V4: add 6 postprocess modules + enhanced audio pipeline

- beat_sync: librosa-based beat detection with 3 shot-alignment modes
- deflicker: histogram/optical_flow temporal flicker removal + cross-shot harmonize
- lip_sync: multi-backend (musetalk/wav2lip/sadtalker/hallo2/basic) with face detection cascade
- mtv_sync: Demucs source separation + per-frame audio-video sync mapping
- style_harmonize: cross-shot color/brightness/texture consistency (LAB/k-means/VGG)
- video_extend: loop_blend, reverse_bounce, RIFE interpolation, VACE continuation
- audio: voice profiles, generate_narration_with_voice(), storyboard voice loading

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Author: charliewwdev

animatediff/postprocess/__init__.py

@@ -4,12 +4,40 @@
 Modules:
 - interpolation: RIFE frame interpolation for smoother motion
 - upscale: Real-ESRGAN anime super-resolution
-- audio: F5-TTS voice generation + BGM
+- audio: F5-TTS voice generation + BGM + per-character voice cloning
 - compositor: shot assembly, transitions, audio mixing
+- deflicker: temporal flicker removal within and across shots
+- beat_sync: beat detection and shot-to-music alignment (requires librosa)
+- mtv_sync: multi-track audio separation + frame-level video sync
+- video_extend: extend short video clips to longer duration
+- style_harmonize: cross-shot visual consistency
+- lip_sync: multi-backend lip synchronization (MuseTalk, Wav2Lip, SadTalker, Hallo2)
 """
 
 from animatediff.postprocess.interpolation import FrameInterpolator
 from animatediff.postprocess.upscale import VideoUpscaler
-from animatediff.postprocess.audio import AudioGenerator, VoiceProfile
+from animatediff.postprocess.audio import (
+    AudioGenerator,
+    VoiceProfile,
+    generate_narration_with_voice,
+    normalize_audio,
+    add_reverb,
+    crossfade_audio,
+    concatenate_audio,
+    get_audio_duration,
+    prepare_reference_audio,
+)
 from animatediff.postprocess.compositor import VideoCompositor
 from animatediff.postprocess.lipsync import LipSyncProcessor, apply_lipsync_to_shots
+from animatediff.postprocess.lip_sync import LipSyncer, FaceDetector
+from animatediff.postprocess.deflicker import VideoDeflicker
+from animatediff.postprocess.beat_sync import BeatAnalyzer, ShotBeatAligner, BeatInfo
+from animatediff.postprocess.mtv_sync import (
+    AudioStreamAnalyzer,
+    VideoAudioSync,
+    AudioStreams,
+    TimeSegment,
+    SyncMap,
+)
+from animatediff.postprocess.video_extend import VideoExtender
+from animatediff.postprocess.style_harmonize import StyleHarmonizer, StyleDescriptor