From Manuscript to Audiobook

Turning a finished manuscript into a full-cast audiobook looks daunting from the outside. In practice it's a short pipeline with a few decision points, and most of the effort goes into two of them: casting and review. Here's the whole path, in the order you'll actually walk it.

1. Prepare the manuscript

VoiceYourNovel accepts EPUB, PDF, DOCX, and plain text. EPUB and DOCX carry the cleanest structure; PDF works, but a few minutes of cleanup beforehand saves an hour of correction later. Before you upload:

• Strip front and back matter you don't want narrated — copyright pages, review blurbs, mailing-list calls to action.
• Remove page headers, footers, and page numbers if your export tool bakes them into the text.
• Check that scene breaks are marked consistently (a blank line, ***, or similar) — they become natural audio pauses.
• Make sure dialogue uses consistent quotation marks. Mismatched or missing quotes are the main cause of misattributed lines.

2. Map the chapters

After upload, the service splits the book into chapters using your headings. Skim the chapter list once: merge any prologue that got split in two, and mark interludes or letters that need special handling. Chapter titles here become the track titles in your final export, so fix typos now.

3. Review the detected characters

The character pass reads the entire book and returns a cast list: every speaking character, an estimated line count, and an inferred profile — age range, gender, temperament, narrative role. Three things to check:

• Merges: "Liz," "Elizabeth," and "her sister" should be one character, not three.
• Ghosts: a quoted proverb or a letter read aloud sometimes registers as a speaker. Fold those into the narrator.
• Minor roles: characters with a handful of lines can share a small pool of supporting voices instead of each getting a unique one.

4. Assign the voices

Each character gets a suggested voice matched to their profile. Audition the leads on a demanding scene, try an alternate or two, and lock your choices. (We've written a full guide on this — see "Casting Voices for Your Characters.") Keep the narrator neutral and grounded; let dialogue carry the color.

5. Render and review, line by line

Render your first chapter and listen with the transcript view open — every audio line is paired with its source sentence. You're listening for three failure modes: a line attributed to the wrong speaker, a mispronounced name or invented word, and a delivery that misses the emotional mark.

This is where the workflow pays off: fixes stay inside the chapter. Reassign the speaker, add a pronunciation hint, or adjust the delivery note, then re-render just that chapter — every other chapter stays exactly as it was. Review a chapter, patch its handful of lines, re-render, move on.

Front-load your effort on chapters one and two. Most systematic issues — a name the engine consistently mangles, a character cast slightly wrong — surface early, and fixing them at the source corrects the whole book.

6. Export

When the chapters pass review, export:

• One ZIP — every rendered chapter as its own audio file, numbered in reading order, ready for podcast feeds, direct sales, or simple listening.

Render settings, casting, and per-line fixes are all saved with the project, so when you revise the manuscript, you're updating a living production — not starting over.