Audio formats explained: a podcaster’s guide
A podcast passes through four or five format decisions between the microphone and the listener. Only one of them is irreversible, and it is the first one.
Contents
You meet audio formats at every stage of a podcast — capture, edit, master, publish, archive. Most of those choices can be changed later. The one at the start cannot, because a lossy encoder throws information away and nothing downstream puts it back.
WAV: the format to record and edit in
WAV is a container for uncompressed PCM. Every sample your interface converted is stored exactly as captured, which is what makes it the sensible working format: editing, noise reduction, levelling and EQ all operate on the signal, and you want that signal intact while you are still making decisions about it.
The cost is size, and the size is predictable rather than mysterious. Take a one-hour mono episode at 44.1 kHz and 16 bits:
- 44,100 × 16 × 1 × 3600 ÷ 8 ÷ 1,000,000 ≈ 318 MB, approximately — the exact figure depends on the header and on whether your editor writes any extra chunks.
- The same hour in stereo doubles it, because channel count is a plain multiplier.
- Moving from 16-bit to 24-bit multiplies it by 1.5 for the same reason.
That is a large file for an hour of speech, and it is the correct file to be holding while you edit. Storage is recoverable; discarded audio is not. The case for keeping uncompressed masters is the same argument applied to your archive.
MP3: the format to publish in
MP3 is lossy: a psychoacoustic model removes parts of the signal it judges least likely to be noticed, permanently. For a finished, mixed episode being streamed to a phone, that trade is usually the right one — the file shrinks by an order of magnitude and speech survives it well.
The widely recommended range for spoken-word podcasts is around 128 kbps mono, rising to roughly 192–256 kbps stereo when an episode carries music beds or musical segments. Treat that as a starting point rather than a rule: it reflects common practice across podcast hosts and directories, not a specification anyone publishes as a requirement. Check your own host's stated upload guidance before you settle on an encoder preset.
Bitrate alone determines an MP3's size, which makes the sizes as checkable as the WAV ones: 128,000 × 3600 ÷ 8 ÷ 1,000,000 ≈ 58 MB for that same hour.
AAC, FLAC and OGG: where each one actually fits
AAC / M4A
AAC (Advanced Audio Coding), usually delivered in an M4A container, is a lossy codec that generally holds up better than MP3 at low bitrates. It is broadly supported and is commonly used for delivery inside the Apple ecosystem. That is a description of where you will encounter it, not a claim about any platform's stated preference — publication requirements change, so read your host's current upload documentation rather than trusting a second-hand summary.
FLAC
FLAC compresses losslessly: the decoded output is bit-identical to the WAV it came from, typically at roughly half to two-thirds the size. That makes it an excellent archive format and a good transfer format between editors. It is a poor publication format, because acceptance across podcast hosts and player apps is inconsistent.
OGG Vorbis
OGG Vorbis is an open, royalty-free lossy codec of broadly comparable quality to MP3. It is well supported on the open web and in game engines, and less consistently supported across podcast directories and mobile players — which is why it is rarely the right choice as a primary distribution format for a show.
Which format at which stage
| Stage | Format | Why |
|---|---|---|
| Record | WAV, 44.1 or 48 kHz, 16 or 24 bit | Nothing is discarded before you have edited anything. |
| Edit | WAV | Processing operates on an intact signal; no repeated decode cycles. |
| Master | WAV | One authoritative full-quality file per episode. |
| Publish | MP3 (AAC where your host supports it) | Small, and supported across podcast apps and directories. |
| Archive | WAV or FLAC | FLAC halves the storage with no loss; WAV needs no decoder. |
The shape of that table is the whole method: stay lossless until the last step, then compress exactly once. Every extra encode-decode round trip compounds the loss, which is why re-exporting a published MP3 as a "new" master is a mistake even though the file will open perfectly well.
When a guest sends you an MP3
This is the common exception, and it is worth being honest about what you can and cannot do with it. Converting a guest's MP3 to WAV before you import it gives your editor a plain PCM source and avoids repeated decoding during the session — but it does not restore anything the encoder removed. The audio you received is the ceiling.
What it does buy you is a clean, consistent session: one format across all tracks, no format-specific import quirks, and no further compression until you export. If you want the technical comparison in more depth, MP3 vs FLAC vs WAV covers the three formats side by side, and what actually differs between podcast hosts covers the upload requirements you will meet at the other end.
Need to convert a file right now?
Free and private — your files are processed in your browser and are never uploaded. Up to 100 MB per file, 50 files per batch.
Convert MP3 to WAVSources & method
File sizes below are shown as arithmetic you can repeat from the sample rate, bit depth and channel count rather than quoted as measurements. Codec behaviour is described from the published definitions of the formats. Where a specific platform’s current stated preference could not be verified against live documentation, this guide describes what the format does instead of what a vendor is said to prefer.
- FLAC — Free Lossless Audio Codec, Xiph.Org Foundation. xiph.org Accessed 2026-08-29
- mp3 — licensing program terminated, Fraunhofer IIS. iis.fraunhofer.de Accessed 2026-08-29