πΌ Load Image From Path
Load Any Image by Full Path, Not Just From ComfyUI's Input Folder
- image
ComfyUI's built-in Load Image only shows you the input folder. Fine most of the time, infuriating the moment your frames live somewhere else - a render farm dump, a shared drive, an output directory from another tool. Load Image From Path [DVB] does the obvious thing the core node won't: it takes an absolute filesystem path as a string and loads that file. One image, one path, no browsing a fixed list.
That makes it a glue node more than a loader. Pair it with the pack's file-iteration tools (For Each Filename [DVB]) and you can loop over a directory of externally-produced frames without ever copying them into ComfyUI's input folder. Or use it to re-ingest an image your own workflow just wrote to disk. If you're mixing DVB's frame-set pipeline with frames that came from outside ComfyUI, this is the front door.
How it works
Under the hood it reads the file at image_path, loads it with its alpha channel intact, and hands it back as a normal IMAGE tensor. One detail is worth knowing: the node hashes the file's contents to decide whether it needs to re-run, so if you overwrite the file on disk, the workflow picks up the change on the next run instead of silently reusing a cached result. That's a genuinely useful behavior for iterative pipelines - edit the file externally, hit Run, get the new content.
The inputs and outputs that matter
image_path(default"") - the full path to the image file. Not relative to the input folder, not a filename - the whole path. A non-existent path won't error on load but won't trigger a re-run either, so it's worth double-checking you've got the exact file name.- Output:
image- a singleIMAGEtensor, ready for any image node in the graph.
That's the entire interface. Single file in, single image out.
Installing
It's one node in the Dream Video Batches pack (Alt Key Project / Dream Project):
cd ComfyUI/custom_nodes
git clone https://github.com/alt-key-project/comfyui-dream-video-batches.git
cd comfyui-dream-video-batches
pip install -r requirements.txt
Or search "Dream Video Batches" in ComfyUI Manager and restart. No model files involved.
Where people get burned
Path formatting is the whole game. Windows users: backslashes are fine in a string input, but if you're feeding the path through another node that manipulates strings, forward slashes are the safer bet. And don't expect this to load a folder - it's one file, and the file-iteration node is the separate tool that walks directories for you. If your image loads as the wrong colors or disappears, check whether the file actually has an alpha channel; this loader keeps alpha, and some downstream nodes assume RGB.
Inputs (1)
| Name | Type | Default | Description |
|---|---|---|---|
| image_path | STRING | β |
Outputs (1)
| Name | Type | Description |
|---|---|---|
| image | IMAGE | β |