Hildegard Plan:
The Hildegard Plan node is designed to optimize the process of upscaling images by calculating the most efficient dimensions and tile grid configuration. It ensures that the upscaled image meets specific criteria for overlap and scale factor, which are crucial for maintaining image quality and consistency. This node does not rely on a model for upscaling but instead uses a chosen scaling method to achieve the desired results. By snapping tile dimensions to specific multiples, it ensures alignment and precision, making it an essential tool for artists looking to enhance their images while preserving detail and structure.
Hildegard Plan Input Parameters:
image
This parameter represents the input image that you wish to upscale. It serves as the base for all subsequent calculations and transformations performed by the node.
tile_width
This parameter defines the width of each tile in the grid. It has a default value of 1536 pixels, with a minimum value set by LATENT_ALIGN and a maximum of 8192 pixels. Adjusting this parameter affects the granularity of the tiling process and can influence the final image's detail and processing time.
tile_height
Similar to tile_width, this parameter specifies the height of each tile. It shares the same default, minimum, and maximum values as tile_width, ensuring that tiles are consistently shaped and aligned.
min_overlap
This parameter determines the minimum overlap between tiles, which is crucial for seamless blending and continuity in the upscaled image. It offers various options, with "1/32 Tile" as the default, allowing you to control how much adjacent tiles share in terms of content.
min_scale_factor
This parameter sets the minimum scale factor for the upscaling process, with a default value of 3.0. It ranges from 1.0 to 8.0, allowing you to specify how much larger the upscaled image should be compared to the original. A higher scale factor results in a larger image but may require more processing power.
tile_order
This parameter defines the order in which tiles are processed, with options such as "spiral" as the default. The order can impact the efficiency and visual outcome of the upscaling process, especially in terms of how details are prioritized and rendered.
scaling_method
This parameter specifies the method used for upscaling the image. The default method is "lanczos," which is known for its ability to preserve detail and minimize artifacts. Choosing the right scaling method can significantly affect the quality of the final image.
Hildegard Plan Output Parameters:
IMAGE
This output is the upscaled version of the input image. It reflects the dimensions and quality enhancements determined by the node's calculations and parameters, providing a higher-resolution version suitable for further artistic use or display.
dac_data
This output contains the data necessary for processing the tile grid, including information about overlaps and ordering. It is essential for ensuring that subsequent nodes in the workflow can accurately interpret and manipulate the upscaled image.
Hildegard Plan Usage Tips:
- To achieve the best results, experiment with different
tile_widthandtile_heightsettings to find the optimal balance between detail and processing time for your specific image. - Use the
min_overlapparameter to ensure smooth transitions between tiles, especially when working with images that have intricate patterns or textures. - Select the
scaling_methodthat best suits your artistic goals; for instance, "lanczos" is ideal for maintaining sharpness, while other methods might offer different aesthetic qualities.
Hildegard Plan Common Errors and Solutions:
"Invalid tile dimensions"
- Explanation: This error occurs when the specified
tile_widthortile_heightdoes not align with the required multiples set byLATENT_ALIGN. - Solution: Adjust the tile dimensions to ensure they are multiples of
LATENT_ALIGN.
"Scale factor out of range"
- Explanation: The
min_scale_factoris set outside the allowable range of 1.0 to 8.0. - Solution: Ensure that the scale factor is within the specified range to avoid this error.
"Unsupported scaling method"
- Explanation: The chosen
scaling_methodis not recognized by the node. - Solution: Verify that the scaling method is one of the supported options, such as "lanczos".
