@opengamehq/skills 1.3.0 → 1.5.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/CHANGELOG.md +44 -0
- package/README.md +12 -4
- package/dist/.claude-plugin/marketplace.json +26 -0
- package/dist/skills/manifest.json +1409 -145
- package/dist/skills/opengame-accessibility/CHANGELOG.md +4 -0
- package/dist/skills/opengame-accessibility/SKILL.md +48 -32
- package/dist/skills/opengame-all-in-one/CHANGELOG.md +21 -0
- package/dist/skills/opengame-all-in-one/SKILL.md +195 -0
- package/dist/skills/opengame-all-in-one/references/opengame-accessibility/README.md +102 -0
- package/dist/skills/opengame-all-in-one/references/opengame-api/README.md +107 -0
- package/dist/skills/opengame-all-in-one/references/opengame-assets/README.md +168 -0
- package/dist/skills/opengame-all-in-one/references/opengame-assets/references/pipeline.md +164 -0
- package/dist/skills/opengame-all-in-one/references/opengame-audio/README.md +238 -0
- package/dist/skills/opengame-all-in-one/references/opengame-audio/helpers/audio-chain.js +325 -0
- package/dist/skills/opengame-all-in-one/references/opengame-audio/references/mix-notes.md +125 -0
- package/dist/skills/opengame-all-in-one/references/opengame-audio/references/qa-without-ears.md +177 -0
- package/dist/skills/opengame-all-in-one/references/opengame-capture/README.md +128 -0
- package/dist/skills/opengame-all-in-one/references/opengame-capture/references/pipeline.md +114 -0
- package/dist/skills/opengame-all-in-one/references/opengame-capture/references/threejs.md +161 -0
- package/dist/skills/opengame-all-in-one/references/opengame-characters/README.md +180 -0
- package/dist/skills/opengame-all-in-one/references/opengame-characters/references/first-person-body.md +108 -0
- package/dist/skills/opengame-all-in-one/references/opengame-characters/references/integration.md +144 -0
- package/dist/skills/opengame-all-in-one/references/opengame-characters/references/player-og.md +160 -0
- package/dist/skills/opengame-all-in-one/references/opengame-cli/README.md +189 -0
- package/dist/skills/opengame-all-in-one/references/opengame-deploy/README.md +223 -0
- package/dist/skills/opengame-all-in-one/references/opengame-deploy/references/json-events.md +63 -0
- package/dist/skills/opengame-all-in-one/references/opengame-deploy/references/manifest.md +89 -0
- package/dist/skills/opengame-all-in-one/references/opengame-deploy/references/packaging.md +119 -0
- package/dist/skills/opengame-all-in-one/references/opengame-depth/README.md +154 -0
- package/dist/skills/opengame-all-in-one/references/opengame-enemies/README.md +155 -0
- package/dist/skills/opengame-all-in-one/references/opengame-enemies/references/pacing-and-bots.md +108 -0
- package/dist/skills/opengame-all-in-one/references/opengame-enemies/references/threejs.md +193 -0
- package/dist/skills/opengame-all-in-one/references/opengame-front-end/README.md +185 -0
- package/dist/skills/opengame-all-in-one/references/opengame-front-end/references/launch.md +50 -0
- package/dist/skills/opengame-all-in-one/references/opengame-front-end/references/menus.md +81 -0
- package/dist/skills/opengame-all-in-one/references/opengame-game-feel/README.md +181 -0
- package/dist/skills/opengame-all-in-one/references/opengame-game-feel/helpers/spring.js +146 -0
- package/dist/skills/opengame-all-in-one/references/opengame-game-feel/references/feel-numbers.md +142 -0
- package/dist/skills/opengame-all-in-one/references/opengame-game-feel/references/killcam.md +88 -0
- package/dist/skills/opengame-all-in-one/references/opengame-jam/README.md +114 -0
- package/dist/skills/opengame-all-in-one/references/opengame-lighting/README.md +141 -0
- package/dist/skills/opengame-all-in-one/references/opengame-lighting/helpers/light-pool.js +185 -0
- package/dist/skills/opengame-all-in-one/references/opengame-loading/README.md +135 -0
- package/dist/skills/opengame-all-in-one/references/opengame-loading/helpers/loading-count.mjs +115 -0
