ComfyUI Node
AsymFLUX2 Oklab Encode
Encodes a ComfyUI IMAGE into the 3-channel Oklab pixel latent that the AsymFLUX2 transformer denoises in. This is a deterministic perceptual-color transform — there is no learned VAE.
AsymFLUX2 Oklab Encode
- image
- latent
◄mean_l0.560►
◄mean_a0.000►
◄mean_b0.010►
◄std0.160►
CategoryAsymFLUX2
Inputs (5)
| Name | Type | Default | Description |
|---|---|---|---|
| image | IMAGE | — | |
| mean_l | FLOAT | 0.560-1–1 | — |
| mean_a | FLOAT | 0.000-1–1 | — |
| mean_b | FLOAT | 0.010-1–1 | — |
| std | FLOAT | 0.1600.001–1 | — |
Outputs (1)
| Name | Type | Description |
|---|---|---|
| latent | LATENT | — |