ComfyUI Node: LayerUtility: H/L Frequency Detail Restore

LayerUtility: H/L Frequency Detail Restore:

The LayerUtility: HLFrequencyDetailRestore node is designed to enhance and restore image details by manipulating high and low-frequency components. This node is particularly useful for AI artists who want to refine the details in their images, ensuring that high-frequency details are preserved while low-frequency noise is minimized. The node works by blending a detailed image with a background image, applying Gaussian blur to control the frequency components, and optionally using a mask to selectively apply these effects. This process helps in achieving a balanced and visually appealing result, making it an essential tool for image enhancement and restoration tasks.

LayerUtility: H/L Frequency Detail Restore Input Parameters:

image

This parameter represents the main image to which the detail restoration will be applied. It is the base layer that will be enhanced with high-frequency details from the detail image. The image should be provided in a compatible format, typically as a tensor.

detail_image

The detail_image parameter is the source of high-frequency details that will be blended into the main image. This image should contain the fine details that you want to preserve and enhance in the final output. It is also provided as a tensor.

keep_high_freq

This integer parameter controls the level of Gaussian blur applied to the detail image to isolate high-frequency components. A higher value results in more blur, which means fewer high-frequency details are kept. The default value is 64, with a minimum of 0 and a maximum of 1023.

erase_low_freq

This integer parameter determines the level of Gaussian blur applied to the main image to reduce low-frequency components. A higher value results in more blur, effectively erasing more low-frequency noise. The default value is 32, with a minimum of 0 and a maximum of 1023.

mask_blur

The mask_blur parameter specifies the amount of blur applied to the mask, if provided. This helps in smoothing the edges of the mask, ensuring a more seamless blend between the detailed and background images. The default value is 16, with a minimum of 0 and a maximum of 1023.

mask

This optional parameter allows you to provide a mask that defines the areas where the detail restoration should be applied. The mask should be in a compatible format, typically as a tensor. If not provided, the entire image will be processed.

LayerUtility: H/L Frequency Detail Restore Output Parameters:

image

The output parameter is the enhanced image with restored high-frequency details and minimized low-frequency noise. This image is returned as a tensor and represents the final result of the detail restoration process.

LayerUtility: H/L Frequency Detail Restore Usage Tips:

• To achieve the best results, experiment with different values for keep_high_freq and erase_low_freq to find the optimal balance between preserving details and reducing noise.
• Use the mask parameter to selectively apply detail restoration to specific areas of the image, such as focusing on facial features while leaving the background untouched.
• Adjust the mask_blur parameter to smooth the edges of the mask, ensuring a more natural blend between the detailed and background areas.

LayerUtility: H/L Frequency Detail Restore Common Errors and Solutions:

Image tensor dimensions mismatch

• Explanation: This error occurs when the dimensions of the input images (image and detail_image) do not match.
• Solution: Ensure that both the main image and the detail image have the same dimensions before processing.

Invalid mask format

• Explanation: This error happens when the provided mask is not in the expected format or dimensions.
• Solution: Verify that the mask is a tensor and has the same dimensions as the input images. Convert the mask to the correct format if necessary.

High-frequency and low-frequency values out of range

• Explanation: This error occurs when the values for keep_high_freq or erase_low_freq are outside the allowed range.
• Solution: Ensure that the values for keep_high_freq and erase_low_freq are within the specified range (0 to 1023). Adjust the values accordingly.