- package/dist/skills/opengame-all-in-one/references/opengame-loading/references/threejs.md +141 -0
- package/dist/skills/opengame-all-in-one/references/opengame-look/README.md +181 -0
- package/dist/skills/opengame-all-in-one/references/opengame-look/references/matching-a-reference.md +73 -0
- package/dist/skills/opengame-all-in-one/references/opengame-look/references/painterly.md +69 -0
- package/dist/skills/opengame-all-in-one/references/opengame-look/references/post-and-lens.md +73 -0
- package/dist/skills/opengame-all-in-one/references/opengame-maps/README.md +74 -0
- package/dist/skills/opengame-all-in-one/references/opengame-maps/references/map-spec.md +77 -0
- package/dist/skills/opengame-all-in-one/references/opengame-mobile/README.md +86 -0
- package/dist/skills/opengame-all-in-one/references/opengame-montage-make/README.md +183 -0
- package/dist/skills/opengame-all-in-one/references/opengame-montage-make/references/briefs.md +293 -0
- package/dist/skills/opengame-all-in-one/references/opengame-montage-make/references/day-one-questions.md +46 -0
- package/dist/skills/opengame-all-in-one/references/opengame-montage-make/references/edge-cases.md +27 -0
- package/dist/skills/opengame-all-in-one/references/opengame-montage-make/references/filmable.md +244 -0
- package/dist/skills/opengame-all-in-one/references/opengame-montage-make/references/finishing.md +126 -0
- package/dist/skills/opengame-all-in-one/references/opengame-montage-make/references/measure.md +282 -0
- package/dist/skills/opengame-all-in-one/references/opengame-montage-make/references/montage-capture.md +151 -0
- package/dist/skills/opengame-all-in-one/references/opengame-montage-make/references/montage-gates.md +84 -0
- package/dist/skills/opengame-all-in-one/references/opengame-montage-make/references/montage-genres.md +198 -0
- package/dist/skills/opengame-all-in-one/references/opengame-montage-make/references/montage-grammar.md +320 -0
- package/dist/skills/opengame-all-in-one/references/opengame-montage-make/references/montage-ops.md +127 -0
- package/dist/skills/opengame-all-in-one/references/opengame-montage-make/references/og-snipers-record.md +110 -0
- package/dist/skills/opengame-all-in-one/references/opengame-montage-make/references/owner-eye.md +214 -0
- package/dist/skills/opengame-all-in-one/references/opengame-montage-make/references/reference-study.md +293 -0
- package/dist/skills/opengame-all-in-one/references/opengame-montage-make/references/song.md +64 -0
- package/dist/skills/opengame-all-in-one/references/opengame-montage-plan/README.md +193 -0
- package/dist/skills/opengame-all-in-one/references/opengame-montage-plan/references/briefs.md +293 -0
- package/dist/skills/opengame-all-in-one/references/opengame-montage-plan/references/day-one-questions.md +46 -0
- package/dist/skills/opengame-all-in-one/references/opengame-montage-plan/references/filmable.md +244 -0
- package/dist/skills/opengame-all-in-one/references/opengame-montage-plan/references/measure.md +282 -0
- package/dist/skills/opengame-all-in-one/references/opengame-montage-plan/references/montage-capture.md +151 -0
- package/dist/skills/opengame-all-in-one/references/opengame-montage-plan/references/montage-gates.md +84 -0
- package/dist/skills/opengame-all-in-one/references/opengame-montage-plan/references/montage-genres.md +198 -0
- package/dist/skills/opengame-all-in-one/references/opengame-montage-plan/references/montage-grammar.md +320 -0
- package/dist/skills/opengame-all-in-one/references/opengame-montage-plan/references/montage-ops.md +127 -0
- package/dist/skills/opengame-all-in-one/references/opengame-montage-plan/references/og-snipers-record.md +110 -0
- package/dist/skills/opengame-all-in-one/references/opengame-montage-plan/references/owner-eye.md +214 -0
- package/dist/skills/opengame-all-in-one/references/opengame-montage-plan/references/reference-study.md +293 -0
