Technical
AI Color Grading Still Isn't a Timeline Feature. It's a Round Trip
ยท RenderBob team
Imagen Video returns a new Premiere or Resolve project. LunaBloom exports a LUT. Colourlab can grade inside Resolve. None of that is a live grade on a generative timeline.
AI colour tools in 2026 sit beside Premiere Pro and DaVinci Resolve, and they do not all behave the same way. Imagen Video takes a finished .prproj or .drp, corrects the clips, and hands back a new project: Lumetri layers in Premiere, nodes in Resolve. LunaBloom's grading flow analyses footage and exports a .cube LUT for Resolve, Premiere, or Final Cut. Colourlab AI's Resolve OFX plugin is the exception that matters. Colourlab's own page says you grade in the node tree with no fetch, push, or round trip. Its Premiere and Final Cut path still fetches the timeline and pushes it back.
A LUT is a file. A grade on the clip is a control
A 3D LUT captures a colour transform you can apply broadly. It does not give you a live, per-clip control inside the timeline that produced the shot. Imagen's returned project is editable once you are back in Premiere or Resolve, which is better than a frozen LUT, and it is still a trip out of the session and back. Applying the look, refining it, and managing how it changes from shot to shot happens in the traditional editor.
Log footage needs a technical transform before the AI looks at it
Imagen's LUT guide is explicit about log and raw: a technical LUT converts that footage to a display space such as Rec.709, and you can assign a different technical LUT per camera before the creative look. Correction LUTs built for display-referred Rec.709 are the wrong tool on flat log, which is stored that way to keep dynamic range. A pipeline that detects the source colour space and applies the camera transform before analysis saves the mistake. Leaving it as a step the operator must remember is how the grade misreads the shot.
Theatrical finishing is still a colourist's job
Colourlab markets an OFX plugin for Resolve, Nuke, Baselight, and Flame, and it advertises a matching engine that runs in ACES. That is assistance inside a high-end suite. Dolby Vision trim passes and a theatrical digital intermediate are still led by a senior colourist in a dedicated pipeline. AI matching is useful for a consistent, presentable base. It does not replace that finishing room when the grade itself is the deliverable.
Generated clips need matching across models
Generated video adds a problem camera tools were not built for. Different models produce different colour and tone, so a timeline that mixes a local draft and a cloud final needs matching between generations, not only correction of each clip. Current graders expect captured footage. Current generators do not share a colour science. An editor that can match those sources on the timeline, without exporting a LUT first, is solving a generative problem the grading market has barely started on.
RenderBob keeps generation on the timeline. A grade that can follow the clip, including a match between a local draft and a cloud final, belongs in that same session.
More from the blog
- Human Authorship Is Becoming the Unit of Value
The Oscars' 2026 AI rules, Cannes' unpublished competition stance, a patchwork of festival policies, and the Thaler copyright denial all ask how much of a film a human authored.
- Ideation Is the Green Zone: Where Studios Allow AI and Where They Stop
Netflix's generative AI guidance for production partners treats concept work as the open case and final deliverables as the restricted one. That split is spreading as a working template.