Release v2.5.22: FFmpeg 10-bit video backend, MPS bicubic fix, cross-platform histogram matching

Note: index_select(out=) optimization removed as it caused color polarization on MPS; using simple indexing instead

README.md

@@ -36,6 +36,13 @@ We're actively working on improvements and new features. To stay informed:
 
 ## 🚀 Release Notes
 
+**2025.12.13 - Version 2.5.22**
+
+- **🎬 CLI: FFmpeg video backend with 10-bit support** - New `--video_backend ffmpeg` and `--10bit` flags enable x265 encoding with 10-bit color depth, reducing banding artifacts in gradients compared to 8-bit OpenCV output *(based on PR by [@thehhmdb](https://github.com/thehhmdb) - thank you!)*
+- **🍎 Fix: MPS bicubic upscaling compatibility** - Added CPU fallback for bicubic+antialias interpolation on PyTorch versions before 2.8.0, resolving RGBA alpha upscaling errors on Apple Silicon
+- **⚡ Fix: Cross-platform histogram matching** - Replaced scatter_ operation with argsort+index_select for improved reliability across CUDA, ROCm, and MPS backends
+- **🧹 MPS: Remove sync overhead** - Reverted unnecessary `torch.mps.synchronize()` calls introduced in v2.5.21 for consistent behavior with CUDA pipeline
+
 **2025.12.12 - Version 2.5.21**
 
 - **🛠️ Fix: GGUF dequantization error on MPS** - Resolved shape mismatch error introduced in 2.5.20 by skipping GGUF quantized buffers in precision conversion - these must remain in packed format for on-the-fly dequantization during inference

pyproject.toml

@@ -1,7 +1,7 @@
 [project]
 name = "seedvr2_videoupscaler"
 description = "SeedVR2 official ComfyUI integration: ByteDance-Seed's one-step diffusion-based video/image upscaling with memory-efficient inference"
-version = "2.5.21"
+version = "2.5.22"
 authors = [
     {name = "numz"},
     {name = "adrientoupet"}

src/utils/color_fix.py

@@ -506,15 +506,15 @@ def _histogram_matching_channel(source: Tensor, reference: Tensor, device: torch
         matched_sorted = reference_sorted[ref_indices]
         del source_quantiles, ref_indices, reference_sorted
     
-    del source_sorted
+    del source_sorted, source_flat
     
-    # Reconstruct using index_select with out param (portable, memory-efficient)
+    # Reconstruct using argsort (portable across CUDA/ROCm/MPS)
     inverse_indices = torch.argsort(source_indices)
     del source_indices
-    torch.index_select(matched_sorted, 0, inverse_indices, out=source_flat)
+    matched_flat = matched_sorted[inverse_indices]
     del matched_sorted, inverse_indices
     
-    return source_flat.reshape(original_shape)
+    return matched_flat.reshape(original_shape)
 
 
 def hsv_saturation_histogram_match(content_feat: Tensor, style_feat: Tensor, debug: Optional['Debug'] = None) -> Tensor:
@@ -756,11 +756,10 @@ def _histogram_match_1d(source: Tensor, reference: Tensor, device: torch.device)
     
     del source_sorted
     
-    # Reconstruct using index_select with out param (portable, memory-efficient)
-    matched = torch.empty_like(source)
+    # Reconstruct using argsort (portable across CUDA/ROCm/MPS)
     inverse_indices = torch.argsort(source_indices)
     del source_indices
-    torch.index_select(matched_sorted, 0, inverse_indices, out=matched)
+    matched = matched_sorted[inverse_indices]
     del matched_sorted, inverse_indices
     
     return matched

src/utils/constants.py

@@ -4,7 +4,7 @@ Only includes constants actually used in the codebase
 """
 
 # Version information
-__version__ = "2.5.21"
+__version__ = "2.5.22"
 
 import os
 import warnings