Tts Openai

TTS (Text-to-Speech) Skill

Generate professional voiceover audio using OpenAI's TTS API.

Prerequisites

• OpenAI API key set as environment variable: OPENAI_API_KEY
• curl installed (standard on most systems)

Available Voices

Available Models

Generate Voiceover

Using curl (Recommended)

# Generate voiceover
curl https://api.openai.com/v1/audio/speech \
  -H "Authorization: Bearer $OPENAI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "tts-1-hd",
    "input": "Your script text goes here. This will be converted to speech.",
    "voice": "nova"
  }' \
  --output output/voiceover.mp3

For Long Scripts

Split into chunks if text is very long (max ~4000 characters per request):

# Generate part 1
curl https://api.openai.com/v1/audio/speech \
  -H "Authorization: Bearer $OPENAI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "tts-1-hd",
    "input": "First part of the script...",
    "voice": "nova"
  }' \
  --output output/vo_part1.mp3

# Generate part 2
curl https://api.openai.com/v1/audio/speech \
  -H "Authorization: Bearer $OPENAI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "tts-1-hd",
    "input": "Second part of the script...",
    "voice": "nova"
  }' \
  --output output/vo_part2.mp3

# Combine parts
ffmpeg -i "concat:output/vo_part1.mp3|output/vo_part2.mp3" \
  -acodec copy output/voiceover_combined.mp3

Audio Post-Processing

Normalize Audio Levels

ffmpeg -i output/voiceover.mp3 \
  -filter:a loudnorm=I=-16:TP=-1.5:LRA=11 \
  output/voiceover_normalized.mp3

Add Background Music

# Mix voiceover with background music (music at 15% volume)
ffmpeg -i output/voiceover.mp3 -i assets/music/background.mp3 \
  -filter_complex "[1:a]volume=0.15[bg];[0:a][bg]amix=inputs=2:duration=first" \
  output/voiceover_with_music.mp3

Get Audio Duration

ffprobe -v error -show_entries format=duration \
  -of default=noprint_wrappers=1:nokey=1 \
  output/voiceover.mp3

Voice Selection Guide

Best Practices

1. Clean the script: Remove stage directions, keep only spoken text
2. Add punctuation: Periods and commas create natural pauses
3. Use SSML-like formatting: "..." creates pauses in speech
4. Test voices: Generate samples before full production
5. Normalize audio: Ensure consistent levels
6. Save originals: Keep unprocessed audio for re-editing

Cost Estimation

(Based on ~150 words/min, ~6 chars/word = ~900 chars/min)

Output Files

Save audio files to:

• Raw TTS: output/voiceover_{project}.mp3
• Normalized: output/voiceover_{project}_normalized.mp3
• With music: output/voiceover_{project}_final.mp3