- package/dist/skills/opengame-all-in-one/references/opengame-montage-plan/references/song.md +64 -0
- package/dist/skills/opengame-all-in-one/references/opengame-montage-review-and-improve/README.md +174 -0
- package/dist/skills/opengame-all-in-one/references/opengame-montage-review-and-improve/references/briefs.md +293 -0
- package/dist/skills/opengame-all-in-one/references/opengame-montage-review-and-improve/references/day-one-questions.md +46 -0
- package/dist/skills/opengame-all-in-one/references/opengame-montage-review-and-improve/references/filmable.md +244 -0
- package/dist/skills/opengame-all-in-one/references/opengame-montage-review-and-improve/references/finishing.md +126 -0
- package/dist/skills/opengame-all-in-one/references/opengame-montage-review-and-improve/references/measure.md +282 -0
- package/dist/skills/opengame-all-in-one/references/opengame-montage-review-and-improve/references/montage-capture.md +151 -0
- package/dist/skills/opengame-all-in-one/references/opengame-montage-review-and-improve/references/montage-gates.md +84 -0
- package/dist/skills/opengame-all-in-one/references/opengame-montage-review-and-improve/references/montage-genres.md +198 -0
- package/dist/skills/opengame-all-in-one/references/opengame-montage-review-and-improve/references/montage-grammar.md +320 -0
- package/dist/skills/opengame-all-in-one/references/opengame-montage-review-and-improve/references/montage-ops.md +127 -0
- package/dist/skills/opengame-all-in-one/references/opengame-montage-review-and-improve/references/og-snipers-record.md +110 -0
- package/dist/skills/opengame-all-in-one/references/opengame-montage-review-and-improve/references/owner-eye.md +214 -0
- package/dist/skills/opengame-all-in-one/references/opengame-montage-review-and-improve/references/reference-study.md +293 -0
- package/dist/skills/opengame-all-in-one/references/opengame-montage-review-and-improve/references/song.md +64 -0
- package/dist/skills/opengame-all-in-one/references/opengame-multiplayer/README.md +139 -0
- package/dist/skills/opengame-all-in-one/references/opengame-multiplayer/references/bot-lobbies.md +71 -0
- package/dist/skills/opengame-all-in-one/references/opengame-multiplayer/references/netcode.md +165 -0
- package/dist/skills/opengame-all-in-one/references/opengame-performance/README.md +177 -0
- package/dist/skills/opengame-all-in-one/references/opengame-performance/helpers/landing-pacer.mjs +86 -0
- package/dist/skills/opengame-all-in-one/references/opengame-performance/references/measuring.md +123 -0
- package/dist/skills/opengame-all-in-one/references/opengame-performance/references/webgpu.md +86 -0
- package/dist/skills/opengame-all-in-one/references/opengame-production/README.md +192 -0
- package/dist/skills/opengame-all-in-one/references/opengame-production/references/team.md +107 -0
- package/dist/skills/opengame-all-in-one/references/opengame-production/references/templates.md +153 -0
- package/dist/skills/opengame-all-in-one/references/opengame-quality-bar/README.md +170 -0
- package/dist/skills/opengame-all-in-one/references/opengame-quality-bar/references/cheap-tells.md +43 -0
- package/dist/skills/opengame-all-in-one/references/opengame-quality-bar/references/judges.md +95 -0
- package/dist/skills/opengame-all-in-one/references/opengame-quality-bar/references/review-prompt.md +32 -0
- package/dist/skills/opengame-all-in-one/references/opengame-rhythm/README.md +134 -0
- package/dist/skills/opengame-all-in-one/references/opengame-rhythm/references/charting.md +109 -0
- package/dist/skills/opengame-all-in-one/references/opengame-rhythm/references/timing.md +92 -0
- package/dist/skills/opengame-all-in-one/references/opengame-router/README.md +54 -0
- package/dist/skills/opengame-all-in-one/references/opengame-router/references/process.md +38 -0
- package/dist/skills/opengame-all-in-one/references/opengame-scene-audit/README.md +136 -0
