Visit ComfyUI Online for ready-to-use ComfyUI environment
Define and manage substeps in OCS framework for detailed AI art generation with configurable sampling control.
The OCS Substeps node is designed to define and manage substeps within the Overly Complicated Sampling (OCS) framework. This node allows you to specify the number of substeps and the method of sampling to be used, providing fine-grained control over the sampling process. By breaking down each step into multiple substeps, you can achieve more detailed and nuanced results, which is particularly useful for complex AI art generation tasks. The node is highly configurable, enabling you to tailor the sampling process to your specific needs and preferences, thereby enhancing the quality and precision of the generated outputs.
This parameter defines the number of substeps to use for each step. Essentially, it allows you to split a single step into multiple smaller steps, which can lead to more refined and detailed results. The minimum value for this parameter is 1, the maximum is 1000, and the default value is 1. Adjusting this parameter can significantly impact the granularity and quality of the sampling process.
This parameter specifies the sampling method to be used. It is a tuple that includes various sampler types, allowing you to choose the most appropriate one for your task. The tooltip for this parameter simply states "In other words, the sampler," indicating that it directly influences the sampling technique applied during the substeps.
This optional parameter allows you to connect another OCS Substeps node. When connected, the substeps will run in order, starting from the OCS Substeps node furthest from the OCS Group node. This can be useful for creating complex, multi-layered sampling processes.
This optional parameter lets you connect additional parameters, such as custom noise, from OCS Param or OCS MultiParam nodes. This provides further customization options for the sampling process, allowing you to fine-tune various aspects to achieve the desired results.
This optional text parameter block allows you to set custom parameters using YAML (recommended) or JSON. It supports multiline input and does not use dynamic prompts. The default value is a predefined YAML parameter set, but you can leave it blank if no additional customization is needed.
This output can be connected to another OCS Substeps node or an OCS Group node. It serves as the link between different stages of the sampling process, enabling the creation of complex, multi-step sampling workflows. The output essentially carries the configuration and results of the substeps, which can then be further processed or combined with other nodes.
substeps
parameter to find the optimal balance between performance and result quality. Higher values can lead to more detailed outputs but may also increase processing time.step_method
parameter to test various sampling techniques and determine which one best suits your specific project needs.params_opt
parameter to introduce custom noise or other parameters, which can add unique characteristics to your generated art.parameters
text block to fine-tune the sampling process with custom YAML or JSON settings, allowing for highly specific adjustments.substeps
parameter is outside the allowed range (1-1000).substeps
value is within the specified range. Adjust the value to be between 1 and 1000.step_method
is not recognized or supported by the node.step_method
is one of the supported sampler types listed in the node's documentation. Select a valid sampler type from the provided options.substeps_opt
parameter is expected to be connected to another OCS Substeps node but is not.substeps_opt
parameter to another OCS Substeps node to ensure the substeps run in the correct order.parameters
text block contain syntax errors or are not formatted correctly.© Copyright 2024 RunComfy. All Rights Reserved.