ComfyUI Node
Anima Triple-Pass KSampler
A ComfyUI node in HaDeS/Anima Samplers with 30 inputs and 5 outputs.
Anima Triple-Pass KSampler
- model
- positive
- negative
- latent_image
- model_pass2
- model_pass3
- vae
- final_latent
- pass1_latent
- pass2_latent
- decoded_image
- info
◄triple_presetManual (Use Widget Settings)►
◄seed0►
◄upscale_methodLatent 16ch Variance-Preserving Bicubic►
◄steps_stage128►
◄cfg_stage14.5►
◄sampler_stage1er_sde►
◄scheduler_stage1normal►
◄denoise_stage11.00►
◄shift_stage10.00►
◄scale_stage21.35►
◄steps_stage216►
◄cfg_stage24.0►
◄sampler_stage2er_sde►
◄scheduler_stage2beta►
◄denoise_stage20.48►
◄shift_stage23.50►
◄scale_stage31.75►
◄steps_stage312►
◄cfg_stage33.4►
◄sampler_stage3euler►
◄scheduler_stage3anima_beta57►
◄denoise_stage30.26►
◄shift_stage34.00►
CategoryHaDeS/Anima Samplers
Inputs (30)
| Name | Type | Default | Description |
|---|---|---|---|
| model | MODEL | — | |
| positive | CONDITIONING | — | |
| negative | CONDITIONING | — | |
| latent_image | LATENT | — | |
| triple_preset | COMBO | Manual (Use Widget Settings) | 5 options: Manual (Use Widget Settings), Progressive Staircase 1.0x -> 1.35x -> 1.75x (Ultra-Res Poster Pipeline), Progressive 2x Master 1.0x -> 1.5x -> 2.0x (4 MP Studio Illustration), Concept -> Aesthetic Bridge 1.5x -> Micro-Polish 1.5x (Lock Res at Stage 2), Hybrid Base -> Aesthetic -> Turbo Polish (Fast 3-Stage Studio) |
| seed | INT | 00–18446744073709550000 | — |
| upscale_method | COMBO | Latent 16ch Variance-Preserving Bicubic | 4 options: Latent 16ch Variance-Preserving Bicubic, Latent 16ch Sharp Bislerp, Pixel-Space VAE Round-Trip (Lanczos 32px), Direct Latent Pass-Through (1.0x) |
| steps_stage1 | INT | 281–200 | — |
| cfg_stage1 | FLOAT | 4.51–20 | — |
| sampler_stage1 | COMBO | er_sde | 44 options: er_sde, euler, euler_ancestral, dpmpp_2m_sde_gpu, dpmpp_2m, euler_cfg_pp, +38 |
| scheduler_stage1 | COMBO | normal | 11 options: normal, beta, anima_beta57, anima_lightning_tdm, simple, sgm_uniform, +5 |
| denoise_stage1 | FLOAT | 1.000–1 | — |
| shift_stage1 | FLOAT | 0.000–10 | Rectified Flow shift for this pass. 0 = inherit the shift already set on the model (Model Loader flow_shift or ModelSampling nodes). |
| scale_stage2 | FLOAT | 1.351–2.5 | — |
| steps_stage2 | INT | 161–200 | — |
| cfg_stage2 | FLOAT | 4.01–20 | — |
| sampler_stage2 | COMBO | er_sde | 44 options: er_sde, euler, euler_ancestral, dpmpp_2m_sde_gpu, dpmpp_2m, euler_cfg_pp, +38 |
| scheduler_stage2 | COMBO | beta | 11 options: normal, beta, anima_beta57, anima_lightning_tdm, simple, sgm_uniform, +5 |
| denoise_stage2 | FLOAT | 0.480.05–1 | — |
| shift_stage2 | FLOAT | 3.500–10 | Rectified Flow shift for this pass. 0 = inherit the shift already set on the model (Model Loader flow_shift or ModelSampling nodes). |
| scale_stage3 | FLOAT | 1.751–3 | — |
| steps_stage3 | INT | 121–200 | — |
| cfg_stage3 | FLOAT | 3.41–20 | — |
| sampler_stage3 | COMBO | euler | 44 options: er_sde, euler, euler_ancestral, dpmpp_2m_sde_gpu, dpmpp_2m, euler_cfg_pp, +38 |
| scheduler_stage3 | COMBO | anima_beta57 | 11 options: normal, beta, anima_beta57, anima_lightning_tdm, simple, sgm_uniform, +5 |
| denoise_stage3 | FLOAT | 0.260.05–1 | — |
| shift_stage3 | FLOAT | 4.000–10 | Rectified Flow shift for this pass. 0 = inherit the shift already set on the model (Model Loader flow_shift or ModelSampling nodes). |
| model_pass2opt | MODEL | — | |
| model_pass3opt | MODEL | — | |
| vaeopt | VAE | — |
Outputs (5)
| Name | Type | Description |
|---|---|---|
| final_latent | LATENT | — |
| pass1_latent | LATENT | — |
| pass2_latent | LATENT | — |
| decoded_image | IMAGE | — |
| info | STRING | — |