ACN_TimestepKeyframeInterpolation Node Documentation

Overview

The ACN_TimestepKeyframeInterpolation node is part of the ComfyUI-Advanced-ControlNet extension. This node allows users to create a series of Timestep Keyframes with interpolated strength values over a specified percentage range during the sampling process. It's a powerful tool for dynamically adjusting the influence of ControlNet elements across different sampling steps, enabling more precise and flexible control during image generation.

Functionality

This node creates a sequence of Timestep Keyframes that gradually change strength according to specified interpolation methods over a defined range of sampling steps. It helps in smoothly transitioning the effects of control elements, providing gradual changes rather than abrupt shifts.

Inputs

• prev_timestep_kf (optional): This input allows you to chain Timestep Keyframes together to form a cohesive schedule. The sequence is automatically sorted by their start percentages, with any overlapping start percentages being overwritten in favor of the newly added keyframe.

• cn_weights (optional): Defines the weights to apply to the ControlNet while this keyframe is in effect. If set to inherit, missing weights are obtained from previous keyframes in the schedule. However, if an override weight is specified in the Apply Advanced ControlNet node, it will take precedence.

• latent_keyframe (optional): Specifies latent keyframes for the ControlNet during this keyframe's activation. Missing latent keyframes can inherit values from previous keyframes if set to do so. Overrides, if set in the Apply Advanced ControlNet node, will take precedence.

• mask_optional (optional): Specifies attention masks to be applied to control elements, determining areas of the image affected by ControlNet, potentially with varying strengths if non-binary masks are used. Inherited from earlier keyframes if not set and inherit is enabled. Masks defined here are used in conjunction with any specified in the Apply Advanced ControlNet node.

• start_percent (required as a widget): The percentage of sampling steps at which the first Timestep Keyframe begins to apply its influence.

• end_percent (required as a widget): The percentage of sampling steps at which the last Timestep Keyframe's influence concludes.

• strength_start (required as a widget): Specifies the strength of the Timestep Keyframe at the beginning of the range.

• strength_end (required as a widget): Specifies the strength of the Timestep Keyframe at the end of the range.

• interpolation (required as a widget): Defines the method used for interpolation between the strength_start and strength_end values.

• intervals (required as a widget): The total number of keyframes to generate, with specific percentages applied to the start and end, and linearly distributed steps between.

• null_latent_kf_strength (required as a widget): The strength for latents not accounted for in the specified latent_keyframes.

• inherit_missing (required as a widget): Indicates whether the node should reuse values from preceding keyframes for optional settings not provided in this keyframe input.

• print_keyframes (required as a widget): When enabled, outputs the generated Timestep Keyframes for debugging purposes.

Outputs

• TIMESTEP_KF: This output provides the generated Timestep Keyframe, which can be linked into a larger Timestep Keyframe sequence or into a Timestep Keyframe input for further processing or application.

Usage in ComfyUI Workflows

The ACN_TimestepKeyframeInterpolation node is typically utilized when users require a smooth transition in strength or influence of ControlNet settings across the generation of a batch. It is especially useful in workflows where:

• Gradual transitions are desired over abrupt changes.
• Precise control over the impact of specific conditioning steps is needed.
• Complex scheduling of ControlNet influences over sampling steps must be achieved.

For instance, in a scenario where a user's artistic vision requires the progressive intensification of a stylistic element, this node can adjust the related parameters gradually and effectively.

Special Features and Considerations

• Interpolation Options: Users can select different interpolation methods, allowing flexibility in how strengths are calculated and applied between the start and end percentages.

• Inheritance: The node provides an inheritance feature which enables continuity by reusing values from previous keyframes, ensuring that only necessary parameters are changed while others remain consistent.

• Debugging Output: By enabling the print_keyframes option, users can better understand and debug the Timestep Keyframes generated, facilitating more precise adjustments and optimizations.

This node represents a key component in advanced scheduling and modulation of ControlNet effects, enhancing the capability to manipulate and refine outputs within the ComfyUI framework.