𝙆 Color Adjustment

This is two color tools in one node. The first half is what you'd expect - exposure, contrast, temperature, tint, saturation, all as plain sliders. The second half is more interesting: a dropdown of 25 named presets straight out of the pilgram Python library, which is a faithful reimplementation of classic Instagram filters (Aden, Brannan, Brooklyn, Clarendon, Earlybird, Gingham, Hudson, Inkwell, Kelvin, Lark, Lo-Fi, and more). It's a pure image-processing node - no model weights, no GPU dependency beyond whatever's already running your pipeline - so it's cheap to drop in as a final stylization pass.

How it works

The five manual sliders do standard tone adjustments on top of the image. Separately (or together, see below), the style dropdown applies one of the pilgram filter presets, and strength blends how much of that filter's effect shows through - 0 leaves the image untouched, 100 applies it at full force. There's also an All toggle in the optional inputs; per the author's own notes it's a one-click way to turn on every filter/adjustment combination at once rather than dialing exposure, contrast, temperature, tint and saturation in one at a time - useful if you're experimenting and want to see the maximal effect before backing off individual sliders.

Because the style filters and the manual sliders live on the same node, you can layer a preset like Clarendon on top of your own exposure and contrast tweaks rather than choosing one or the other.

The inputs and output that matter

• image (IMAGE) - required input.
• exposure / contrast / temperature / tint / saturation (FLOAT, -100 to 100, default 0) - the manual grade.
• style (choice of 26: None plus 25 named filters) - the pilgram preset.
• strength (FLOAT, 0–100, default 100) - how strongly the chosen style applies.
• All (optional BOOLEAN) - enables every filter combination at once.
• IMAGE (output) - the graded/filtered result, ready for a preview or save node.

Installing it

It ships as part of the full KayTool pack - there's no way to install just this node. ComfyUI Manager: search KayTool, install, restart. Manually: cd ComfyUI/custom_nodes && git clone https://github.com/kk8bit/KayTool, restart. Unlike some of KayTool's other nodes, this one has no model to download and no API dependency - once the pack's Python requirements are satisfied (pilgram is a lightweight pure-image library, not a heavy install), it works offline.

Common issues

If the pilgram dependency didn't install cleanly during setup - which can happen if ComfyUI Manager's pip step got interrupted or conflicted with another node's requirements, a known friction point across custom-node packs in general since everything shares one Python environment - the style dropdown may error out even though the manual sliders keep working fine; that's a sign to reinstall the pack's requirements rather than a sign the node itself is broken. Beyond that, this is a straightforward, low-risk node: since it does pure math on pixel values with no external service call, if the output looks wrong the first place to check is whether strength is actually above 0 and style isn't sitting on None while you're expecting a filter to show.