FLAC to Ogg converter
A release bundle is a bandwidth problem with uneven material: quiet tracks and loud tracks need different Vorbis settings, so one global quality choice either wastes bytes or damages the quiet one. The FLAC masters stay authoritative; the Ogg files are the shipped copies.
The target codec is lossy, so the original samples are not restored bit for bit.
Processed on your device. Files aren’t sent anywhere.
The release constraint and the uneven material
Illustrative: a label releases a four-track EP mastered in FLAC, averaging about 38 MB per track. The release page streams on a slow connection and the bundle is distributed through open tooling. The masters are uniform; the tracks are not, so the quality decision is made per track.
- Drag one FLAC master onto this page; the local FFmpeg worker loads on first use
- Set quality for that track: Balanced for loud material, High for quiet or ambient material
- Convert, then read the reported codec, duration and channel count
- Repeat per track, then listen to the whole bundle in order before publishing
Release packaging questions
Why Ogg for an open-source release?
What does the quality setting mean?
Does Ogg keep the FLAC intact?
Why not one quality setting for the whole release?
Is the master uploaded?
The quality ladder, per track
Illustrative figures; the point is the shape of the trade, not the exact numbers.
| Track | Character | Quality | Illustrative size |
|---|---|---|---|
| Track 1 — synth, loud | Dense and loud | Balanced | ~7 MB |
| Track 2 — ambient, quiet | Soft, sustained, no transients to mask artefacts | High | ~11 MB |
| Track 3 — acoustic duo | Wide stereo, quiet room tone | High | ~9 MB |
| Track 4 — voice + piano | Speech-forward | Balanced | ~6 MB |
Why quiet material is the hard case
Vorbis spends bits on loud material first. Sustained quiet content has no transients to hide behind, so a low setting makes it shimmer. Encode the quiet tracks higher and listen to them on headphones, where the artefact is obvious.
- Ambient pads and room tone: High
- Wide acoustic stereo: High, then check the image does not collapse
- Loud dense mixes: Balanced is usually enough
- Speech-forward tracks: Balanced, unless the room tone matters
Listening to a quiet track is not the same as listening to a track
A loud track hides artefacts behind its own energy, so a low setting passes a quick audition. A quiet track exposes everything, which is why the check for tracks 2 and 3 is a headphone listen through the first thirty seconds rather than a skim of the whole EP.
Naming the release and keeping the masters separate
The shipped bundle and the master shelf must never be the same folder, or a future re-encode will pick up the lossy files.
- Masters folder holds FLAC only; release folder holds Ogg only
- Same stem in both folders, so a re-encode is a copy operation
- Record the quality setting used per track in the release notes
- One file per run, up to 60 MiB and 600 seconds of audio
Re-encoding when the master changes
A revised master means a fresh encode from the FLAC file, never from the shipped Ogg. Encoding from a shipped copy applies a second loss on top of the first.
- For MP3 delivery to legacy hosts: FLAC to MP3, bitrate chosen per recipient
- For editing or measurement: FLAC to WAV, which gives codec-free PCM
- If a shipped Ogg is all that exists, treat it as a lossy source: Ogg to FLAC consolidates it, but the loss stays