Fix cropping logic bug and update defaults/documentation
Critical fixes: - Fixed inverted cropping logic that caused odd dimensions and math errors * When original is wider: now correctly resizes by height and crops width * When original is taller: now correctly resizes by width and crops height * This fixes the 1792x1024 → 1280x731 bug (now correctly becomes 1280x736) Configuration updates: - Changed default multiple_of from 8 to 2 (more appropriate for modern models) - Updated all SDXL-specific references to be model-agnostic - Added mentions of modern models (Flux, Stable Diffusion, etc.) Documentation improvements: - Updated README to remove dated SDXL references - Updated tooltips and descriptions to be model-agnostic - Changed example workflows to use generic "diffusion model" terminology
This commit is contained in:
Claude
@@ -9,7 +9,7 @@ A [ComfyUI](https://github.com/comfyanonymous/ComfyUI) node that intelligently r
|
||||
- **Intelligent Image Resizing**: Automatically resizes images using high-quality bilinear interpolation
|
||||
- **Aspect Ratio Preservation**: Maintains original aspect ratio through smart cropping when needed
|
||||
- **Flexible Constraint Modes**: Choose between prioritizing minimum resolution or strictly enforcing maximum limits
|
||||
- **Multiple Alignment**: Ensures dimensions are multiples of specified values (e.g., 8 for SDXL, 16/32/64 for performance)
|
||||
- **Multiple Alignment**: Ensures dimensions are multiples of specified values (e.g., 2, 8, 16, 32, 64 for performance)
|
||||
- **Smart Cropping**: Optional cropping with configurable position (center, top, bottom, left, right)
|
||||
- **Dual Outputs**: Provides both resized and original images for workflow flexibility
|
||||
- **Comprehensive Validation**: Input validation prevents invalid configurations
|
||||
@@ -19,7 +19,7 @@ A [ComfyUI](https://github.com/comfyanonymous/ComfyUI) node that intelligently r
|
||||
|
||||
Many AI models have strict dimension requirements:
|
||||
- **Image-to-Video models** often require exact resolutions (e.g., 1024x576, 768x768)
|
||||
- **SDXL** works best with dimensions divisible by 8
|
||||
- **Modern diffusion models** (Flux, Stable Diffusion, etc.) typically work best with dimensions divisible by 2, 8, or higher
|
||||
- **Some models** require dimensions divisible by 16, 32, or 64 for optimal performance
|
||||
- **VRAM constraints** may require strict maximum resolution limits
|
||||
|
||||
@@ -33,9 +33,9 @@ This node handles all these requirements intelligently, ensuring your images are
|
||||
### Resolution Constraints
|
||||
- **min_res** (default: 704, range: 1-65536): Minimum resolution in pixels for both width and height. Images with any dimension smaller than this will be upscaled to meet the requirement.
|
||||
- **max_res** (default: 1280, range: 1-65536): Maximum resolution in pixels for both width and height. Images with any dimension larger than this will be downscaled.
|
||||
- **multiple_of** (default: 8, range: 1-256): Ensures output dimensions are multiples of this number. Common values:
|
||||
- `8` - Standard for SDXL and most Stable Diffusion models
|
||||
- `16` - Some models with stricter alignment requirements
|
||||
- **multiple_of** (default: 2, range: 1-256): Ensures output dimensions are multiples of this number. Common values:
|
||||
- `2` - Standard for most diffusion models (Flux, Stable Diffusion, etc.)
|
||||
- `8` or `16` - Some models with stricter alignment requirements
|
||||
- `32` or `64` - Optimal for certain architectures and performance
|
||||
- `1` - Disable rounding (use exact calculated dimensions)
|
||||
|
||||
@@ -74,7 +74,7 @@ The node provides six outputs for maximum workflow flexibility:
|
||||
1. Add the **Constrain Resolution** node to your workflow
|
||||
2. Connect your image source to the `image` input
|
||||
3. Set your desired `min_res` and `max_res` constraints
|
||||
4. Set `multiple_of` based on your model requirements (8 for SDXL, 16/32/64 for others)
|
||||
4. Set `multiple_of` based on your model requirements (2 for most models, 8/16/32/64 for specific architectures)
|
||||
5. Choose your `constraint_mode` (keep default for most cases)
|
||||
6. Keep `crop_as_required` enabled for exact dimensions (recommended)
|
||||
7. Connect the `resized_image` output to your next node (resize node, image-to-video, etc.)
|
||||
@@ -91,12 +91,12 @@ Settings: min_res=768, max_res=1024, multiple_of=8, crop_as_required=True
|
||||
```
|
||||
Ensures images meet exact dimension requirements for video generation models.
