Set Value For MC I2V V2

This node doesn't generate anything. It's a settings panel - five fields in, the same five values straight back out as separate wires. If you've ever built a big workflow and gotten tired of five separate primitive nodes cluttering the top of your canvas, that's exactly the itch this scratches: one box to fill in, five clean outputs to route wherever they're needed.

The catch is who it was built for. This is huyl3-cpu's own front-end for an image-to-video ("I2V") pipeline - "MC" isn't explained anywhere in the pack, and I couldn't turn up what it stands for. The defaults give the game away regardless: paths point at /content/drive/MyDrive, which only makes sense if you're running ComfyUI inside Google Colab with a Drive mount, and the resolution choices (480p through 1080p) read like short-form video sizing, not a general-purpose render setting. The actual I2V model call this feeds into isn't part of this pack - it lives in whatever template the author distributes separately. Wire this into your own generation nodes and you'll get exactly as much value as the pass-through fields provide.

The inputs and outputs that matter

Five fields, five matching outputs:

• URL image / URL audio - plain string paths (or URLs) to your source image and an audio track. Come out the other side as path_image and path_audio, unchanged.
• Độ phân giải ("resolution") - enum, 480p / 600p / 720p / 1080p, default 720p. Comes out as resolution, typed INT. Nothing in the schema documents what integer each label maps to (unlike its Set-Value-For-MC-V2V-V2 sibling, which does spell this out) - if the exact pixel height matters downstream, check it once with a Preview/Text node before trusting it.
• THƯ MỤC lưu ("save folder") - where output presumably gets written later in the graph. Passes through as save_dir.
• Prompt mô tả cử động (tiếng việt/tiếng anh) ("motion description prompt, Vietnamese or English") - a multiline text field for describing the desired movement. Passes through as prompt.

That's the whole node. No validation, no transformation - it just gives your graph a labeled front panel instead of five loose string/enum widgets.

Installing it

Via ComfyUI Manager: search the pack title, comfyui-huyl2-nodes, install, restart. Or by hand:

cd ComfyUI/custom_nodes
git clone https://github.com/huyl3-cpu/comfyui-sortlist

then restart ComfyUI. The pack's own README only documents one node - a video-segment-cutting tool that isn't even in this batch - and is written entirely for a Google Colab setup, not local install. There's no requirements.txt guidance and no mention of dependencies for this specific node, which tracks: it's a pure settings hub with no processing logic of its own, so it shouldn't need anything beyond a working ComfyUI install.

Common issues & troubleshooting

Because it's a pass-through, "this node is broken" almost always turns out to be a downstream problem instead - the model/API node actually consuming path_image, path_audio, or prompt is where to look first if generation fails.

The one thing worth double-checking yourself: the resolution output is an INT with no documented mapping in this version. Drop a quick text/preview node on that wire once, confirm it resolves to the pixel value you expect for each of the four choices, and you won't have to think about it again.

Given the Google-Drive-shaped defaults and the fact that the README never mentions this node by name, treat it as exactly what it looks like: infrastructure for one specific person's Colab template, bundled into a public pack. If you're not following that template, there's nothing wrong with reusing it purely as a labeled config panel - just don't expect the defaults to mean anything for your setup.