VECTOR4 (JOV)

VECTOR4 (JOV) takes four numbers and puts them on one VEC4 wire. That sounds dull until you remember what four-number bundles actually mean in image work: RGBA colors, and the "top/left/top/right"-style corner quadruples that Jovimetrix's own projection and crop nodes love (see the TRANSFORM node's TL-TR and BL-BR inputs). This is the node that lets you drive those from individual animated values instead of typing constants into a widget.

What it is

Four numeric inputs - X, Y, Z, W (labeled 1st through 4th channel value) - each accepting booleans, floats or ints over the full signed 64-bit float range. Leave one unconnected and it contributes 0. Build the pair any way you like: static sliders, WAVE GEN outputs, a VALUE node, or other vector math. Jovimetrix advertises compatibility with ComfyMath Vector nodes, so the VEC4 you build here can flow into ComfyMath's vector operations too.

Outputs

Two, same pattern as the rest of the family:

• VEC4 - the quadruple as floats.
• VEC4INT - the quadruple rounded to integers.

The integer output is genuinely useful for RGBA work: when you're feeding a color into a node that wants 0–255 channels, grab VEC4INT and skip the rounding step yourself.

Install

Ships with Jovimetrix: git clone https://github.com/Amorano/Jovimetrix.git into ComfyUI/custom_nodes, then pip install -r requirements.txt (Windows portable: .\python_embed\python.exe -s -m pip install -r requirements.txt), or search "Jovimetrix" in ComfyUI Manager. No models, no keys.

Two small things to keep straight. First, remember the "W" input is the fourth channel, not a width - despite the naming pattern in some Jovimetrix nodes, this node treats all four as plain values. Second, vector type discipline matters in this pack: a VEC4 is its own type, so don't try to feed it where a VEC2 is expected and expect a silent truncation. Wire the right one and this node is as painless as it gets.