☼ Image Brightness Adjustment

Every animation needs a fade at some point - a fade-in from black at the start, a fade-out at the end, a slow dip to black mid-shot for a scene change. Image Brightness Adjustment [Dream] is the node that does it: one factor input, one image out, where 1.0 is untouched and 0.0 is pure black. Wire the factor to a curve and the fade happens automatically over the animation.

It's a color node from the Dream Project Animation Nodes pack (alt-key-project), the 2023 Deforum-style animation toolkit. Like its siblings it's not actively maintained, but a single Pillow brightness call is timeless.

How it works

Under the hood it's ImageEnhance.Brightness(...).enhance(factor) - a Pillow operation with no dependencies beyond what the pack already installs. The factor is a FLOAT with a floor of 0 and no upper clamp:

• 1.0 - no change
• > 1.0 - brighter (2.0 = roughly double brightness; blown-out whites, careful)
• < 1.0 - darker
• 0.0 - black

There's no negative territory - a factor of 0 is as low as it goes, which is exactly right for fades.

Because factor is a socket (not just a slider), you can drive it from any of the pack's curve nodes. The classic setup: a Linear Curve going 0 → 1 over the first few frames for a fade-in, or 1 → 0 near the end for a fade-out. Same mechanism, zero keyframing by hand.

The uses that justify the install

1. Fades, obviously. In and out of shots, between scenes, down to black at the end. In a frame-by-frame render the alternative is re-rendering or hand-editing every frame - this does it as a node in the chain.
2. Exposure drift correction. Long Deforum-style animations slowly darken as colors accumulate; a mild brightness boost tied to a palette comparison keeps things level.
3. Compositing feel. A subtle brightness dip during a prompt-morph (the pack's prompt-morphing example does exactly this with brightness and contrast) sells a transition far better than an instant cut.

You can stack it with the pack's Image Contrast Adjustment - brightness first, contrast after - and both accept curve inputs, so the whole grade can be animated.

Installing

cd ComfyUI/custom_nodes
git clone https://github.com/alt-key-project/comfyui-dream-project.git

Or via ComfyUI Manager under "Dream Project Animation". No model downloads; the pack's deps are imageio, scipy, torchvision, pilgram, and evalidate. One thing to remember: it pins numpy<2.0, so if NumPy errors show up after installing, that's the first suspect. Find it under Dream → image → color.