EasyUse/PreSamplinggenerated

PreSampling(samplerSettings)

The PreSampling node (class samplerSettings) from ComfyUI-Easy-Use is a pipeline-level sampler configuration node.

sampler Settings

pipe
image_to_latent
latent
pipe
steps
20
cfg
8
sampler_name
COMBO
scheduler
COMBO
denoise
1.00
seed
INT
Easy Use

The PreSampling node (class samplerSettings) from ComfyUI-Easy-Use is a pipeline-level sampler configuration node. It takes a PIPE_LINE, applies your sampling choices, and returns the updated PIPE_LINE for downstream sampling nodes. It belongs to the EasyUse/PreSampling category.

Inputs

InputTypeRequiredDefaultDescription
pipePIPE_LINEYes—The pipeline object carrying model, conditioning, VAE, latent, and other workflow state.
stepsINTYes20Number of sampling steps. Range: 1–10000.
cfgFLOATYes8Classifier-free guidance scale. Range: 0–100.
sampler_nameCOMBOYes—Sampler algorithm to use. The dropdown is populated by the samplers available in your ComfyUI environment.
schedulerCOMBOYes—Scheduler to pair with the selected sampler. The dropdown is populated by the schedulers available in your ComfyUI environment.
denoiseFLOATYes1Denoising strength, from 0 to 1 in 0.01 steps. Use 1 for full sampling from the latent; lower values preserve more of the starting image or latent structure.
seedINTYes—Seed used to control sampling noise. Changing it produces different results.
image_to_latentIMAGENo—Optional input image. When connected, the image is converted into a latent and used as the starting point, which is useful for img2img-style workflows.
latentLATENTNo—Optional existing latent to use as the starting point. Connect this when you already have an encoded latent instead of an image.

Outputs

OutputTypeDescription
PIPE_LINEPIPE_LINEThe updated pipeline with the sampling settings applied, ready to pass to a node that performs sampling.

Usage Notes

  • Place this node before the sampler or generation node in your graph. It only configures settings; it does not run the sampling itself.
  • For text-to-image workflows, leave the optional image/latent inputs disconnected and make sure the pipe already contains a latent, for example from an Empty Latent node.
  • For image-to-image or refinement workflows, connect image_to_latent to start from an image without manually adding a VAE Encode step, or connect latent if you already have an encoded latent.
  • The denoise value is especially important when starting from an image or existing latent. A value of 1 performs full denoising; lower values retain more of the original structure, which is useful for light edits or upscaling workflows.

Comments

Sign in with GitHub to join the discussion.

Loading comments…