- package/dist/skills/opengame-all-in-one/references/opengame-scene-audit/helpers/scene-audit.js +677 -0
- package/dist/skills/opengame-all-in-one/references/opengame-sdk/README.md +283 -0
- package/dist/skills/opengame-all-in-one/references/opengame-sdk/references/engines.md +93 -0
- package/dist/skills/opengame-all-in-one/references/opengame-sdk/references/frame.md +107 -0
- package/dist/skills/opengame-all-in-one/references/opengame-sdk/references/identity.md +111 -0
- package/dist/skills/opengame-all-in-one/references/opengame-sdk/references/leaderboards.md +115 -0
- package/dist/skills/opengame-all-in-one/references/opengame-sdk/references/radio.md +71 -0
- package/dist/skills/opengame-all-in-one/references/opengame-stability/README.md +112 -0
- package/dist/skills/opengame-all-in-one/references/opengame-trailer-make/README.md +160 -0
- package/dist/skills/opengame-all-in-one/references/opengame-trailer-make/references/archetypes.md +113 -0
- package/dist/skills/opengame-all-in-one/references/opengame-trailer-make/references/briefs.md +273 -0
- package/dist/skills/opengame-all-in-one/references/opengame-trailer-make/references/core.md +129 -0
- package/dist/skills/opengame-all-in-one/references/opengame-trailer-make/references/craft.md +299 -0
- package/dist/skills/opengame-all-in-one/references/opengame-trailer-make/references/edge-cases.md +27 -0
- package/dist/skills/opengame-all-in-one/references/opengame-trailer-make/references/filmable.md +244 -0
- package/dist/skills/opengame-all-in-one/references/opengame-trailer-make/references/finishing.md +126 -0
- package/dist/skills/opengame-all-in-one/references/opengame-trailer-make/references/gates.md +184 -0
- package/dist/skills/opengame-all-in-one/references/opengame-trailer-make/references/measure.md +282 -0
- package/dist/skills/opengame-all-in-one/references/opengame-trailer-make/references/music.md +148 -0
- package/dist/skills/opengame-all-in-one/references/opengame-trailer-make/references/og-clash-record.md +133 -0
- package/dist/skills/opengame-all-in-one/references/opengame-trailer-make/references/ops.md +154 -0
- package/dist/skills/opengame-all-in-one/references/opengame-trailer-make/references/owner-eye.md +214 -0
- package/dist/skills/opengame-all-in-one/references/opengame-trailer-make/references/review.md +141 -0
- package/dist/skills/opengame-all-in-one/references/opengame-trailer-make/references/title.md +92 -0
- package/dist/skills/opengame-all-in-one/references/opengame-trailer-plan/README.md +172 -0
- package/dist/skills/opengame-all-in-one/references/opengame-trailer-plan/references/archetypes.md +113 -0
- package/dist/skills/opengame-all-in-one/references/opengame-trailer-plan/references/briefs.md +273 -0
- package/dist/skills/opengame-all-in-one/references/opengame-trailer-plan/references/core.md +129 -0
- package/dist/skills/opengame-all-in-one/references/opengame-trailer-plan/references/craft.md +299 -0
- package/dist/skills/opengame-all-in-one/references/opengame-trailer-plan/references/edge-cases.md +27 -0
- package/dist/skills/opengame-all-in-one/references/opengame-trailer-plan/references/filmable.md +244 -0
- package/dist/skills/opengame-all-in-one/references/opengame-trailer-plan/references/finishing.md +126 -0
- package/dist/skills/opengame-all-in-one/references/opengame-trailer-plan/references/gates.md +184 -0
- package/dist/skills/opengame-all-in-one/references/opengame-trailer-plan/references/measure.md +282 -0
- package/dist/skills/opengame-all-in-one/references/opengame-trailer-plan/references/music.md +148 -0
- package/dist/skills/opengame-all-in-one/references/opengame-trailer-plan/references/og-clash-record.md +133 -0
- package/dist/skills/opengame-all-in-one/references/opengame-trailer-plan/references/ops.md +154 -0
- package/dist/skills/opengame-all-in-one/references/opengame-trailer-plan/references/owner-eye.md +214 -0
