Nodes/comfyUI-TiledWan/TiledWan Image Statistics
ComfyUI Node

TiledWan Image Statistics

A pocket console readout for your tensors

By Baverne·Created about a year ago·Updated 7 months ago· 8
TiledWan Image Statistics
  • image
  • image
  • min_value
  • max_value
  • mean_value
  • variance
  • median_value
show_per_channeltrue

Sometimes the node you need does nothing but talk. TiledWan Image Statistics takes an image, prints min, max, mean, variance, and median to your ComfyUI console, and passes the image straight through. That's the whole job, and it's a genuinely handy one when you're debugging a video pipeline.

It ships in Baverne's comfyUI-TiledWan pack, whose real business is tiled Wan 2.1 VACE video inpainting - but debugging utilities like this are why the pack is a set rather than a single node. Where it earns its keep: checking whether an image is properly normalized before it hits a model, catching a near-black or near-white frame that would silently wreck a batch, or verifying that a channel actually spans the range you think it does. The stats land in the console because the node is an output node by design - it's a diagnostic, not a data source.

Inputs: image (IMAGE) and show_per_channel (BOOLEAN, on by default), which adds separate R/G/B breakdowns to the console output. Outputs: the image pass-through plus five FLOAT values - min_value, max_value, mean_value, variance, median_value - in case you want to wire a threshold somewhere.

It's useful precisely because it's boring: no knobs to mis-set, no failure modes to hit. Install it as part of the pack - ComfyUI Manager (search "comfyUI-TiledWan"), or cd ComfyUI/custom_nodes && git clone https://github.com/Baverne/comfyUI-TiledWan, then restart - and it just works. The pack declares no extra dependencies.

If your problem is masks rather than images, the pack's TiledWan Mask Statistics is the sibling for that.

CategoryTiledWan

Inputs (2)

NameTypeDefaultDescription
imageIMAGEInput image to analyze. Statistics will be calculated across all pixels and channels.
show_per_channelBOOLEANtrueDisplay separate statistics for each color channel (R, G, B) in addition to overall statistics.

Outputs (6)

NameTypeDescription
imageIMAGEPass-through of the input image
min_valueFLOATMinimum pixel value found in the image
max_valueFLOATMaximum pixel value found in the image
mean_valueFLOATAverage pixel value across the entire image
varianceFLOATVariance of pixel values (measure of spread/contrast)
median_valueFLOATMedian pixel value (middle value when sorted)