☁️BizyAir Apply ControlNet (OLD)
The legacy single-conditioning apply node
- conditioning
- control_net
- image
- CONDITIONING
BizyAir ships this node with "(OLD)" right in its display name, which is about as direct a hint as a node pack ever gives you: use the other one. This is the simple, original-style ControlNet apply - one conditioning in, a control net and reference image, a strength, one conditioning out. No positive/negative split, no start/end-percent window. It exists for backward compatibility with older workflows built before BizyAir added the advanced version, not because it's the better choice for anything new.
What it does, mechanically: ControlNet conditions generation on spatial structure pulled from a reference image - edges, depth, a pose skeleton, whatever the loaded ControlNet model was trained on - so the prompt still decides what gets generated while the ControlNet decides where it goes. This node takes an already-encoded conditioning, layers the ControlNet's spatial guidance on top of it at the given strength, and hands back a new conditioning carrying both. That's genuinely useful and still the core mechanism behind every modern ControlNet Union model - the limitation here isn't the concept, it's the missing controls.
What "(OLD)" costs you specifically: the biggest one is no start/end-percent window. The community's long-standing advice - and it's held up since the very first ControlNet release in 2023 - is to release the structural guidance partway through denoising rather than holding it for the whole sample, so composition gets locked in early and the model still gets to add its own detail in later steps. This node can't do that; the ControlNet's influence is on for the entire generation or not at all, at a single fixed strength. The other gap is the single-conditioning shape: modern workflows almost always want the ControlNet applied to positive and negative conditioning separately (that's what the Advanced version's positive/negative split is for), and this node just doesn't offer that.
Inputs that matter: conditioning, control_net (BIZYAIR_CONTROL_NET, from a BizyAir ControlNet loader), and image, all required. strength (default 1, range 0–10, step 0.01) is the only dial - how hard the spatial condition pushes against the prompt. Note the output is labeled CONDITIONING in the node's own display but is actually typed BIZYAIR_CONDITIONING, so it still only connects to other BizyAir nodes downstream, same as everywhere else in this pack.
When to actually use this instead of the Advanced version: honestly, close to never for new work. The only real case is reproducing an old workflow exactly as it was built, where introducing the extra positive/negative split and step-range controls would change behavior you're deliberately trying to keep fixed. For anything you're building fresh, reach for BizyAir_ControlNetApplyAdvanced instead - same underlying mechanism, strictly more control, no real downside.
Install: search "BizyAir" in ComfyUI Manager and install, git clone https://github.com/siliconflow/BizyAir.git into custom_nodes and restart, or comfy node install bizyair - one pack install covers both this node and its Advanced sibling. As with the rest of the pack, it's a cloud call, so register a BizyAir account and set your API key before anything runs.
Troubleshooting: if the ControlNet doesn't seem to be having any visible effect, try raising strength before assuming something's broken - the default of 1 is a reasonable starting point but not always enough depending on your image and control net pairing. A stall or unclear failure, as always in this pack, is worth an API-key check first.
Inputs (4)
| Name | Type | Default | Description |
|---|---|---|---|
| conditioning | BIZYAIR_CONDITIONING | — | |
| control_net | BIZYAIR_CONTROL_NET | — | |
| image | IMAGE | — | |
| strength | FLOAT | 1.000–10 | — |
Outputs (1)
| Name | Type | Description |
|---|---|---|
| CONDITIONING | BIZYAIR_CONDITIONING | — |