TTS支持参考音频逻辑

2025-10-31 15:50:35 +08:00
parent 7c3a49f55b
commit 2fe5e7c0fa
8 changed files with 435 additions and 316 deletions
--- a/pixelle_video/models/storyboard.py
+++ b/pixelle_video/models/storyboard.py
@@ -29,6 +29,7 @@ class StoryboardConfig:
    voice_id: str = "[Chinese] zh-CN Yunjian"     # Default voice
    tts_workflow: Optional[str] = None         # TTS workflow filename (None = use default)
    tts_speed: float = 1.2                     # TTS speed multiplier (1.0 = normal, >1.0 = faster)
+    ref_audio: Optional[str] = None            # Reference audio for voice cloning (only some workflows support this)
    
    # Image parameters
    image_width: int = 1024