ComfyUI Node: Sigmas SetFloor

Sigmas SetFloor:

The Sigmas SetFloor node is designed to adjust a sequence of sigma values by setting a new minimum threshold, or "floor," for these values. This node is particularly useful in scenarios where you need to ensure that all sigma values in a sequence meet a certain minimum level, which can be crucial for maintaining stability in variance-locked Stochastic Differential Equation (SDE) sampling processes. By replacing any sigma values that fall below a specified floor with a new floor value, this node helps prevent issues that can arise from excessively low sigma values, such as numerical instability or undesirable sampling behavior. This adjustment ensures that the sigma values remain within a range that is conducive to effective sampling, thereby enhancing the overall robustness and reliability of the sampling process.

Sigmas SetFloor Input Parameters:

sigmas

The sigmas parameter represents the sequence of sigma values that you want to process. These values are crucial in controlling the variance during the sampling process. The node requires this input to identify which sigma values fall below the specified floor and need adjustment. The sigmas parameter is mandatory and must be provided for the node to function.

floor

The floor parameter specifies the current minimum threshold for the sigma values. Any sigma value in the sequence that is less than or equal to this floor will be replaced by the new_floor value. This parameter allows you to define the boundary below which sigma values are considered too low. The default value is 0.0291675, with a minimum of -10000 and a maximum of 10000, adjustable in steps of 0.01.

new_floor

The new_floor parameter defines the new minimum value that will replace any sigma values falling below or equal to the specified floor. This allows you to set a new baseline for the sigma values, ensuring they remain within a desired range. Like the floor parameter, the default value is 0.0291675, with a minimum of -10000 and a maximum of 10000, adjustable in steps of 0.01.

Sigmas SetFloor Output Parameters:

SIGMAS

The output SIGMAS is the adjusted sequence of sigma values after applying the new floor. This output ensures that all sigma values meet the new minimum threshold, thereby maintaining the desired level of variance control in the sampling process. The adjusted sigma values are returned in the same format as the input, ready for further processing or analysis.

Sigmas SetFloor Usage Tips:

• Use the floor parameter to identify and adjust sigma values that are too low for your specific application, ensuring stability in your sampling process.
• Adjust the new_floor parameter to set a new baseline for sigma values, which can help maintain consistency and prevent numerical issues during sampling.

Sigmas SetFloor Common Errors and Solutions:

"sigmas is not provided"

• Explanation: This error occurs when the sigmas input is not supplied to the node, which is required for its operation.
• Solution: Ensure that you provide a valid sequence of sigma values as input to the node.

"Invalid floor or new_floor value"

• Explanation: This error may arise if the floor or new_floor values are set outside their permissible range.
• Solution: Check that the floor and new_floor values are within the specified range of -10000 to 10000 and adjust them accordingly.