A custom node designed for ComfyUI, this tool facilitates the generation of empty latent variables specifically for Stable Cascade. It enhances the user experience by providing customizable options for aspect ratios and dimensions.
- Allows for a default increment of 64 in width and height, enabling consistent scaling.
- Features a locking mechanism for aspect ratios, ensuring proportional adjustments when one dimension is modified.
- Supports separate batch size configurations for stages, enhancing functionality when using the Latent From Batch node.
Context
This custom node serves as an enhancement to the ComfyUI framework, specifically tailored for generating empty latents utilized in Stable Cascade workflows. Its primary purpose is to streamline the process of setting up latent variables, making it easier for users to manage dimensions and aspect ratios.
Key Features & Benefits
One of the standout features is the default increment of 64 pixels for both width and height, which standardizes the dimensions for users. The aspect ratio locking feature ensures that when one dimension is modified, the other adjusts automatically, maintaining the intended proportions without additional calculations. Furthermore, the ability to separate batch sizes for different stages improves the usability of the Latent From Batch node, allowing for more precise control over processing.
Advanced Functionalities
The node includes a runtime feature that allows users to switch the width and height during execution, although this option is not visible within the node's interface. This capability provides flexibility in adjusting dimensions dynamically based on specific requirements during the processing phase.
Practical Benefits
By incorporating this custom node into their workflow, ComfyUI users can achieve greater control over latent variable generation, leading to enhanced efficiency and quality in their AI art projects. The straightforward interface and intuitive features reduce the complexity involved in managing dimensions, ultimately improving the overall user experience.
Credits/Acknowledgments
This tool was developed by Guillaume-Fgt, and it is available under an open-source license on GitHub. Users can access the repository for further contributions or updates.