About
V3KTR is a suite of browser apps for depth-aware image and video work. They share one effects library and one way of reading a picture: work out how far back everything sits first, then let every effect use that map.
Processor, for images and video, is live. Kompositor, for editing, and Dimensor, for 3D, are in build. All three open in a browser tab, with nothing to install, and nothing you make is uploaded.
See the range in one page — all 111 frames of the reel on the home page, every one made in Processor, no two alike.
Deep FX for images and video.
Effects read depth and other attributes of your image or clip, and weave together into combinations of your own. Roll generators to explore, or work a look down to a single setting. Out come snippets, clips and final art.
Edit video, or generate the edit.
Cut your own footage, start to finish — Kompositor reads its depth the same way Processor does, and uses it to generate the edit: rules over your clips that land the cuts on the beat, instead of a timeline you drag by hand. Bring in Processor clips too, if you've got them. Export 16:9 and 9:16 from one timeline.
Tell me when it’s readyDeep FX for 3D.
Bring in a logo, type, a shape or your own 3D model and stack geometry effects on it — push, fold, shatter, grow. Light it, animate it, and render it to video or stills, ready to use on its own. Export GLB or OBJ to Blender, or send the render to Processor for its depth-aware effects on top, with exact depth and normals so they land precisely.
Tell me when it’s readyBuilt different
Layer editors, timelines and template tools treat depth as an add-on. V3KTR is built on it. A per-pixel depth map comes first, and everything else reads from it.
A monocular depth model reads a depth map from a single image: no depth scanner, no second lens, no mask painted by hand. It runs on your device and never sends the picture anywhere. Then the effects use that map: fog the background and hold the subject crisp, relight from a direction the picture was never lit from, sweep an effect front to back through the scene. Most of the library is still the flat work you already do, and it runs the way you’d expect.
Load an image or a clip and a depth model maps how far back every pixel sits, on your own device. Every effect can read that map, and all three apps share one depth space.
A look is a non-destructive stack: every effect and every parameter, saved as a blueprint you can reopen, edit and apply to anything else. Never a flattened export.
Generators roll a stack; every parameter stays exposed for fine control. No generative model touches your pixels. Each one comes from an effect you can open and turn down.
Processor, Kompositor and Dimensor pass looks, clips and 3D renders between browser tabs, depth included. WebGPU, no install, nothing uploaded.
Photoshop, After Effects, DaVinci Resolve and CapCut are built around layers and timelines. Blender is built around scenes. Canva and Figma are built around layouts. They’re great at that, and V3KTR works alongside them. It’s built around depth.
Built different. Go deep.
What’s in it
| Feature | App |
|---|---|
| Depth read from one flat image, on your device, and used by the effects as a channel |  |
| Depth-aware effects: fog, focus, relight, zoom, shatter, slice, wireframe | |
| Looks saved as blueprints: a non-destructive stack, reopened and edited | |
| Generators that roll a stack, with every parameter still exposed | |
| Real algorithms, not presets: JPEG DCT, real datamosh, Sabattier solarise, scan-processor relief | |
| Feature | App |
|---|---|
| Paste in from any app, drop a file, copy the result back out | |
| Drop an image or a video; depth is read per image, and one track runs across a video's frames | |
| Three workspaces: build by hand (FX Edit), roll variations (FX Explorer), animate (FX Motion) | |
| Masks on any layer: shape, luma, depth or paint | |
| Export stills, MP4 and WebM loops up to 4K, with no watermark | |
| Save and share Looks; import and export packs | |
| Save a whole project (image, stack, masks and motion) in one file | |
| Feature | App |
|---|---|
| Handles on the canvas: drag the points, lines, boxes and masks you can see | |
| Precise control: arrow keys nudge one display unit, Shift-drag for fine moves | |
| The preview adapts to your device; export is always full resolution | |
| Says so when it lowers quality, and recovers from a lost GPU rather than failing | |
| Feature | App |
|---|---|
| Everything runs in your browser. Nothing is sent to us. | |
| No account to start | |
| A live compatibility check tells you whether your browser will run it | |
| Free for personal use; one licence for commercial work |   |
| Feature | App |
|---|---|
| One design system across the apps and the site | |
| The whole library in one place: 102 effects in ten families | |
| Kept current: new effects arrive curated, as they are tuned | |
Nothing in V3KTR invents image content. There’s no prompt box, and nothing here will hallucinate a hand onto your image.
The Generate button rolls a blueprint, not pixels: it picks which effects go in the stack and where their settings land, and you shape the roll with faders for complexity, intensity and how hard it leans on depth. Every pixel you end up with was put there by an effect you can open, inspect and turn down. Kompositor works the same way for edits — it deals cuts from rules over your own footage.
Everything is free during the open beta — commercial work included. No account needed.
Personal use stays free after the beta. It isn’t a beta perk and it doesn’t expire.
When the beta ends, commercial use is paid: earning money with V3KTR, or promoting a business with it — client work, sales, monetised or sponsored content. There are two licences: Individual, for one person, and Studio, per seat with a bulk discount for teams. One licence covers every V3KTR app, paid once, no subscription. The price isn’t set yet; it will be announced here first, with notice before anything changes. What you’ve already made is unaffected: your exports are yours, and your saved .v3ktr files keep opening.
Content creators: if your content earns, talk to us. We work out creator deals — a rate or a promotion arrangement — rather than a one-size licence. Get in touch.
V3KTR is made by Patrick Martin, a designer who builds. The aim is an ecosystem of apps that work together, not another pile of separate tools: Processor for effects, Kompositor for editing, and Dimensor, now in the works. The name is the idea: a vector is a direction with magnitude. Each app is one, and they all start from the same depth.
Each one starts as the tool Patrick wanted to use and is shaped so anyone can pick it up: considered UX and UI, simple workflows, nothing to install. The apps share one library, so what you make in one carries straight into the next.
Three apps is where the suite stands today, not where it stops.
If something is missing, say what you’re trying to make and what’s in the way: get in touch.
Go deep.
Kompositor and Dimensor are on the way. Pick what you want: one email when each app opens, and new effects as they land.