# Clone or navigate to the skill directory
cd skills/ai-voice-generator

# Install dependencies
pip install -r requirements.txt

# Configure API keys (choose your providers)
export ELEVENLABS_API_KEY="your-elevenlabs-key"     # For high-quality voices
export AZURE_SPEECH_KEY="your-azure-key"           # For multi-language
export OPENAI_API_KEY="your-openai-key"            # For fast generation
export GOOGLE_TTS_KEY="your-google-key"            # Optional
export COQUI_API_KEY="your-coqui-key"              # Optional: for local TTS

# Generate character dialogue (auto-selects best provider)
python scripts/generate.py dialogue "你好，我叫小明，很高兴认识你" --voice "young female" --emotion happy --style anime

# Generate with specific provider
python scripts/generate.py dialogue "Hello, my name is Alice" --provider elevenlabs --voice "young_female_1"

# Direct text-to-speech
python scripts/generate.py tts "这是一段测试语音" --provider azure --language zh-CN

# Clone a voice from audio sample
python scripts/clone.py reference.wav --name "main_character"

# List available voices
python scripts/list_voices.py

# Generate with cloned voice
python scripts/generate.py dialogue "新的对话内容" --voice "main_character" --emotion neutral

# Generate from dialogue script
python scripts/batch.py generate dialogue.txt --voice "young female" --output ./audio/

# Generate multiple languages
python scripts/batch.py translate dialogue_en.txt --languages zh,ja,ko --provider azure

from lib.voice_generator import AnimeVoiceGenerator

generator = AnimeVoiceGenerator()

# Generate character dialogue
audio = await generator.generate_dialogue(
    text="你好，我叫小明！",
    voice="young_female",
    emotion="happy",
    provider="elevenlabs"
)

# Clone a voice
voice_id = await generator.clone_voice(
    audio_sample="reference.wav",
    name="protagonist_voice"
)

# Generate with cloned voice
audio = await generator.generate_dialogue(
    text="今天天气真好呢！",
    voice_id=voice_id,
    emotion="cheerful"
)

# Batch generation
results = await generator.batch_generate(
    script="dialogue.txt",
    voice="young_male",
    emotions={"happy", "sad", "angry"}
)

export ELEVENLABS_API_KEY="your-key"

# Usage
python scripts/generate.py dialogue "Hello!" --provider elevenlabs --voice "young_female_1"

export AZURE_SPEECH_KEY="your-key"
export AZURE_SPEECH_REGION="eastus"

# Usage
python scripts/generate.py tts "你好世界" --provider azure --language zh-CN --voice "zh-CN-XiaoxiaoNeural"

export OPENAI_API_KEY="your-key"

# Usage
python scripts/generate.py tts "Hello world" --provider openai --voice "alloy"

export GOOGLE_TTS_KEY="your-key"

# Usage
python scripts/generate.py tts "こんにちは" --provider google --language ja --voice "ja-JP-Wavenet-A"

# No API key needed
python scripts/generate.py tts "Test" --provider coqui --model "tacotron2"

# configs/voice_presets.yaml
presets:
  young_female:
    voices:
      elevenlabs: "young_female_1"
      azure: "zh-CN-XiaoxiaoNeural"
      openai: "nova"
      google: "ja-JP-Wavenet-A"
    emotion_range: ["happy", "sad", "angry", "surprised", "neutral"]
    pitch: "+2st"
    speed: 1.0
    
  young_male:
    voices:
      elevenlabs: "young_male_1"
      azure: "zh-CN-YunxiNeural"
      openai: "alloy"
      google: "ja-JP-Wavenet-B"
    emotion_range: ["happy", "serious", "sad", "angry", "neutral"]
    pitch: "-1st"
    speed: 1.0
    
  older_female:
    voices:
      elevenlabs: "older_female_1"
      azure: "zh-CN-XiaoyouNeural"
      openai: "shimmer"
    emotion_range: ["gentle", "strict", "kind", "neutral"]
    pitch: "-1st"
    speed: 0.95
    
  robot:
    voices:
      elevenlabs: "robot_voice"
      azure: "zh-CN-XiaoshuangNeural"
    emotion_range: ["neutral", "serious"]
    pitch: "+3st"
    speed: 1.1

