LTFloat_Step_005

LTFloat_Step_005 is the 0.05 member of the Latent Tools step-helper family - the point where a sweep stops being about smoothness and becomes about coverage. It's a calibrated constant: type a value, get the same number back as a FLOAT and a STRING, with the widget stepping by 0.05.

That's the whole deal with this family: sweeping a parameter across a batch of images with the seed locked, changing one variable per run. At 0.05 you cover a lot of ground fast - twenty runs moves a parameter a full unit. The outputs will be clearly distinct from each other, which is exactly right for finding the interesting zone in a noise parameter. Once you've bracketed it, swap in a finer step (0.005 or 0.001) to map it properly. Nobody uses the coarse steps to make animations; they use them to scout.

How it works

The honest mechanism, because the name invites the wrong assumption: the node does not auto-increment. It takes the value you give it, rounds to five decimals, and returns it as a FLOAT and a STRING. The step is the widget's spinner increment and the label for the constant. You bump the value between runs; the node keeps it deliberate.

The inputs and outputs

• value - FLOAT, default 0, range ±1,000,000, step 0.05.
• FLOAT - the value.
• STRING - the value as text, for filename prefixes so each output records its parameter.

Wire the FLOAT into a mean/std/cfg input, the STRING into a Save Image filename, keep the seed fixed, and read the range.

Installing it

Part of xl0's Latent Tools pack. ComfyUI Manager → search Latent Tools → install → restart, or:

cd ComfyUI/custom_nodes
git clone https://github.com/xl0/latent-tools

Restart, and it's under LatentTools. The pack's only dependency is lovely-tensors (Manager installs it; manual cloners may need pip install lovely-tensors). No model downloads.