The Complete Audiobook Production Process

Audiobooks are the fastest-growing segment in publishing, and for good reason. They let readers absorb content during commutes, workouts, and chores. But producing a professional audiobook involves more than reading a manuscript into a microphone. Here is what the process actually looks like, from first read-through to final distribution.

Stage 1: Manuscript Preparation

Before any recording begins, the manuscript needs to be prepared for audio. This means:

• Pronunciation guides — scientific terms, proper nouns, foreign words. A narrator who mispronounces "epitome" or "nuclear" will lose the listener's trust immediately.
• Visual content adaptation — charts, tables, and images in the book need verbal descriptions or need to be cut. Audio listeners cannot see a bar graph.
• Structural markers — chapter breaks, section transitions, and any front/back matter that should be included (dedication, foreword, appendices).

Stage 2: Voice Selection

The voice sets the tone for the entire listening experience. Key decisions include:

• Single vs. multi-narrator — most nonfiction works well with a single authoritative voice. Multi-narrator works for dialogue-heavy or interview-based books.
• Tone matching — a kids' science book needs warmth and enthusiasm. An adult medical reference needs calm authority. The voice must match the reader's expectations for the genre.
• Neural TTS vs. human narration — modern neural text-to-speech (like Kokoro) produces remarkably natural results at a fraction of the cost and time. For large catalogs, this is the practical choice.

Stage 3: Recording and Generation

Whether using a human narrator or TTS, the raw audio needs careful attention:

1. Chapter-by-chapter processing — each chapter is recorded or generated as a separate file for easier editing and quality control.
2. Pacing control — narration speed should be consistent (typically 150-170 words per minute for nonfiction). Faster for exciting passages, slightly slower for complex explanations.
3. Breath and pause management — natural pauses between sentences and sections. No run-on paragraphs.
4. Error correction — mispronunciations, plosives, and artifacts get flagged and re-generated or re-recorded.

Stage 4: Post-Production and Mastering

Raw audio is never the final product. Post-production makes the difference between amateur and professional:

• Noise floor reduction — background hiss, hum, and room tone are removed or reduced to -60dB or below.
• Dynamic range compression — ensures quiet passages are audible and loud passages don't clip. Listeners shouldn't need to adjust volume every chapter.
• EQ and de-essing — balances the frequency spectrum and tames harsh sibilance ("s" and "sh" sounds).
• Loudness normalization — targets -14 LUFS for streaming platforms (Audible, Apple Books) or -16 LUFS for direct distribution.
• Chapter markers and metadata — embedded chapter markers so listeners can navigate, plus title, author, and cover art in the file metadata.

Stage 5: Quality Control

Every finished audiobook goes through a QC pass before delivery:

• Listen to the first and last 60 seconds of every chapter.
• Spot-check at least 3 random points per chapter for pronunciation and pacing.
• Verify chapter order and metadata accuracy.
• Confirm total runtime is proportional to word count (roughly 1 hour per 9,000-10,000 words).
• Test playback on at least two devices (phone speaker, headphones).

Stage 6: Distribution

The final master files are delivered in the format required by each platform. Common requirements include:

• MP3 (192kbps CBR) for most platforms.
• M4B for Apple Books (with chapter markers).
• WAV or FLAC for archival and future re-mastering.

At BellerCreatives, we generate all three formats from every production run, so the author always has a lossless archive regardless of where the audiobook is distributed.