Lacus Post
May 25, 2026
Argus comes from identifying those friction points in real productions and building a tool that eliminates them, with automatic calculations, native exporters to VFX software, and AI integration to accelerate data entry without sacrificing precision.
VFX on-set documentation rethought from the ground up, designed for modern production workflows.
What is Argus?
Argus is a VFX on-set documentation system developed internally at Lacus Post, built as a native iPad application. Its purpose is to capture, organize, and export all critical shot information directly from set, aligned with the VES On-Set Reference Guide standards but designed to function as a modern tool. not as a digitized form.
The name Argus, taken from the hundred-eyed giant of Greek mythology, reflects the software's philosophy: to watch over and record every detail of the set that will be critical when the footage reaches post-production.
The problem Argus solves
On-set documentation has evolved, but not enough. Today, most productions work with editable PDFs of the VES Guide, shared spreadsheets in Google Drive, generic note-taking apps, or custom forms built in platforms like FileMaker or Airtable. They work, but each carries its own limitations: data fragmentation across multiple tools, lack of specific fields for modern VFX setups (LED walls, witness cams, lidar, motion capture), reliance on connectivity for cloud sync, and exporters that don't speak the language of the DCCs.
Argus comes from identifying those friction points in real productions and building a tool that eliminates them, with automatic calculations, native exporters to VFX software, and AI integration to accelerate data entry without sacrificing precision.
What Argus does
Structured Shot Entry across 6 specialized tabs. Each shot is documented through tabs organized by category: Camera & Lens, GPS & Location, VFX Classification, Lighting & Scans, Notes, and Plate Deliverables. More than 70 fields cover priorities 1, 2, and 3 of the VES On-Set Reference Guide, accounting for modern setups: virtual production with LED walls, witness cameras, lidar captures, motion capture, HDRi, and multi-reference workflows.
GPS + Sun Position calculated automatically. Argus pulls GPS coordinates from the iPad and the exact capture time, then computes solar azimuth and elevation in real time. It detects golden hour and blue hour. This gives the CG lighting team a concrete, verifiable data point about the sun's position for every shot, without having to reconstruct it from photo references in post.
One-click DCC exporters. From the solar calculation, Argus generates ready-to-paste scripts for Nuke, Houdini, Maya (MEL), and Blender. The TD or lighting artist opens the script, runs it, and gets a sun light positioned with the same orientation it had on set. This replaces a process that normally takes minutes of manual work per shot and eliminates transcription errors.
Reference Hub with annotation tools. Integrated multi-document viewer for PDFs and images, with pen and eraser tools. Up to 10 documents open in simultaneous tabs, storyboards, lookbooks, set diagrams, visual references, each with its own annotation set and notes. All reference documentation lives on the same device where the set data is being captured.
Customizable On-Set Checklist. Per-project templates with a default VFX supervision checklist organized in phases (Before / During / Data & Backup / End of Day), plus the ability to create custom templates tailored to the specifics of each production.
Professional PDF export. Multi-page breakdown report with cover page, shot cards color-coded by VFX complexity, and a summary grouped by complexity tier. Designed to be legible for the post team as well as production and client.
Offline-first. All data lives locally on the iPad. There's no forced cloud sync, no telemetry, no external service dependencies during use. This is critical in remote locations where WiFi or cellular connectivity is unstable or nonexistent, and also from a privacy and data-ownership perspective in productions where on-set information is confidential.
Claude AI integration - AI Auto-Fill
Argus's most distinctive feature is its integration with the Claude API (Anthropic), which
fundamentally changes how a breakdown sheet gets filled out on set.
What does the API do?
Traditionally, documenting a shot requires manually typing between 20 and 40 fields per setup. On productions with 80–120 setups per day, that translates into hours of typing that compete with actual supervision work. AI Auto-Fill solves this by allowing the information to be captured through photos and free-form description.
How it works in production.
The user uploads one or several photos of the set from the iPad camera — typically one of the slate, one of the camera/lens, and one of the general setup. They optionally write a brief description ("Sc 14A, shot 030, T2, greenscreen, HDRi captured") or use Quick Tags for the most common setups (Greenscreen, HDRi, Hero Shot, LED Wall, etc.).
Argus sends the images and text to the Claude API with the project context. Claude visually analyzes the photos, reads the slate, identifies the camera setup, detects lighting type and background, and returns up to 38 completed fields in 10–30 seconds. The supervisor reviews, adjusts whatever is necessary, and saves the shot.
Intelligent Override mode. By default, a second AI pass only fills empty fields without overwriting whatever the supervisor entered manually. If the user wants to replace everything with a new analysis (for example, when the first pass was inaccurate due to missing photos), they simply toggle the override on. For text note fields, instead of overwriting, the AI appends its analysis at the end with an [AI added:] prefix, always preserving what the user wrote. The user's API key is stored locally once. It's the user's own Anthropic key, no middleman, no hidden billing, no project data leaving toward third-party services beyond the call to the model itself.
Technical stack
Argus is built as a single-file HTML application with vanilla JavaScript — no heavy frameworks, no build step, no unnecessary dependencies. This architectural decision is deliberate: a single HTML page is trivial to audit, maintain, and deploy; and it eliminates the update complexity that native App Store apps carry.
It's optimized for Safari on iPad and installs as a PWA via "Add to Home Screen," opening in standalone mode without browser bars. jsPDF and PDF.js are loaded on demand from a public CDN only when PDF functions are needed, the rest of the app runs 100% offline once installed.
The innovation behind Argus
Argus is not a digitization of the VES breakdown sheet. It's a complete rethinking of the on-set documentation process, built around three principles:
Visual capture over textual capture. The iPad's cameras combined with analytical AI make it more efficient to photograph the slate than to transcribe it. Argus bets on that workflow.
Computation over note-taking. Sun position, golden hour, unit conversions, anything that can be computed shouldn't have to be written. Argus computes.
Output to pipeline, not to a file. Argus's primary output is not a static PDF; it's executable scripts, structured data, and exporters that connect directly with the VFX tools that will be used in post. The breakdown stops being a reference document and becomes the starting point of the post-production work.
Argus - VFX On-Set, by Lacus Post.
