FLAC to MP3 for Music Production

The first thing to be clear about: MP3 is not a production format. Never bring an MP3 into a session as source material if a FLAC or WAV exists, never master from one, and never bounce your stems to MP3 for archival. Every lossy encode is irreversible, and layering one under further processing — EQ, compression, limiting — amplifies its artifacts rather than hiding them. Inside the session, keep FLAC or WAV.

Where MP3 is genuinely correct is everything that leaves the session. Reference mixes you want to check in the car or on earbuds. Rough bounces sent to a client, a vocalist, or a bandmate. Demos on a drive. Anything where the file has to travel through email, a messaging app, or a phone — none of which cope well with a 40 MB lossless file, and none of which need to.

For those uses, encode from the FLAC at 320 kbps. Because FLAC is lossless, the encoder sees the complete mix and produces the cleanest MP3 that material can yield — noticeably better than an MP3 made from an already-lossy file. 320 kbps is transparent enough that a client will hear the mix, not the codec, which is the entire point of a reference send.

One caveat worth knowing when you use MP3 as a reference: lossy encoding can slightly raise peak levels, so a mix that reads exactly 0 dBFS can clip after encoding. If you are sending references regularly, leave a little headroom (around -1 dBTP) on the bounce you encode from. That is standard practice for distribution masters and it applies to reference MP3s too.

Keep FLAC as the deliverable master and the archive. Distributors and streaming platforms want lossless (FLAC or WAV) — never upload an MP3 for distribution, because the platform will transcode it again and you will have stacked two lossy generations on your release. Convert to MP3 for humans; keep lossless for machines.