TT Apply LUT
Steal a colorist's LUT and grade your renders like a film
- image
- IMAGE
A LUT - look-up table - is how film colorists make a teal-and-orange blockbuster look the way it does. It's a .cube file that remaps every color to another color, wholesale. TT Apply LUT lets you drop those same files onto your generated images. It's the one node in this pack that takes an actual file you have to go fetch, and the payoff is instantly filmic color grades you didn't have to hand-tune.
From TenserTensor's postproduction section, it's simple: image in, IMAGE out, and three controls deciding how the grade is applied.
The three controls
lut_file- a dropdown of.cubefiles. This is where the "go fetch a file" part lives: the node scans a specific folder on your machine and lists what it finds. No.cubefiles, empty dropdown. (See below for where to put them.)strength(0–1, default 1.0) - blend between the graded image and the original. 1.0 is the full grade; 0.5 is a half-strength grade. This is the knob you'll actually ride, because most LUTs are designed for footage and land too hard on a crisp AI render - 0.3–0.6 is often the sweet spot.colorspace-linearorlogarithmic. This is the author's field and it encodes a real decision: the node applies a1/2.2gamma lift before the LUT and restores it after in logarithmic mode. If your LUT was made for log-encoded video, uselogarithmic; if it's a display-referred "look" LUT, uselinear. Wrong choice and the grade comes out washed out or crushed. The field is marked advanced, but it's the one that determines whether the grade looks right.
Under the hood it uses the colour (colour-science) library to parse the .cube and apply it - which is worth knowing because it means this node pulls a real color-science dependency into the pack.
Where your LUTs go
The node scans ComfyUI/custom_nodes/ComfyUI-TenserTensor/lut/. The pack ships that folder with a put_luts_files_here.txt placeholder. Drop .cube files there and refresh:
ComfyUI/custom_nodes/ComfyUI-TenserTensor/lut/
└── my-teal-lut.cube
Good free sources: the LUT packs bundled with DaVinci Resolve tutorials, or the open .cube collections on GitHub. Any .cube will show up in the dropdown after a refresh.
Install and caveats
cd ComfyUI/custom_nodes
git clone https://github.com/tenser-tensor/ComfyUI-TenserTensor
or ComfyUI Manager → "TenserTensor" → install → restart.
Two gotchas worth flagging. First, the pack's requirements.txt only lists gguf and kornia, but this node's code imports colour - it usually rides in with ComfyUI's own dependency set, but if the postproduction group throws an import error, pip install colour-science is the fix. Second, the standing V1 situation: this is a legacy node, moved to Deprecated/ after the pack's migration to ComfyUI's API V3, with removal planned for a future major release. It works today. And if the dropdown is empty, that's not a bug - it's telling you the LUT folder has no files in it yet.
Inputs (4)
| Name | Type | Default | Description |
|---|---|---|---|
| image | IMAGE | — | |
| lut_file | COMBO | 0 options: | |
| strength | FLOAT | 1.000–1 | — |
| colorspace | COMBO | 2 options: linear, logarithmic |
Outputs (1)
| Name | Type | Description |
|---|---|---|
| IMAGE | IMAGE | — |