- package/dist/skills/opengame-all-in-one/references/opengame-trailer-plan/references/reference-set.md +92 -0
- package/dist/skills/opengame-all-in-one/references/opengame-trailer-plan/references/review.md +141 -0
- package/dist/skills/opengame-all-in-one/references/opengame-trailer-plan/references/title.md +92 -0
- package/dist/skills/opengame-all-in-one/references/opengame-trailer-review-and-improve/README.md +141 -0
- package/dist/skills/opengame-all-in-one/references/opengame-trailer-review-and-improve/references/archetypes.md +113 -0
- package/dist/skills/opengame-all-in-one/references/opengame-trailer-review-and-improve/references/briefs.md +273 -0
- package/dist/skills/opengame-all-in-one/references/opengame-trailer-review-and-improve/references/core.md +129 -0
- package/dist/skills/opengame-all-in-one/references/opengame-trailer-review-and-improve/references/craft.md +299 -0
- package/dist/skills/opengame-all-in-one/references/opengame-trailer-review-and-improve/references/edge-cases.md +27 -0
- package/dist/skills/opengame-all-in-one/references/opengame-trailer-review-and-improve/references/filmable.md +244 -0
- package/dist/skills/opengame-all-in-one/references/opengame-trailer-review-and-improve/references/finishing.md +126 -0
- package/dist/skills/opengame-all-in-one/references/opengame-trailer-review-and-improve/references/gates.md +184 -0
- package/dist/skills/opengame-all-in-one/references/opengame-trailer-review-and-improve/references/measure.md +282 -0
- package/dist/skills/opengame-all-in-one/references/opengame-trailer-review-and-improve/references/music.md +148 -0
- package/dist/skills/opengame-all-in-one/references/opengame-trailer-review-and-improve/references/og-clash-record.md +133 -0
- package/dist/skills/opengame-all-in-one/references/opengame-trailer-review-and-improve/references/ops.md +154 -0
- package/dist/skills/opengame-all-in-one/references/opengame-trailer-review-and-improve/references/owner-eye.md +214 -0
- package/dist/skills/opengame-all-in-one/references/opengame-trailer-review-and-improve/references/review.md +141 -0
- package/dist/skills/opengame-all-in-one/references/opengame-trailer-review-and-improve/references/title.md +92 -0
- package/dist/skills/opengame-all-in-one/references/opengame-verification/README.md +197 -0
- package/dist/skills/opengame-all-in-one/references/opengame-verification/helpers/browser-evidence.mjs +74 -0
- package/dist/skills/opengame-all-in-one/references/opengame-weapons/README.md +179 -0
- package/dist/skills/opengame-all-in-one/references/opengame-weapons/references/handling-numbers.md +114 -0
- package/dist/skills/opengame-all-in-one/references/opengame-weapons/references/reloads-and-hands.md +96 -0
- package/dist/skills/opengame-all-in-one/references/opengame-weapons/references/sights-and-scopes.md +98 -0
- package/dist/skills/opengame-all-in-one/references/opengame-weapons/references/threejs.md +240 -0
- package/dist/skills/opengame-all-in-one/references/opengame-world-physics/README.md +141 -0
- package/dist/skills/opengame-all-in-one/references/opengame-world-physics/references/measured-failures.md +79 -0
- package/dist/skills/opengame-api/CHANGELOG.md +12 -0
- package/dist/skills/opengame-api/SKILL.md +39 -13
- package/dist/skills/opengame-assets/CHANGELOG.md +11 -0
- package/dist/skills/opengame-assets/SKILL.md +79 -31
- package/dist/skills/opengame-assets/references/pipeline.md +67 -0
- package/dist/skills/opengame-audio/CHANGELOG.md +11 -0
- package/dist/skills/opengame-audio/SKILL.md +129 -41
- package/dist/skills/opengame-audio/references/mix-notes.md +44 -1
- package/dist/skills/opengame-audio/references/qa-without-ears.md +177 -0
- package/dist/skills/opengame-capture/CHANGELOG.md +10 -0
- package/dist/skills/opengame-capture/SKILL.md +44 -20
- package/dist/skills/opengame-capture/references/pipeline.md +33 -1
