Visit ComfyUI Online for ready-to-use ComfyUI environment
Seamlessly blend latent representations with various operations for creating unique AI-generated art.
The Blend Latents (PPF Noise) node is designed to seamlessly combine two latent representations, or "latents," using various blending operations. This node is particularly useful for AI artists who want to merge different latent spaces to create unique and complex visual outputs. By adjusting parameters such as blend ratio, blend strength, and blending mode, you can achieve a wide range of effects, from subtle transitions to dramatic transformations. The node also supports the use of masks to control the blending process more precisely, allowing for targeted blending in specific areas of the latent space. This flexibility makes it a powerful tool for generating diverse and intricate AI-generated art.
This parameter represents the first latent input that you want to blend. It is a tensor containing the latent representation of an image or other data. The shape of this tensor must match that of latent_b
to ensure a proper blend.
This parameter represents the second latent input that you want to blend with latent_a
. Like latent_a
, it is a tensor containing the latent representation of an image or other data. The shapes of latent_a
and latent_b
must be identical.
This parameter specifies the blending operation to be used. Common operations include "add," "subtract," "multiply," and other blending modes. The choice of operation will significantly impact the resulting blend.
This parameter controls the proportion of latent_a
and latent_b
in the final blend. A value of 0.5 means an equal mix of both latents, while values closer to 0 or 1 will favor one latent over the other. The range is typically from 0.0 to 1.0.
This parameter determines the intensity of the blending effect. Higher values result in a more pronounced blend, while lower values produce a subtler effect. The range is usually from 0.0 to 1.0.
An optional parameter that allows you to apply a mask to control the blending process. The mask should have the same spatial dimensions as the latents and can be used to blend specific areas selectively.
This optional parameter, when set to 'true,' will apply a noise mask to the blended latent. This can add an additional layer of complexity and texture to the blend.
An optional parameter that, when set to 'true,' normalizes the blended latent to a specified range. This can help in maintaining consistent output values.
This optional parameter sets the minimum value for clamping the blended latent. It ensures that the output values do not fall below this threshold.
This optional parameter sets the maximum value for clamping the blended latent. It ensures that the output values do not exceed this threshold.
An optional parameter that, when set to 'true,' generates a preview of the blended latent in RGB format. This can be useful for visualizing the blend before finalizing it.
This output parameter contains the blended latent representation. It is a tensor that combines the input latents according to the specified blending parameters and operations.
If the set_noise_mask
parameter is set to 'true,' this output will include the noise mask applied to the blended latent. This can be useful for further processing or analysis.
mask
parameter to blend specific areas of the latents selectively. This can help in creating more complex and targeted visual effects.blend_ratio
and blend_strength
parameters to fine-tune the blending effect. Small changes in these values can lead to significantly different results.latent2rgb_preview
option to visualize the blend before finalizing it. This can save time and help in making more informed adjustments.{latent_a.shape}
, b {latent_b.shape}
latent_a
and latent_b
do not match.{', '.join(list(blending_modes.keys()))}
operation
parameter.normalize
parameter is set to 'true,' but the normalization function is not available.© Copyright 2024 RunComfy. All Rights Reserved.