🅐 Image Make Slideshow

ComfyUI can generate a hundred images, but making a slideshow video out of them is usually several steps of glue. 🅐 Image Make Slideshow is that glue: you feed in a batch of images, it holds each one on screen for a while, crossfades or slides between them, and hands back a big batch of frames in the right order. Feed that to a video encoder and you've got a real, watchable slideshow without leaving the graph. One honest note before the details: the class name is spelled ACE_ImageMakeSlieshow - the "Slideshow" typo is in the pack itself, so don't go hunting for a correctly-spelled version.

How it works

The required images input is your IMAGE batch. Then you dial in the timing:

• fps - INT, default 25. The frame rate the output is intended for.
• image_display_time - FLOAT, default 5.0 seconds. How long each image stays static.
• transition_time - FLOAT, default 0.5 seconds. Duration of the transition between images.
• transition_effect - enum: fade, slide_left, slide_right, slide_up, slide_down. Default fade.

For each image after the first, the node generates fps × transition_time transition frames blending it with the previous one (crossfade or a directional slide), then appends fps × image_display_time copies of the image itself. Stack it all up and you get a single batched IMAGE output: first image held, transition, second image held, transition, and so on. Because it's just a batch of tensors, you're free to run it through any further image processing before encoding - that's a feature, since you can grade or upscale the frames in-graph.

Inputs and outputs

• images - IMAGE, required.
• fps, image_display_time, transition_time, transition_effect - as above.
• Output: IMAGE - the assembled frame sequence.

Where it fits

The output wants to end up in a video-encoding node (VHS or similar) that turns the frame batch into an actual mp4. Keep the fps setting here consistent with what the encoder uses, or your timing math will be off. It's also worth noting the output size: 25 fps × 5 seconds × N images means a big tensor in memory - fine for a handful of frames, heavy for a long show.

Install

Ships in hay86/ComfyUI_AceNodes:

cd ComfyUI/custom_nodes
git clone https://github.com/hay86/ComfyUI_AceNodes
cd ComfyUI_AceNodes
pip install -r requirements.txt

Restart ComfyUI and look for 🅐 Image Make Slideshow under Ace Nodes. Manager users search ComfyUI_AceNodes.

Gotchas

The hard requirement hidden in the code: all input images must be the same size. Different dimensions throw an assertion error mid-run, so normalize with 🅐 Image Constrain first if your batch is mixed. Keep the x out of your timing values (floats work, but the display and transition times have a 0.1 minimum). And the pack's heavy requirements.txt (rembg, transformers, boto3…) installs even though this node is pure tensor work. For turning a batch into a watchable sequence, though, it's one of the pack's more rewarding nodes.