- package/dist/skills/opengame-capture/references/threejs.md +17 -1
- package/dist/skills/opengame-characters/CHANGELOG.md +15 -0
- package/dist/skills/opengame-characters/SKILL.md +94 -39
- package/dist/skills/opengame-characters/references/first-person-body.md +108 -0
- package/dist/skills/opengame-characters/references/player-og.md +1 -1
- package/dist/skills/opengame-cli/CHANGELOG.md +20 -0
- package/dist/skills/opengame-cli/SKILL.md +12 -11
- package/dist/skills/opengame-deploy/CHANGELOG.md +30 -0
- package/dist/skills/opengame-deploy/SKILL.md +43 -23
- package/dist/skills/opengame-deploy/references/json-events.md +6 -5
- package/dist/skills/opengame-deploy/references/manifest.md +9 -3
- package/dist/skills/opengame-deploy/references/packaging.md +32 -30
- package/dist/skills/opengame-depth/CHANGELOG.md +8 -0
- package/dist/skills/opengame-depth/SKILL.md +61 -33
- package/dist/skills/opengame-enemies/CHANGELOG.md +10 -0
- package/dist/skills/opengame-enemies/SKILL.md +96 -40
- package/dist/skills/opengame-enemies/references/pacing-and-bots.md +108 -0
- package/dist/skills/opengame-enemies/references/threejs.md +59 -0
- package/dist/skills/opengame-front-end/CHANGELOG.md +4 -0
- package/dist/skills/opengame-front-end/SKILL.md +85 -39
- package/dist/skills/opengame-front-end/references/menus.md +42 -8
- package/dist/skills/opengame-game-feel/CHANGELOG.md +11 -0
- package/dist/skills/opengame-game-feel/SKILL.md +95 -47
- package/dist/skills/opengame-game-feel/references/feel-numbers.md +59 -0
- package/dist/skills/opengame-game-feel/references/killcam.md +88 -0
- package/dist/skills/opengame-jam/CHANGELOG.md +12 -0
- package/dist/skills/opengame-jam/SKILL.md +48 -33
- package/dist/skills/opengame-lighting/CHANGELOG.md +9 -0
- package/dist/skills/opengame-lighting/SKILL.md +85 -34
- package/dist/skills/opengame-loading/CHANGELOG.md +4 -0
- package/dist/skills/opengame-loading/SKILL.md +78 -38
- package/dist/skills/opengame-loading/helpers/loading-count.mjs +115 -0
- package/dist/skills/opengame-loading/references/threejs.md +32 -7
- package/dist/skills/opengame-look/CHANGELOG.md +9 -0
- package/dist/skills/opengame-look/SKILL.md +122 -51
- package/dist/skills/opengame-look/references/matching-a-reference.md +73 -0
- package/dist/skills/opengame-look/references/painterly.md +69 -0
- package/dist/skills/opengame-look/references/post-and-lens.md +10 -3
- package/dist/skills/opengame-maps/CHANGELOG.md +9 -0
- package/dist/skills/opengame-maps/SKILL.md +78 -0
- package/dist/skills/opengame-maps/references/map-spec.md +77 -0
- package/dist/skills/opengame-mobile/CHANGELOG.md +4 -0
- package/dist/skills/opengame-mobile/SKILL.md +18 -18
- package/dist/skills/opengame-montage-make/CHANGELOG.md +5 -0
- package/dist/skills/opengame-montage-make/SKILL.md +187 -0
- package/dist/skills/opengame-montage-make/references/briefs.md +293 -0
- package/dist/skills/opengame-montage-make/references/day-one-questions.md +46 -0
- package/dist/skills/opengame-montage-make/references/edge-cases.md +27 -0
- package/dist/skills/opengame-montage-make/references/filmable.md +244 -0
- package/dist/skills/opengame-montage-make/references/finishing.md +126 -0
- package/dist/skills/opengame-montage-make/references/measure.md +282 -0
- package/dist/skills/opengame-montage-make/references/montage-capture.md +151 -0
- package/dist/skills/opengame-montage-make/references/montage-gates.md +84 -0
- package/dist/skills/opengame-montage-make/references/montage-genres.md +198 -0
- package/dist/skills/opengame-montage-make/references/montage-grammar.md +320 -0
- package/dist/skills/opengame-montage-make/references/montage-ops.md +127 -0
- package/dist/skills/opengame-montage-make/references/og-snipers-record.md +110 -0
- package/dist/skills/opengame-montage-make/references/owner-eye.md +214 -0
- package/dist/skills/opengame-montage-make/references/reference-study.md +293 -0
