Core Concepts
A quick map of how the app organizes your work, and how a chapter becomes audio.
At a glance
- Library, then Project, then Chapter, then Segment.
- A Voice has Variants, built from Samples.
- Audio goes through five steps: analysis, queue, synthesis, baking, assembly.
How your book is organized
- Library: all your projects.
- Project: one audiobook, with its author, series and cover.
- Chapter: one division of the book, holding its text and audio.
- Segment: a sentence or paragraph in a chapter, the smallest piece of text the app stores.
- Chunk group: what you see in the Performance and Production tabs. One group can hold several neighboring segments that belong together.
- Character: a speaker you assign to lines. By default a chapter is read by the Narrator.
How voices are organized
- Voice: an identity you assign to characters.
- Variant: one style or mood of that voice, built from its own recordings.
- Sample: a reference recording used to clone the voice.
From text to audiobook
- Analysis: the app looks for sentences over 500 characters and splits them.
- Queuing: when you click Generate, the groups join the background queue.
- Synthesis: each group is spoken by its assigned voice.
- Baking: the finished groups are stitched into one chapter file, also saved as an M4A.
- Assembly: the M4A chapters are joined into the final
.m4b. Because they are saved, rebuilding is nearly instant.