# configs/emotions.yaml
emotions:
  happy:
    stability: 0.5
    similarity: 0.75
    style: 0.5
    description: "Cheerful, upbeat tone"
    
  sad:
    stability: 0.7
    similarity: 0.8
    style: 0.2
    description: "Melancholic, gentle tone"
    
  angry:
    stability: 0.3
    similarity: 0.7
    style: 0.8
    description: "Intense, forceful tone"
    
  surprised:
    stability: 0.4
    similarity: 0.7
    style: 0.6
    description: "Excited, shocked tone"
    
  neutral:
    stability: 0.7
    similarity: 0.8
    style: 0.0
    description: "Calm, natural tone"

# configs/providers.yaml
providers:
  primary: elevenlabs
  fallback_order:
    - elevenlabs
    - azure
    - openai
    - google
    - coqui
    
  elevenlabs:
    api_url: https://api.elevenlabs.io/v1
    enabled: true
    voices_per_minute: 100
    
  azure:
    api_url: https://eastus.tts.speech.microsoft.com
    enabled: true
    voices_per_minute: 200
    
  openai:
    api_url: https://api.openai.com/v1
    enabled: true
    voices_per_minute: 50
    
  google:
    api_url: https://texttospeech.googleapis.com/v1
    enabled: true
    voices_per_minute: 300
    
  coqui:
    api_url: http://localhost:5002
    enabled: true
    voices_per_minute: 1000

# configs/audio_settings.yaml
output:
  format: mp3        # mp3, wav, ogg
  sample_rate: 44100 # 16000, 22050, 44100, 48000
  bitrate: 192k      # 128k, 192k, 256k, 320k
  channels: 1        # 1 (mono), 2 (stereo)
  
normalization:
  target_lufs: -16
  max_peak_dbfs: -3

from lib.voice_generator import AnimeVoiceGenerator

generator = AnimeVoiceGenerator()

# Generate with SSML for advanced control
audio = await generator.generate_ssml("""
<speak>
  <prosody rate="fast" pitch="+2st">
    大家好，我是小明！
  </prosody>
  <break time="500ms"/>
  <amazon:effect name="whispered">
    这是一个秘密...
  </amazon:effect>
</speak>
""")

# Create and use character voice profiles
profile = generator.create_voice_profile(
    name="protagonist",
    base_voice="young_female",
    pitch_offset=1.5,
    speed=0.95,
    custom_replacements={
        "小明": {"pronunciation": "xiǎo míng", "emphasis": True}
    }
)

# Generate with profile
audio = await generator.generate_with_profile(
    text="大家好！",
    profile=profile
)

# Apply effects to generated audio
from lib.voice_generator import AudioProcessor

processor = AudioProcessor()

# Add reverb
processed = processor.add_reverb(
    audio,
    room_size=0.5,
    wetness=0.3
)

# Normalize audio
normalized = processor.normalize(audio, target_lufs=-16)

# Apply fade in/out
faded = processor.apply_fade(audio, in_ms=500, out_ms=1000)

# dialogue.txt
Scene 1: 学校门口
[happy] 小明: 早上好！今天天气真好啊！
[neutral] 小红: 是啊，阳光明媚呢。
[surprised] 小明: 哇，你看那只小猫！
[cute] 小猫: 喵～

Scene 2: 教室
[serious] 老师: 同学们，上课了！

# Generate all dialogue
python scripts/batch.py generate dialogue.txt --output ./audio/ --format mp3

# Generate with specific voice
python scripts/batch.py generate dialogue.txt --voice "young_female" --emotion auto

# Generate with timestamps
python scripts/batch.py generate dialogue.txt --timestamps --output dialogue.srt

# Generate voice and sync with video
from lib.voice_generator import AnimeVoiceGenerator
from lib.video_generator import AnimeVideoGenerator

voice_gen = AnimeVoiceGenerator()
video_gen = AnimeVideoGenerator()

# Generate voice first
voice = await voice_gen.generate_dialogue(
    text="你好！我是小明。",
    voice="young_female",
    emotion="happy"
)

# Generate video
video = await video_gen.generate_from_text(
    prompt="anime girl saying hello",
    duration=voice.duration
)

# Sync in post-processing
# (Use video editor for precise lip-sync)

# Generate audio and subtitles together
python scripts/generate.py with_subs dialogue.txt --output ./project/