📷 FoW - Shot Weight

The dial for "close-up means close-up"

The Shot bucket is where the camera lives: Close-Up, Wide Shot, Low Angle, Rule of Thirds, Panning. These tags fight a constant battle with the subject and scene buckets - write "extreme close-up, full body, wide castle" and the model has no idea who wins. FoW - Shot Weight is the tiebreaker: one slider that scales the whole Shot conditioning slice, so you can make the camera language actually stick.

It ships in SirWillance's FoW_Suite_LIGHT, the free Light tier of the "Force of Will" suite, built by a self-taught dev over two months and licensed MIT. Everything about the Light tier is "one simple knob per concern," and this node is exactly that. It's also the most useful of the pack's weight nodes, because shot vs. content conflict is a real, recurring problem that usually ends in people rewriting prompts.

How it works

Standard suite mechanism:

weighted.append((tensor.clone() * Weight, metadata))

The Shot conditioning (normally the Shot Conditioning output from a Shot Category / ShotAgent node) comes in, gets multiplied by the Weight value, and goes out scaled as one block. 1.0 = no change, below 1 = damped, above 1 = amplified. No token-level weighting in Light tier - Close-Up and Dutch Angle move together.

The inputs that matter

• RAW Condition (CONDITIONING) - the Shot conditioning to scale. Wire it from a Shot Agent node's Shot Conditioning output, or the positive_shot output of Prompt Refiner Light.
• Weight (FLOAT, 0–2, default 1, step 0.1) - the dial.

The single output, W8ed Shot Condition (CONDITIONING), feeds a positive Fusion node or a KSampler.

Installing it

The suite is one install:

cd ComfyUI/custom_nodes
git clone https://github.com/SirWillance/FoW_Suite_LIGHT

Restart ComfyUI, and all nodes land under "🧿 FoW - Light". ComfyUI Manager also has it - search "Force of Will Suite Light". No extra dependencies, no model downloads.

Where people get burned

The classic move is cranking Shot to 2.0 because "the model keeps ignoring my close-up" - and sometimes that genuinely fixes it, but more often the shot and content buckets were simply contradicting each other (close-up + full body + wide shot in one prompt). A weight dial can't resolve a contradiction; it just picks a winner. Clean the contradiction first, then weight. And the standing caveat: on distilled/turbo models at CFG 1, conditioning scaling is far weaker than on SDXL-lineage checkpoints at normal CFG. On the models where it works, though, this is the one weight node in the suite I'd actually reach for - it solves a real, named problem instead of just adding gain.