docs(prompting): lower-third renders behind the subject — match film behavior

docs/prompting/capstone.mdx

@@ -57,7 +57,7 @@ This is the full prompt, nothing paraphrased or trimmed. Two of its clauses —
 >
 > **Everywhere (21–28s).** The wire arcs and the ground falls away: adapt the **`world-map` registry block** — the dotted world map scrolls beneath while the wire becomes a great-circle route arcing across it. Real render infrastructure, truthfully labeled: pulses light up at AWS and Google Cloud regions (US-EAST-1 · EU-WEST-1 · AP-SOUTH-1) with a "CLOUD RENDER · LAMBDA + CLOUD RUN" kicker — HyperFrames' actual deploy targets. The map tilts away as the wire descends into the next region.
 >
-> **Material (28–39s).** The wire threads into a working canvas where footage is the material on the timeline: a real talking-head clip (generate a short clip of a person speaking one neutral line via the media pipeline's avatar video generation — it is material being processed, not a testimonial; **transcode it to HEVC `hvc1`** so the automatic proxy subsystem carries preview) sits as a clip on the wire. The order of operations IS the story: as the camera arrives and BEFORE the person speaks, the framework mattes the footage — **the background peels away via background removal**, sliding off along the travel direction and leaving the cutout standing alone on the brand ground. THEN they speak, and the main **keywords of their own line — derived from the clip's transcription — land word-synced as huge display text BEHIND the cutout**, each keyword slamming in on its spoken moment with the subject's silhouette occluding it (the two-layer text-behind-subject plate); style the keyword type by adapting a bold **catalog caption component** (`caption-kinetic-slam` or similar) at display scale. A designed lower-third chip docks beside them. The clip's own audio ducks the BGM briefly; the VO resumes as the camera pulls away and the worked footage recedes by parallax.
+> **Material (28–39s).** The wire threads into a working canvas where footage is the material on the timeline: a real talking-head clip (generate a short clip of a person speaking one neutral line via the media pipeline's avatar video generation — it is material being processed, not a testimonial; **transcode it to HEVC `hvc1`** so the automatic proxy subsystem carries preview) sits as a clip on the wire. The order of operations IS the story: as the camera arrives and BEFORE the person speaks, the framework mattes the footage — **the background peels away via background removal**, sliding off along the travel direction and leaving the cutout standing alone on the brand ground. THEN they speak, and the main **keywords of their own line — derived from the clip's transcription — land word-synced as huge display text BEHIND the cutout**, each keyword slamming in on its spoken moment with the subject's silhouette occluding it (the two-layer text-behind-subject plate); style the keyword type by adapting a bold **catalog caption component** (`caption-kinetic-slam` or similar) at display scale. A designed lower-third chip renders behind the subject. The clip's own audio ducks the BGM briefly; the VO resumes as the camera pulls away and the worked footage recedes by parallax.
 >
 > **Surface (39–45s).** SANCTIONED SEAM #1: the camera pushes through an **`sdf-iris` shader transition** — the iris opens ONTO the continuation of the same wire (the wire is visible through the iris throughout; this is a lens the journey passes through, not a cut). Inside: frosted-glass inspector panels (real translucency — blur over what's behind; the product's Studio design language) hover above the wire carrying live values (`ease: power3.out`, a color token, an fps readout), over a **generated-artwork backdrop** — a flat abstract mural evoking motion trails, generated **palette-neutral (grayscale) and duotoned live by CSS** from the two variables so it re-skins with everything else. **The mural is pinned in world space and never translates: the widening iris circle alone reveals it (the mural is already in place behind the mask), and when the camera moves on it exits by fading out — it must never slide into or out of frame like a panel.** Drifting `data-*` tags ride between mural and glass, clear of all text.
 >
@@ -89,7 +89,7 @@ Every region pins a different technique, and every technique has a chapter:
 | Mount | The timeline's own vocabulary — tracks, clips, `data-*` labels | [Prompt anatomy](/prompting/anatomy) |
 | Measure | `data-chart` block; the wire as chart baseline (matched motion) | [Data and maps](/prompting/data-and-maps), [Transitions](/prompting/transitions) |
 | Everywhere | `world-map` block; real Lambda / Cloud Run regions | [Data and maps](/prompting/data-and-maps), [Rendering and output](/prompting/rendering-and-output) |
-| Material | Generated clip → HEVC proxy → background removed *before* the speech → the line's keywords slam in word-synced as display text BEHIND the cutout (silhouette occlusion) → lower-third | [Media and audio](/prompting/media-and-audio), [Caption styles](/prompting/captions-catalog), [Overlays and lower thirds](/prompting/overlays-and-lower-thirds) |
+| Material | Generated clip → HEVC proxy → background removed *before* the speech → the line's keywords slam in word-synced as display text BEHIND the cutout (silhouette occlusion) → lower-third behind the subject | [Media and audio](/prompting/media-and-audio), [Caption styles](/prompting/captions-catalog), [Overlays and lower thirds](/prompting/overlays-and-lower-thirds) |
 | Surface | `sdf-iris` shader lens, frosted glass, palette-neutral generated mural | [Transitions](/prompting/transitions), [VFX and liquid glass](/prompting/vfx-and-liquid-glass), [When to generate artwork](/prompting/generated-artwork) |
 | Rhythm | Cuts on a real analyzed beat grid (`hyperframes beats`) | [Music videos and slideshows](/prompting/music-and-slideshows) |
 | Depth | Real Three.js via the frame adapter, wire as 3D coil | [Runtimes and 3D](/prompting/runtimes-and-3d) |