BD TRELLIS2 Dual Conditioning (Info)

Some nodes are machinery; some are sticky notes. BD TRELLIS2 Dual Conditioning is the sticky note - its whole job is to explain the dual-conditioning workflow for TRELLIS.2 and hand you the text. It doesn't process an image, it doesn't run a model, it doesn't even take a real input: connect it (or not), and it emits workflow_info, a plain-text STRING containing the setup instructions. If you've ever wished a pack would just document its own advanced workflow inside the graph, this is that wish granted.

What it is

A documentation node with a display name that ends in "(Info)" - which is the honest tell. The info_trigger input (default: "Connect to see workflow info") is a dummy; there's no processing. The output, workflow_info, is a STRING you can wire to a text display node or read in the node's own output area.

What the info actually says

The embedded text walks through the three useful ways to split TRELLIS.2 conditioning across the shape and texture stages:

1. Dual conditioning, different images - traced outline for shape, clean image for texture. Setup: two BD TRELLIS.2 Get Conditioning nodes; first one's conditioning → Image to Shape; second's → Shape to Textured Mesh as texture_conditioning.
2. Fast shape + detailed texture, different resolutions - a 512 conditioning for shape generation, a 1024_cascade one for the texture voxelgrid. Roughly 2× faster shape with the high-res color data preserved for vertex-color sampling.
3. Combined - both tricks at once: different images and different resolutions, using all four optional inputs together.

That last one is the genuinely useful recipe. The motivating use case is the character/pipeline one: generate the shape fast and loose from a low-res pass, then paint the high-detail texture from a separate, cleaner reference. It's the same philosophy as the "fast shape + detailed texture" recommendation that's baked into the Get Conditioning node's own tooltip.

Why it exists

You could read the README, sure - but the README is 20 screens away and this lives in the graph, one node away from the workflow you're building. The BrainDead pack ships 18 example workflows and a bunch of these affordances; this node is the inline documentation for the single most common TRELLIS2 quality optimization. It's also a reasonable template if you're building your own info/helper nodes - a string output wired to a text display is the cheapest "docs in the graph" pattern there is.

Gotchas

None mechanically, but a couple of expectations to set: it's an output node, so it shows up in the workflow as a sink; the workflow_info is a STRING you need a text-display node to actually read in the UI. And it's documentation, not a processor - it won't wire the dual conditioning for you, and it won't error if you ignore it. If you were hoping for a node that performs dual conditioning, the actual work is done by two Get Conditioning nodes, which this one cheerfully explains.

Installing it

ComfyUI Manager: search "BrainDead" → install. Manual:

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

Restart, find it under 🧠BrainDead/TRELLIS2. Add it to any TRELLIS2 workflow you're tuning, wire workflow_info to a text display, and follow the recipe it prints. No dependencies beyond the pack itself - though the workflows it documents still need ComfyUI-TRELLIS2 installed to actually run.