SpektraLab Puts a Virtual Darkroom on Your Mac, and Agents Can Run the Enlarger

Photo post-processing now has an open, scriptable entry point that goes beyond single LUTs: a full film-emulation pipeline with an agent interface can be wired into automation workflows, batch operations, and personal AI tooling. The Mac-only, free, open-source nature makes it a practical playground for developers who want to experiment with MCP-driven creative tools.

SpektraLab models a full analog darkroom pipeline: a digital photo goes through film exposure and development, then printing, then scan output. The underlying film model comes from the open-source spektrafilm project, which computes spectral response, dye density, grain, and halation. SpektraLab implements that model in C++ and Metal, so the Mac app renders entirely on the GPU with no Python runtime required.

The interface makes the workflow visual-first: every film option shows a live preview of the current photo, and users pick a film and a paper separately (reversal film skips the paper step). The result is a way to taste-test combinations without learning curves or color wheels first. The app supports RAW and JPEG input, exports JPEG, PNG, and TIFF, and lets users save presets or copy settings across photos.

On the automation side, SpektraLab exposes a set of MCP tools that let external agents query the library, import photos, set star ratings and color labels, change film/paper/exposure settings, and batch-export using saved recipes. The exposed controls are limited to film, paper, and exposure; white balance, curves, and HSL remain outside the tool interface, and no preview images are provided, so agents cannot yet judge aesthetics themselves.