Audio Fabric Studio logo Audio FabricStudio
Knowledge base Speech

Voice, podcast, audiobook, and course audio

How speech-oriented audio differs by destination: podcasts, interviews, audiobooks, courses, voiceover, captions, and transcription.

Practical guide Creator focused Preview-first workflow
Try this on your audio Back to knowledge base
Study goals

What you will understand

This guide is written for creators who want practical audio decisions, not abstract engineering theory.

01

Match a speech file to the right profile based on listener context.

02

Understand why audiobook comfort differs from podcast loudness.

03

Know when transcription cleanup should prioritize intelligibility over tone.

Speech map

Quick reference

Use this table as the first pass before choosing a profile or export path.

DestinationPriorityRecommended starting profile
PodcastVoice consistency and listener-friendly loudnessPodcast Voice
InterviewSpeaker balance and reduced room distractionsInterview Balance
AudiobookLong-form comfort and chapter consistencyAudiobook Focus
CourseClear lessons across varied recording setupsLecture Clean
VoiceoverPresence, articulation, and controlled dynamicsVoice Clarity
TranscriptionIntelligibility for speech recognitionTranscription Clean
Field notes

Lessons

Short, practical guidance you can apply while comparing previews and preparing a master.

01

Speech is judged by effort

Listeners should not have to ride the volume, strain to understand words, or get distracted by hiss and room tone. The best speech master feels stable and invisible.

  • Balance speakers before increasing overall loudness.
  • Treat sibilance and plosives before making the track brighter.
  • Avoid over-cleaning long-form narration because artifacts become tiring.
02

Different speech formats need different compromises

A short voiceover can be brighter and more forward. An audiobook should be smoother. A lecture may need more cleanup. A transcription file may trade natural tone for intelligibility.

  • Podcast Voice is a good default for published spoken content.
  • Audiobook Focus is more restrained for long listening.
  • Transcription Clean is for words first, tone second.
Before export

Practical checklist

Use these checks while comparing previews and preparing a full master.

  • Check the loudest speaker and the quietest speaker.
  • Listen through headphones for mouth noise and sibilance.
  • Use a custom profile when every episode needs the same house sound.
Related profiles

Good starting points

🎙️
Voice

Podcast Voice

Crisp, consistent speech with reduced sibilance.

Best for: Solo podcasts, interviews, narration, and spoken-word shows.

  • Brings voices forward without harsh s sounds.
  • Keeps dialogue consistent over long episodes.
  • Targets podcast-friendly loudness around -16 LUFS.
low-cutde-essercompressionloudness target
Try Podcast Voice
🎧
Voice

Interview Balance

Balances levels across multiple speakers.

Best for: Two or more voices with uneven levels.

  • Balances multiple speakers in one file.
  • Great for remote interviews and panels.
  • Helps listeners avoid volume changes between speakers.
speaker levelingnoise reductioncompression
Try Interview Balance
📖
Voice

Audiobook Focus

Comfortable, even narration for long listening.

Best for: Long-form narration, chapters, and spoken books.

  • Keeps narration steady across long chapters.
  • Tames sibilance without dulling tone.
  • Prioritizes comfort over excessive loudness.
low-cutlight compressionsmooth de-essing-18 LUFS target
Try Audiobook Focus
🏫
Education

Lecture Clean

Clean long talks with steady loudness.

Best for: Classroom audio, long talks, courses, and webinars.

  • Keeps long lectures intelligible.
  • Reduces room hum and background noise.
  • Works well for lessons, courses, and training content.
low-cutnoise reductionlong-form leveling
Try Lecture Clean