EasyControlnet++(controlnetPlusPlus)

EasyControlnet++ (controlnetPlusPlus) applies a ControlNet to an EasyUse pipeline and returns the updated pipeline plus the positive and negative conditionings with the ControlNet applied. It is a pipe-based loader node, so it fits directly into EasyUse workflows without manually unpacking and repacking the pipeline.

Inputs

Required

• pipe (PIPE_LINE) — The EasyUse pipeline carrying the model, conditioning, latent, and other workflow state. The ControlNet is applied to the conditionings stored in this pipe.
• image (IMAGE) — The control image for the ControlNet: e.g. an edge map, depth map, pose skeleton, or other preprocessed guide matching the selected ControlNet’s expected modality.
• control_net_name (COMBO) — Selects a ControlNet model available in the models/controlnet folder. This is used to load the ControlNet unless a direct ControlNet object is supplied.

Optional

• control_net (CONTROL_NET) — A directly supplied ControlNet object from another node. When connected, this can be used in place of loading a model by control_net_name.
• strength (FLOAT, default 1, range 0.0–10.0) — Overall strength of the ControlNet effect. 0 disables it; higher values increase its influence.
• start_percent (FLOAT, default 0, range 0.0–1.0) — The denoising progress, as a percentage, at which the ControlNet effect starts. 0 means from the beginning of sampling.
• end_percent (FLOAT, default 1, range 0.0–1.0) — The denoising progress, as a percentage, at which the ControlNet effect ends. 1 means it remains active until the final step.
• scale_soft_weights (FLOAT, default 1, range 0.0–1.0) — Scales the soft-weight channels of the ControlNet. Lower values attenuate these channels; 0 disables them and 1 uses their full value. This is useful for ControlNets that expose per-channel or soft-weight behavior.
• union_type (COMBO) — Selects the control mode for a Union ControlNet. The available choices are populated by the UI and depend on the selected ControlNet model; it is not needed for regular single-mode ControlNets.

Outputs

• PIPE_LINE — The updated EasyUse pipeline. Pass this to later EasyUse nodes so the ControlNet-conditioned state is preserved.
• CONDITIONING — The positive conditioning with the ControlNet applied.
• CONDITIONING — The negative conditioning with the ControlNet applied.

The two CONDITIONING outputs are the positive conditioning first, then the negative conditioning.

Usage Notes

• Connect a preprocessed guide image matching the ControlNet’s expected control type. The node does not do automatic preprocessing of images such as depth or pose before applying the ControlNet.
• Use start_percent and end_percent to restrict the ControlNet to a specific part of the denoising schedule. For example, start_percent=0 and end_percent=0.5 only applies the ControlNet during the first half of sampling.
• The optional control_net input lets you load a ControlNet elsewhere and pass it in directly, which can avoid reloading the same model multiple times in a workflow.
• The pipe output can be used to continue the EasyUse workflow, while the two CONDITIONING outputs allow manual connection to samplers when you need more granular control.
• For Union ControlNets, set union_type to the specific control mode you want to activate. For ordinary ControlNet models, leave it at its default state.