Nodes/ComfyUI-WanVideoWrapper_QQ/Wan Video Image To Video Encode_v2 (QQ)
ComfyUI Node

Wan Video Image To Video Encode_v2 (QQ)

A ComfyUI node in WanVideoWrapper_QQ/utils with 22 inputs and 1 output.

By siraxe·Created 10 months ago·Updated 4 months ago· 71
Wan Video Image To Video Encode_v2 (QQ)
  • vae
  • clip_embeds
  • start_image
  • mid_image
  • end_image
  • control_embeds
  • temporal_mask
  • extra_latents
  • add_cond_latents
  • image_embeds
width832
height480
num_frames81
noise_aug_strength0.000
start_latent_strength1.000
mid_latent_strength1.000
end_latent_strength1.000
end_final_strength0.800
mid_position0.50
end_position1.00
force_offloadtrue
fun_or_fl2v_modeltrue
tiled_vaefalse
CategoryWanVideoWrapper_QQ/utils

Inputs (22)

NameTypeDefaultDescription
widthINT83264–8096Width of the image to encode
heightINT48064–8096Height of the image to encode
num_framesINT811–10000Number of frames to encode
noise_aug_strengthFLOAT0.0000–10Strength of noise augmentation, helpful for I2V where some noise can add motion and give sharper results
start_latent_strengthFLOAT1.0000–10Additional latent multiplier, helpful for I2V where lower values allow for more motion
mid_latent_strengthFLOAT1.0000–10Additional latent multiplier for mid frame, helpful for I2V where lower values allow for more motion
end_latent_strengthFLOAT1.0000–10Additional latent multiplier, helpful for I2V where lower values allow for more motion
end_final_strengthFLOAT0.8000–10Strength for end_image copy placed at final frame for temporal consistency
mid_positionFLOAT0.500–1Position of mid_image as fraction of total frames (0.0 = start, 1.0 = end)
end_positionFLOAT1.000–1Position of end_image relative to remaining timeline after mid_image
force_offloadBOOLEANtrue
vaeoptWANVAE
clip_embedsoptWANVIDIMAGE_CLIPEMBEDSClip vision encoded image
start_imageoptIMAGEImage to encode
mid_imageoptIMAGEmiddle frame
end_imageoptIMAGEend frame
control_embedsoptWANVIDIMAGE_EMBEDSControl signal for the Fun -model
fun_or_fl2v_modeloptBOOLEANtrueEnable when using official FLF2V or Fun model
temporal_maskoptMASKmask
extra_latentsoptLATENTExtra latents to add to the input front, used for Skyreels A2 reference images
tiled_vaeoptBOOLEANfalseUse tiled VAE encoding for reduced memory use
add_cond_latentsoptADD_COND_LATENTSAdditional cond latents WIP

Outputs (1)

NameTypeDescription
image_embedsWANVIDIMAGE_EMBEDS