Back to corpus
research notebacklog referencescore 24

E460 — Duncan Fewkes reel analysis digest

The prior chunk caught his recent product surface (VFX Editor, two-channel audio+motion reactivity, sunset preset, bullet-time/clones). This chunk catches the **rendering tech** under the surface:

Full HTML reader

Read the full artifact

Open in new tab

Extracted abstract or opening context

⚠ No video on disk for this reel — referenced by the playbook but not in the local ingest cache. This file aggregates every playbook chunk section that cites E460. Sections are de-duplicated by heading. The prior chunk caught his recent product surface (VFX Editor, two-channel audio+motion reactivity, sunset preset, bullet-time/clones). This chunk catches the **rendering tech** under the surface: 1. **The product is called "HOLOVIS"** — that's the brand text rendered as 3D letters in nearly every install demo. It's not a placeholder. The letters are the **payload object** of his rendering: blendshape inflated by audio, reflection probe hit-target, particle attractor, light source. Treat HOLOVIS_LETTERS as a first-class object in his architecture, not chrome. 2. **Target hardware = Quadro A6000 @ 60fps** (E470 caption verbatim: *"Target hardware is Quadro A6000 at 60 fps"*). His laptop dev box is a 3070. This sets the LUME perf target ceiling — A6000 ≈ RTX 4080. Anything we plan that wouldn't run on a single 4080 at 60fps is over-budget for parity. 3. **The calibration grid IS the LED stage UI** — the gridded floor + walls + circular degree-marked "calibration markers" you see in every recent reel are actually his runtime spatial reference, used for depth-camera reprojection and multi-screen LED alignment. Not just decoration. He runs an "LED-backed XR / CAVE-like volume" (Gemini on E460/E466/E472). 4. **Two render-mode profiles ship together** (E473 visual analysis): `tint Lights, No Shadows` (160-170 fps, ambient/emissive only) and `Spot Lights, Shadows` (70-110 fps, cast shadows from 7x point/spot lights). Operator can A/B at runtime. **Add to LumeVfxEditor: a `LightingProfile` dropdown** with at least these two modes — fast/cheap and full/cinematic. 5. **Reflection-probe budget**: realtime reflection probe render every frame across 6 faces is a known toggle (E460 caption) — adds depth/colour but he calls it "as an option toggled in settings." Same toggle slot in LUME. 6. **Depth-camera kaleidoscope** is a discrete VFX preset he calls out explicitly (E469 first introduces it, E471 adds metallic vs non-metallic, E472 stretches to full body). **Net new mode for LUME** not in the prior playbook. 7. **Marching Cubes is Keijiro's**, confirmed (E427 caption verbatim: `https://github.com/keijiro/ComputeMarchingCubes`). Source particles → MC compute shader → mesh. He runs **2 surface thresholds simultaneously** to get nested skins (inner gold, outer glass). Caption progression: - E469: *"shoulda done it sooner — Amazing the daft poses that some rotational symmetry prompts you to hit"* - E471: tests metallic vs non-metallic surface, adds "test box environment so the metallic surface has something to reflect" - E472: extends from upper-body-only to full

Promotion decision

What has to happen next

Keep in the searchable backlog until it intersects a live paper or system.

Why this is not always a full paper yet

Corpus pages are public-safe readers for discovered workspace artifacts. They are not automatically final papers. A corpus item becomes a polished paper only after the editable source, evidence checkpoints, references, figures, render path, and release status are attached through the paper schema.