M4B
M4B is the MPEG-4 container used for audiobooks. It holds an entire book as a single file, carries chapter markers, and signals to the player that it should remember your position between sessions. The audio inside is normally AAC. The file extension is the only real difference from an M4A: the same container renamed, so players treat it as a book rather than as music.
M4B against MP3
| M4B | MP3 | |
|---|---|---|
| Structure | One file for the whole book | One file per chapter, typically |
| Chapter markers | Embedded, navigable | Not in the format; the filenames stand in |
| Resume position | Remembered by the player | Depends entirely on the app |
| Cover and metadata | Embedded | Embeddable but often lost |
| Plays everywhere | Most audiobook apps | Effectively everything |
Why both are usually delivered
M4B is the better listening experience and MP3 is the more universal one. Distributors differ on what they accept, and ACX in particular requires separate MP3 files, one per chapter, at 192 kbps constant bitrate and 44.1 kHz. So a production that hands you only an M4B has left you unable to submit to ACX, and one that hands you only MP3s has given you nothing with chapter navigation. Getting both is the point.
What the chapter markers are actually doing
An M4B stores a table of named time offsets. A player reads it and shows a chapter list you can skip through, which is the difference between navigating a ten-hour book and scrubbing through a ten-hour file. Markers only exist if something wrote them, which means chapter structure has to survive from your manuscript all the way to the finished file. That is why manuscripts with real heading styles convert cleanly and manuscripts using manual line breaks to look like chapters do not.
Bitrate and file size
Spoken word does not need music bitrates. AAC at 64 kbps mono is transparent for narration, which puts a ten-hour book at roughly 290 MB as a single M4B. ACX's per-chapter MP3 requirement of 192 kbps constant at 44.1 kHz is considerably more generous than speech needs, and produces a set of files several times larger in total. Neither is wrong: the M4B is sized for listening, the MP3 set is sized to a distributor's specification.
Where you will meet each one
Apple Books and most dedicated audiobook players want M4B. ACX wants per-chapter MP3. Aggregators vary and generally accept both. Keep the M4B as your master and treat the MP3 set as the distribution deliverable.