|
||||
|
||||
#### SDXL Image-to-Image
|
||||
#### Diffusion Model Image-to-Image
|
||||
```
|
||||
Load Image → Constrain Resolution → SDXL Upscale/I2I
|
||||
Settings: min_res=704, max_res=1280, multiple_of=8, constraint_mode="Prioritize Min Resolution"
|
||||
Load Image → Constrain Resolution → Model Upscale/I2I
|
||||
Settings: min_res=704, max_res=1280, multiple_of=2, constraint_mode="Prioritize Min Resolution"
|
||||
```
|
||||
Optimizes images for SDXL processing while maintaining quality.
|
||||
Optimizes images for diffusion model processing while maintaining quality.
|
||||
|
||||
#### Batch Processing with VRAM Limits
|
||||
```
|
||||
|
||||
@@ -40,7 +40,7 @@ class ConstrainResolution(io.ComfyNode):
|
||||
"💡 USAGE TIPS:\n"
|
||||
"• Use 'Prioritize Min Resolution' to ensure images are never too small (may exceed max on one dimension)\n"
|
||||
"• Use 'Prioritize Max Resolution (Strict)' for hard VRAM limits (may go below min on one dimension)\n"
|
||||
"• Set 'Multiple Of' to 8 for SDXL, 16 for some models, or 64 for optimal performance\n"
|
||||
"• Set 'Multiple Of' to 2 for most models, or higher values (8, 16, 32, 64) for optimal performance\n"
|
||||
"• 'Crop as Required' is enabled by default for immediate compatibility with strict dimension requirements\n"
|
||||
"• The node outputs both the resized image and the original for workflow flexibility"
|
||||
),
|
||||
@@ -68,12 +68,12 @@ class ConstrainResolution(io.ComfyNode):
|
||||
),
|
||||
io.Int.Input(
|
||||
"multiple_of",
|
||||
default=8,
|
||||
default=2,
|
||||
min=1,
|
||||
max=256,
|
||||
tooltip=(
|
||||
"Ensures output dimensions are multiples of this number. "
|
||||
"Common values: 8 (SDXL), 16 (some models), 32 or 64 (optimal performance). "
|
||||
"Common values: 2 (most models), 8, 16, 32, or 64 (optimal performance). "
|
||||
"Set to 1 to disable rounding."
|
||||
)
|
||||
),
|
||||
@@ -349,20 +349,22 @@ class ConstrainResolution(io.ComfyNode):
|
||||
aspect_ratio_deviation = abs((final_aspect_ratio - original_aspect_ratio) / original_aspect_ratio * 100)
|
||||
|
||||
if aspect_ratio_deviation > 0.1: # If more than 0.1% deviation
|
||||
# Resize to the dimension that needs to be larger, then crop
|
||||
# Resize to preserve aspect ratio, making one dimension larger than target, then crop
|
||||
if width / height > target_width / target_height:
|
||||
# Original is wider, resize based on width and crop height
|
||||
intermediate_height = int(target_width / (width / height))
|
||||
resized_image = cls.resize_image(image, target_width, intermediate_height)
|
||||
else:
|
||||
# Original is taller, resize based on height and crop width
|
||||
# Original is wider (more landscape), resize based on height and crop width
|
||||
# This ensures height matches target, width will be larger and cropped
|
||||
intermediate_width = int(target_height * (width / height))
|
||||
resized_image = cls.resize_image(image, intermediate_width, target_height)
|
||||
else:
|
||||
# Original is taller (more portrait), resize based on width and crop height
|
||||
# This ensures width matches target, height will be larger and cropped
|
||||
intermediate_height = int(target_width / (width / height))
|
||||
resized_image = cls.resize_image(image, target_width, intermediate_height)
|
||||
|
||||
# Crop to exact target dimensions
|
||||
resized_image = cls.crop_image(resized_image, target_width, target_height, crop_position)
|
||||
|
||||
print(f"Image cropped from center to achieve exact dimensions {target_width}x{target_height}")
|
||||
print(f"Image cropped to achieve exact dimensions {target_width}x{target_height}")
|
||||
|
||||
# Log aspect ratio deviation warning if significant
|
||||
if original_aspect_ratio > 0 and not crop_as_required:
|
||||
|
||||
Reference in New Issue
Block a user