OGG (the container; usually with Vorbis or Opus inside) is the open-source world's answer to MP3. It's what Linux audio recorders default to, what Wikipedia stores its audio samples as, what some Discord and Telegram exports use, and what you'll receive when someone hands you a file from a non-Apple, non-Microsoft pipeline. MDisBetter accepts OGG directly — no need to transcode to MP3 first.
Where OGG comes from
OGG is a container format from Xiph.org, designed as a patent-unencumbered alternative to commercial codecs. OGG Vorbis: the original spec, lossy compression at quality competitive with MP3, common in open-source recording tools (Audacity's default export option for years), Wikimedia audio, some game audio assets. OGG Opus: a more recent codec optimised for both music and voice, used by Discord for voice and by some Telegram exports — covered separately on the Opus to Markdown page. The transcription pipeline handles both flavours from a .ogg extension.
Linux and open-source workflows
If you're on Linux and you record audio with the default tools — sox, ffmpeg with Vorbis output, GNOME Sound Recorder, KDE Krecord — the file you end up with is almost always OGG. Forcing a transcode to MP3 just to upload to a transcription service is a step that costs quality (re-encoding lossy is always slightly lossier than the original) and time. Direct OGG support means the recording goes straight from your Linux box to Markdown.
Common voice-message use cases
Discord voice messages and some older Telegram voice clips export as OGG. If a colleague drops you an OGG voice message and you want to skim it instead of listen, transcription gives you a 30-second read instead of a 2-minute listen. Markdown output with timestamps lets you jump back to the audio if you need to verify a quote or re-listen to tone.
# Project Sync — May 9, 2026
## Status Updates
[00:00:04] OK let's do quick round of updates. Mark, what's the state of the deploy pipeline? Mostly green. Two flaky tests in the integration suite that I'm chasing. Should be fixed by EOD.
## Next Sprint Planning
[00:02:18] Cool. Let's talk next sprint. What's top of mind?
Frequently asked questions
Is OGG supported, the same as MP3 and WAV?
Yes — OGG (both Vorbis and Opus inside the container) is fully supported. Drop the .ogg file in the same way you would an MP3, get the same structured Markdown output with headings and timestamps. No transcoding step required on your end.
What's the difference between OGG Vorbis and OGG Opus?
Vorbis is the older codec (2002, lossy, similar quality tier to MP3 at the same bitrate). Opus is newer (2012, lossy, more efficient especially at low bitrates and for voice). Both ride inside the OGG container. For the transcription engine the input is decoded the same way regardless of inner codec; the output Markdown is identical.
Can I transcribe Wikipedia's audio samples?
Yes — Wikipedia hosts pronunciation samples and historical audio as OGG Vorbis files. Right-click → Save the .ogg, drop into MDisBetter, get the spoken content as Markdown. Useful for language-learning workflows or for indexing the spoken-word components of a Wikipedia article into searchable text.
Will OGG voice messages from Discord transcribe correctly?
Yes. Discord voice messages are typically high-quality enough for clean transcription. Multi-speaker recordings (a Discord call exported as one OGG file) come back as one continuous transcript: we do not identify or separate speakers, so heavy crossing conversation is harder to follow in the text. Sections and timestamps still give you the structure to navigate it.
Are OGG files supported on platforms that don't play them natively (iOS, older Windows)?
Audio playback support is irrelevant to transcription — the file is uploaded and processed on our servers regardless of whether your local machine can play it. So yes, you can transcribe an OGG on an iPhone (where Safari can't play it natively) by uploading it through the web tool.