Ogg to MP3 converter
The partner's validator is the constraint, not the codec: the episode is published as Ogg for your own site, and the syndication feed only accepts MP3. Encoding again means the MP3 pass inherits whatever Vorbis already discarded, so the bitrate is matched to the source rather than defaulted.
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 validator rejects a shape, not a quality
Illustrative: a collective publishes episodes as Ogg because its own site runs open tooling, and a syndication partner's feed validator rejects anything that is not MP3. Nothing is wrong with the audio. The job is one extra deliverable, and the risk is that the extra encode becomes audible.
- Drag the Ogg episode onto this page; the local FFmpeg worker loads on first use
- Set the bitrate to sit above the source's own budget rather than at the low default
- Convert once, then read the reported codec, duration and channel count
- Publish the MP3 copy to the partner and keep the Ogg copy on your own site
What a stacked encode does to the same episode
The same episode, two settings, one listen check. Illustrative figures; the point is which pass becomes the limiting factor.
| Source | MP3 bitrate | What the listener hears |
|---|---|---|
| Vorbis at Balanced | 96 kbps | Room tone and sibilance smear; the MP3 pass is the weak link |
| Vorbis at Balanced | 192 kbps | The MP3 pass has room; the inherited Vorbis loss stays as it was |
| Vorbis at High | 192 kbps | Quiet passages stay clean enough for syndication |
| Vorbis at Low | 128 kbps | Both passes are tight; artefacts read on phone speakers |
Listening for the artefacts a second pass adds
A file that converts cleanly can still read badly on the partner's players. The degradation is specific, so the check is targeted.
- Sibilance — the first thing a stacked lossy pass smears
- Quiet gaps — they can fill with a watery artefact instead of staying silent
- Stereo width — a wide mix can collapse where Vorbis had already narrowed it
What the engine reports after encoding
The engine probes the encoded file and compares codec and duration with the source; drift beyond a small tolerance is reported as a failure rather than handed over as a shortened MP3. It also reports the measured sample rate and channel count, which is worth reading because encoder builds differ.
Two-pass questions
What does the MP3 pass inherit from Vorbis?
How should the bitrate be chosen?
Should both copies be published?
Is the episode uploaded?
Will the sample rate be carried through?
Publishing both copies without version drift
Syndication breaks quietly when the partner and your site serve different encodes of the same episode. One stem, two suffixes.
- Same stem for both copies: episode-042.ogg and episode-042.mp3
- Note the MP3 bitrate in the feed notes so a later re-encode matches it
- One file per run, up to 60 MiB and 600 seconds of audio
What stays authoritative
The Vorbis master is the source for future re-encodes; the MP3 copy is a partner-shaped deliverable. Never encode the next release from the MP3 copy — that would stack a third loss.
- Consolidating the archive into a lossless shelf: Ogg to FLAC, with the lossy history noted
- Editing or measurement: Ogg to WAV for codec-free PCM
- If a real master exists, encode the partner copy from it instead of from Ogg