Gibbstone analyzes a reference image, understands the editing decisions behind it, and builds a Lightroom-ready starting point — presets, profiles, and 3D LUTs. Understand the style, then create your own interpretation.
● Runs entirely in your browser — your photos never leave your device
Gibbstone doesn't clone pixels. It infers and rebuilds photographic editing decisions — tone character, contrast, color cast, saturation hierarchy, highlight rolloff — so the result feels like a starting point, not a destructive edit.
Drop in any photo whose look you admire — a JPEG, an embedded RAW preview, or a test frame. Gibbstone reads its tone and color DNA.
JPEG or RAW, straight from your camera. A worker-based LibRaw WASM decoder handles RAW files right in the browser, with correct orientation and brightness.
One click builds a Lightroom Match Package: preset, profile, 3D LUT, Look DNA, and a README. Or just download the finished JPEG. Your call.
Every match is built from measurable analysis with bounded, explainable corrections — designed to keep people looking like people.
Soft lanes for skin, foliage, sky, and neutrals stop dominant regions — a green lawn, a big sky — from steering the entire look.
Confidence-based skin lanes keep faces and hands believably warm — no yellow-olive drift — while a chroma gate stops red dresses from stealing skin protection.
NEF, CR2/CR3, RAF, ARW, ProRAW DNG and more decode in-browser. High-quality embedded previews first, fast LibRaw decode as fallback — never a tiny thumbnail.
Feed Gibbstone reference frames — grey cards or ColorCheckers work best — and it learns your camera, exporting camera-specific profiles and presets.
No backend exists. Analysis, matching, and export all run on your device — nothing is uploaded, ever.
Installable as an app on desktop and mobile. The service worker caches everything, so you can build looks on a shoot with no signal.
The Lightroom Match Package is a ZIP built for a real workflow. Install the profile, apply the preset, keep shooting. The tone curve, color grading, and camera compensation are all there — as adjustable Lightroom settings, not baked pixels.
Prefer a LUT for video or another editor? It's in the same package.
Try a match →Camera-specific exports become available once Gibbstone learns a camera from your reference images.
This isn't a policy — it's the architecture. Gibbstone is a static app with no backend. There is no server to upload to, no account to create, and no analysis performed anywhere but on your own hardware.
Client work, family photos, unreleased shoots — nothing you load into Gibbstone is ever transmitted.
No. Gibbstone is deterministic color science, not generative AI or segmentation models. It measures the tonal and color character of a reference — contrast, cast, saturation hierarchy, highlight rolloff — and rebuilds those as editable Lightroom decisions. It's a way to learn and interpret a style, not clone an image.
A Lightroom Match Package: a camera profile (.xmp), a develop preset (.xmp) with tone curve, a 3D LUT, a Look DNA JSON, and a README. You can also simply download the matched JPEG.
JPEGs and RAW files including Nikon NEF, Canon CR2/CR3, Fujifilm RAF, Sony ARW, and iPhone ProRAW DNG. RAW files are decoded in your browser with LibRaw compiled to WebAssembly.
Yes — the preset describes the look itself as Lightroom settings. Since it wasn't tuned against each new photo, expect to trim exposure per shot, exactly as with any preset.
Never. Gibbstone has no backend. Everything — decoding, analysis, matching, export — happens on your device, and the app even works offline once loaded.
Give Gibbstone reference images from the camera you want it to learn. Grey card or ColorChecker frames captured with that camera produce the most accurate camera-specific profiles.
Free, in your browser, on your device. No account, no upload, no catch.
Open Gibbstone