Distance Modifier (Single)

Distance Modifier is the spatial-proximity dimension: macro, extreme close-up, portrait distance, medium, full body, wide, aerial, and vertical positions like low ground-level or high overhead. Twenty descriptions appended to your base prompt by ID. It's easy to overlook because it overlaps with the camera dimension, but it's actually the pack's most underrated node - distance vocabulary is concrete, physical, and extraordinarily well-understood by image models, so it moves output more reliably than the moody words do.

The modifiers read as a camera-to-subject ruler: macro at the low end, portrait distance, then the shot-scale progression - medium, full, wide - up through aerial and the vertical positions. It overlaps camera framing (close-up, full shot also live in the camera dimension), but distance is specifically about proximity and scale rather than lens character, and the distinction matters when you want "aerial view" without also implying a specific lens. The format is the pack's usual compact phrase, built to append cleanly after a comma.

The inputs: text (base prompt), modifier_id (1–20, default 6), and preserve_base (default true). Output: single STRING, modified_prompt. Same skeleton as all twelve single-modifier nodes.

Install is the pack-standard zero-dependency story: ComfyUI Manager → search "Variationator" → install → restart, or cd ComfyUI/custom_nodes && git clone https://github.com/MushroomFleet/ComfyUI-Variationator.git. No model downloads, no requirements.txt to install (the repo's install-portable.bat references one that doesn't ship). It sits under Prompt Orchestration, the newest of the single-modifier set.

Where this node earns its keep: batch scale studies. One subject through macro → portrait → full → aerial in a single pass (chain it into a Batch-style workflow, or just swap IDs) is a fast way to see how much scene context your subject needs. And for the pack's automotive roots, distance + camera + lighting is how you build the hero-shot vocabulary ladder. The one real caution is overlap: because distance and camera both describe framing, you can easily end up with "extreme close-up" from one node and "full shot" from another in the same prompt - that's a contradiction any encoder resolves poorly. Decide which dimension owns the framing and let the other one stay quiet. Distance is the cleaner pick for scale; camera is the cleaner pick for angle and lens.