Camera Settings

Camera Settings is one of those "why did nobody make this sooner" nodes. It's a single dropdown of camera types - Medium shot, Close-up, Bird's eye view, Over-the-shoulder, and eleven more - and it turns your selection into a full sentence describing that shot, ready to drop into a prompt.

The whole idea is that framing descriptions are annoying to type well and easy to get wrong. Most people write "close-up" and stop there, when what actually steers a model is the richer phrasing: "Close-up shot of the character's face, highlighting expressions and facial details, neutral lighting." That's the kind of prompt chunk that makes a real difference in a generated image - and it's the kind of thing you'd never type out by hand every time. The node bundles it so you pick from a menu instead.

It's clearly built for the same prompt-assembly flow as the pack's Character nodes: string out a camera shot, a character description, an expression, a style - feed them all into a text concat or the pack's Prompt Assembler - and your positive prompt is constructed from consistent, well-phrased building blocks instead of ad-libbed text. If you're generating a character sheet or a storyboard sequence, the "Orthographic 3-view" option is a gift - it expands to a full character-sheet description with side/front/back views and consistent proportions, which is a real prompt that people hand-craft.

How it works

There's no intelligence here - it's a lookup table. Each dropdown option maps to a prewritten sentence stored in the node. Select one, you get that sentence out as a string. The word choices are deliberate and tested-sounding ("emphasizing height and power" for worm's eye view, "compressing distance" for telephoto), so the phrasing reads like it was written for prompting rather than for a film textbook.

Select None and the node outputs whatever your prefix input is - usually an empty string. That's the graceful way to say "no camera instruction this time."

Inputs and outputs

• camera_type - the 14-way dropdown: None, Orthographic 3-view, Medium shot, Close-up, Long shot, Bird's eye view, Worm's eye view, Over-the-shoulder, First-person, Panoramic, Wide-angle, Telephoto, Portrait, Landscape.
• prefix (optional, wildcard) - text to prepend to the description. Stick a character name or a style tag in here and it lands in front of the shot description.

Output:

• camera_text (STRING) - the shot description, prefix included.

Installing it

Ships in ComfyUI Fictiverse Nodes. ComfyUI Manager → search "ComfyUI Fictiverse Nodes" → install → restart, or:

cd ComfyUI/custom_nodes
git clone https://github.com/Fictiverse/ComfyUI_Fictiverse

No dependencies, no model files. Apache 2.0 licensed.

Common issues

Standard pack gotcha: if you hit ModuleNotFoundError: No module named 'custom_nodes.ComfyUI_Fictiverse' at startup, the install folder isn't named exactly ComfyUI_Fictiverse. Rename it and restart.

The one thing to remember with this node (and its Character siblings): it only describes. It doesn't enforce. The model still has to interpret "worm's eye view" correctly, and on some backbones you may need to pair it with a matching negative prompt or just try a couple of options. Also, the descriptions assume a character subject ("the character's face", "focusing on the character") - for landscape or non-character work you'll want the Landscape option, which is the one phrased around the environment.