same codecs work; the same flags do the same things. Local conversion preserves privacy and avoids upload time on large files.