ComfyUI Node: Blur

Blur:

The Blur node is designed to apply a Gaussian blur effect to an image, which is a common technique in image processing to reduce noise and detail. This node is particularly useful for creating a smooth, soft appearance in your images, which can enhance the overall aesthetic or prepare the image for further processing steps. By adjusting the blur radius and sigma, you can control the intensity and spread of the blur effect, allowing for a wide range of artistic and practical applications. The Gaussian blur method ensures that the transition between blurred and non-blurred areas is smooth and natural, making it a versatile tool for AI artists.

Blur Input Parameters:

image

This parameter represents the input image that you want to apply the blur effect to. The image should be in the form of a tensor with dimensions corresponding to batch size, height, width, and channels.

blur_radius

The blur_radius parameter controls the size of the area around each pixel that will be averaged to produce the blur effect. A larger radius results in a more pronounced blur. The value must be an integer between 1 and 15, with a default value of 1. Increasing the blur radius will make the image appear softer and more out of focus.

sigma

The sigma parameter determines the standard deviation of the Gaussian distribution used to calculate the blur. A higher sigma value results in a wider spread of the blur effect, making the image appear more blurred. The value must be a float between 0.1 and 10.0, with a default value of 1.0. Adjusting the sigma allows for fine-tuning the smoothness and intensity of the blur.

Blur Output Parameters:

image

The output is the blurred version of the input image. This tensor will have the same dimensions as the input image but with the blur effect applied according to the specified blur_radius and sigma values. The blurred image can be used directly in your artwork or as a pre-processed input for further image manipulation tasks.

Blur Usage Tips:

• For a subtle blur effect, use a small blur_radius (e.g., 1 or 2) and a low sigma value (e.g., 0.5 to 1.0).
• To create a strong blur effect, increase both the blur_radius and sigma values. For example, a blur_radius of 10 and a sigma of 5.0 will produce a very soft and smooth image.
• Experiment with different combinations of blur_radius and sigma to achieve the desired artistic effect. Small changes in these parameters can significantly alter the appearance of the blurred image.

Blur Common Errors and Solutions:

"Expected 4-dimensional input for 4-dimensional weight"

• Explanation: This error occurs when the input image tensor does not have the correct dimensions (batch size, height, width, channels).
• Solution: Ensure that your input image tensor is correctly shaped with four dimensions. You may need to reshape or permute your tensor to match the expected format.

"Kernel size should be an odd number"

• Explanation: The kernel size, derived from the blur_radius, must be an odd number to properly center the blur effect.
• Solution: Verify that the blur_radius is set correctly. The node automatically calculates the kernel size as blur_radius * 2 + 1, so ensure that blur_radius is a positive integer.

"CUDA out of memory"

• Explanation: This error indicates that the GPU does not have enough memory to process the image with the specified blur parameters.
• Solution: Try reducing the blur_radius and sigma values to lower the computational load. Alternatively, you can process the image in smaller batches or use a machine with more GPU memory.