Nodes/ComfyUI-SD-CN-Animation/SD-CN Animation Txt2Vid
ComfyUI Node

SD-CN Animation Txt2Vid

Generates a video frame sequence using FloweR optical flow prediction with two-pass SD sampling for temporal coherence.

By pxl-pshr·Created 4 months ago·Updated 4 months ago· 10
SD-CN Animation Txt2Vid
  • model
  • vae
  • positive
  • negative
  • flower_model
  • init_image
  • control_net
  • motion_ctrl
  • frames
seed0
steps15
cfg5.5
sampler_name
scheduler
width512
height512
num_frames30
processing_strength0.85
fix_frame_strength0.15
loop_frames0
cn_strength1.00
CategorySD-CN-Animation

Inputs (20)

NameTypeDefaultDescription
modelMODEL
vaeVAE
positiveCONDITIONING
negativeCONDITIONING
flower_modelFLOWER_MODEL
seedINT00–18446744073709550000
stepsINT151–200
cfgFLOAT5.50–100
sampler_nameCOMBO44 options: euler, euler_cfg_pp, euler_ancestral, euler_ancestral_cfg_pp, heun, heunpp2, +38
schedulerCOMBO9 options: simple, sgm_uniform, karras, exponential, ddim_uniform, beta, +3
widthINT512128–2048
heightINT512128–2048
num_framesINT302–9999
processing_strengthFLOAT0.850–1Denoising strength for inpaint pass (higher = more creative, less coherent)
fix_frame_strengthFLOAT0.150–1Denoising strength for refinement pass (keep low for stability)
loop_framesINT00–9999Number of frames at end to blend back toward first frame for seamless loop. 0 = disabled.
init_imageoptIMAGE
control_netoptCONTROL_NET
cn_strengthoptFLOAT1.000–10
motion_ctrloptMOTION_CTRL

Outputs (1)

NameTypeDescription
framesIMAGE