A real lookup table, not a named approximation
A 3D LUT maps each red, green and blue coordinate to a new colour. Most photographs sit between the recorded grid points, so this tool interpolates between the eight surrounding samples instead of snapping a pixel to its nearest entry. The result is the transform written in your file, blended with the original only when you lower strength.
The parser accepts the common LUT_3D_SIZE, DOMAIN_MIN, DOMAIN_MAX and TITLE fields. It verifies that the cube contains exactly the promised number of samples before touching the photograph.
What an imported LUT cannot decide
- A LUT expects a particular input colour space. A log-camera LUT applied to an ordinary sRGB JPG may look extremely contrasty because the input is not what its author designed for.
- This browser workflow reads the displayed RGB values. It does not interpret ICC profiles or convert camera RAW data.
- A saved JPG carries the camera, lens, focal length, exposure and date across from the original, and the location tags are removed on the way. PNG and WebP carry nothing at all.
Compare an imported transform with the built-in rail
- The look railTwenty-four looks in three packs, every tile drawn from the photograph you opened.
- Film looksThe eight emulations on their own, with what each one does to skin written down.
- Batch filtersRun one named look across up to forty pictures and collect the full-size results in a ZIP.
- A look versus an effectOne test that sorts any filter you meet into the two kinds, and says which site handles which.


