Float to Int
Round a weight curve into discrete frame indices
- floats
- ints
Some things want continuous weights; some things want integers. Float to Int is the pack's one-trick converter for the latter: take any list of floats, round every value to the nearest whole number, and hand back an INTS list of the same length. 0.3 → 0, 0.5 → 1, 2.7 → 3. That's the whole node - but "round to nearest int" is a surprisingly useful step in a video pipeline.
Where it fits
In the Yvann audio pipeline, floats are the language of weights, but a lot of downstream consumers speak integers:
- Frame indices - turn a continuous curve into a set of discrete index values you can use to select specific frames or drive logic nodes.
- Counts - convert a per-section amplitude into a per-section integer count (frames to skip, repeats to apply, batch sizes).
- Bridging to integer-only nodes - plenty of ComfyUI utility and logic nodes want
INT, and this is the clean cast point instead of fighting widget types mid-graph.
The rounding follows standard rules (0.5 rounds up), so 0.5 → 1, not 0. If you need floor or ceiling semantics, that's a different tool - this node is strictly round-to-nearest.
Gotchas
- It's a list converter, not a scalar converter. It takes the pack's
FLOATSlist type and outputsINTS(a list). If you want a single integer from a single float, this isn't the node. - Rounding destroys dynamics. If you round a normalized 0–1 weight curve, you get mostly 0s and 1s - the shape vanishes. Only use this when integer semantics are actually what you want, not as a casual cast.
- No input validation beyond the type - pass a list, get a list, done.
Install
Nothing special:
cd ComfyUI/custom_nodes
git clone https://github.com/yvann-ba/ComfyUI_Yvann-Nodes
or ComfyUI Manager → search ComfyUI_Yvann-Nodes. It's a numpy one-liner under the hood, so it's instant and dependency-light - one of those nodes you forget you have until you need it, and then you're glad it's there.
Inputs (1)
| Name | Type | Default | Description |
|---|---|---|---|
| floats | FLOATS | List of floats to round to nearest integers |
Outputs (1)
| Name | Type | Description |
|---|---|---|
| ints | INTS | — |