ComfyUI Node
🦴 Poser Render (mesh → ControlNet image)
A ComfyUI node in TechNoir/Poser with 11 inputs and 1 output.
🦴 Poser Render (mesh → ControlNet image)
- mesh
- image
â—„width768â–º
â—„height1024â–º
â—„passesdepthâ–º
â—„backgroundblackâ–º
â—„framing1.12â–º
â—„azimuth0.00â–º
â—„elevation5.00â–º
â—„fov30.00â–º
â—„target_y0.00â–º
â—„mesh_colorâ–º
CategoryTechNoir/Poser
Inputs (11)
| Name | Type | Default | Description |
|---|---|---|---|
| mesh | MESH | — | |
| width | INT | 76832–4096 | — |
| height | INT | 102432–4096 | — |
| passes | COMBO | depth | 3 options: depth, normal, rgb |
| background | COMBO | black | 2 options: black, white |
| framing | FLOAT | 1.120.1–10 | Camera framing tightness. 1.5 = loose (65% frame fill). 1.12 = tight portrait (88%). anny_depth ControlNet conditioning uses 1.12. |
| azimuth | FLOAT | 0.00-360–360 | — |
| elevation | FLOAT | 5.00-90–90 | — |
| fov | FLOAT | 30.001–170 | — |
| target_y | FLOAT | 0.00-1–1 | Vertical offset of the camera look-at point in NORMALIZED mesh space. SIGN CONVENTION (empirically verified 2026-07-28): HIGHER target_y = subject HIGHER in frame. target_y=+0.3 → head at row 7/1024 (nearly clipped at top). target_y=0 → centered, full body visible (head row 64, foot row 953). target_y=-0.05 → head row 214, feet CLIP at row 1023. Two prior commits had this sign backwards. 0.0 is the safe default. |
| mesh_color | STRING | Optional RGB albedo override for the rgb pass, as 'r,g,b' floats in [0,1]. Empty = warm gray (0.78,0.75,0.72). Example: '0.6,0.74,0.97' for Pose Studio blue (#98bdf7). Only affects the 'rgb' pass. |
Outputs (1)
| Name | Type | Description |
|---|---|---|
| image | IMAGE | — |