ComfyUI Node: 😼> Noise From Image

😼> Noise From Image:

The "Noise From Image" node is designed to add noise to an image, enhancing its texture and visual complexity. This node is particularly useful for AI artists looking to create more dynamic and interesting images by introducing controlled randomness. The node applies a combination of elastic transformations and noise blending techniques to achieve a unique effect. It allows for fine-tuning various parameters such as noise intensity, blend rate, and saturation correction, providing you with the flexibility to achieve the desired artistic outcome. By manipulating these settings, you can create anything from subtle texture enhancements to dramatic, noise-infused visuals.

😼> Noise From Image Input Parameters:

image

This parameter represents the input image to which noise will be added. The image should be in a tensor format, typically with dimensions corresponding to batch size, channels, height, and width.

magnitude

The alpha value for the elastic transform, which determines the magnitude of displacements in the image. Higher values result in more pronounced distortions. This parameter accepts a float value.

smoothness

The sigma value for the elastic transform, which controls the smoothness of the displacements. A higher sigma value results in smoother transformations. This parameter accepts a float value.

noise_intensity

This parameter controls the intensity of the noise added to the image. It acts as a multiplier for the generated noise, with higher values resulting in more intense noise. This parameter accepts a float value.

noise_resize_factor

A multiplier that enlarges the generated noise. This parameter helps in scaling the noise to match the dimensions of the input image. It accepts an integer value.

noise_blend_rate

The blend rate between the elastic transformation and the noise. This parameter determines how much of the final image is composed of the transformed image versus the added noise. It accepts a float value.

saturation_correction

This parameter adjusts the saturation of the final image. A value of 1.0 means no correction, while values greater or less than 1.0 increase or decrease the saturation, respectively. This parameter accepts a float value.

blend_mode

Specifies the blending mode used to combine the noise with the image. Different modes can produce various artistic effects. This parameter accepts a string value.

blend_rate

The strength of the blending between the original image and the noise. Higher values result in a stronger blend. This parameter accepts a float value.

vae_opt

An optional parameter for a Variational Autoencoder (VAE) that can be used to encode the final image. This parameter is typically used for further processing in generative models.

😼> Noise From Image Output Parameters:

result

The final image with added noise and transformations. This output is a tensor that represents the modified image, ready for further processing or display.

samples

A dictionary containing the latent representation of the final image if a VAE is used. This can be useful for generative tasks or further image manipulation.

😼> Noise From Image Usage Tips:

• Experiment with different values for noise_intensity and noise_blend_rate to achieve a balance between subtle texture and dramatic noise effects.
• Use the saturation_correction parameter to fine-tune the color vibrancy of your final image, especially if the added noise affects the overall color balance.
• Try different blend_mode settings to explore various artistic effects and find the one that best suits your creative vision.

😼> Noise From Image Common Errors and Solutions:

Image tensor dimensions mismatch

• Explanation: The input image tensor dimensions do not match the expected format.
• Solution: Ensure that your input image tensor has the correct dimensions, typically in the format (batch size, channels, height, width).

Invalid parameter values

• Explanation: One or more input parameters have values outside the acceptable range.
• Solution: Double-check the parameter values to ensure they fall within the specified ranges or types. For example, noise_intensity should be a float, and blend_mode should be a valid string.

VAE encoding failure

• Explanation: The optional VAE parameter is not functioning correctly.
• Solution: Verify that the VAE model is properly loaded and compatible with the input image dimensions. If not needed, consider omitting the vae_opt parameter.