- package/dist/skills/opengame-montage-make/references/song.md +64 -0
- package/dist/skills/opengame-montage-plan/CHANGELOG.md +5 -0
- package/dist/skills/opengame-montage-plan/SKILL.md +197 -0
- package/dist/skills/opengame-montage-plan/references/briefs.md +293 -0
- package/dist/skills/opengame-montage-plan/references/day-one-questions.md +46 -0
- package/dist/skills/opengame-montage-plan/references/filmable.md +244 -0
- package/dist/skills/opengame-montage-plan/references/measure.md +282 -0
- package/dist/skills/opengame-montage-plan/references/montage-capture.md +151 -0
- package/dist/skills/opengame-montage-plan/references/montage-gates.md +84 -0
- package/dist/skills/opengame-montage-plan/references/montage-genres.md +198 -0
- package/dist/skills/opengame-montage-plan/references/montage-grammar.md +320 -0
- package/dist/skills/opengame-montage-plan/references/montage-ops.md +127 -0
- package/dist/skills/opengame-montage-plan/references/og-snipers-record.md +110 -0
- package/dist/skills/opengame-montage-plan/references/owner-eye.md +214 -0
- package/dist/skills/opengame-montage-plan/references/reference-study.md +293 -0
- package/dist/skills/opengame-montage-plan/references/song.md +64 -0
- package/dist/skills/opengame-montage-review-and-improve/CHANGELOG.md +5 -0
- package/dist/skills/opengame-montage-review-and-improve/SKILL.md +178 -0
- package/dist/skills/opengame-montage-review-and-improve/references/briefs.md +293 -0
- package/dist/skills/opengame-montage-review-and-improve/references/day-one-questions.md +46 -0
- package/dist/skills/opengame-montage-review-and-improve/references/filmable.md +244 -0
- package/dist/skills/opengame-montage-review-and-improve/references/finishing.md +126 -0
- package/dist/skills/opengame-montage-review-and-improve/references/measure.md +282 -0
- package/dist/skills/opengame-montage-review-and-improve/references/montage-capture.md +151 -0
- package/dist/skills/opengame-montage-review-and-improve/references/montage-gates.md +84 -0
- package/dist/skills/opengame-montage-review-and-improve/references/montage-genres.md +198 -0
- package/dist/skills/opengame-montage-review-and-improve/references/montage-grammar.md +320 -0
- package/dist/skills/opengame-montage-review-and-improve/references/montage-ops.md +127 -0
- package/dist/skills/opengame-montage-review-and-improve/references/og-snipers-record.md +110 -0
- package/dist/skills/opengame-montage-review-and-improve/references/owner-eye.md +214 -0
- package/dist/skills/opengame-montage-review-and-improve/references/reference-study.md +293 -0
- package/dist/skills/opengame-montage-review-and-improve/references/song.md +64 -0
- package/dist/skills/opengame-multiplayer/CHANGELOG.md +15 -0
- package/dist/skills/opengame-multiplayer/SKILL.md +143 -0
- package/dist/skills/opengame-multiplayer/references/bot-lobbies.md +71 -0
- package/dist/skills/opengame-multiplayer/references/netcode.md +165 -0
- package/dist/skills/opengame-performance/CHANGELOG.md +8 -0
- package/dist/skills/opengame-performance/SKILL.md +106 -65
- package/dist/skills/opengame-performance/helpers/landing-pacer.mjs +86 -0
- package/dist/skills/opengame-performance/references/measuring.md +51 -0
- package/dist/skills/opengame-performance/references/webgpu.md +86 -0
- package/dist/skills/opengame-production/CHANGELOG.md +13 -0
- package/dist/skills/opengame-production/SKILL.md +119 -54
- package/dist/skills/opengame-production/references/team.md +107 -0
- package/dist/skills/opengame-production/references/templates.md +106 -16
- package/dist/skills/opengame-quality-bar/CHANGELOG.md +10 -0
- package/dist/skills/opengame-quality-bar/SKILL.md +94 -42
- package/dist/skills/opengame-quality-bar/references/cheap-tells.md +14 -7
- package/dist/skills/opengame-quality-bar/references/judges.md +95 -0
- package/dist/skills/opengame-quality-bar/references/review-prompt.md +10 -5
- package/dist/skills/opengame-rhythm/CHANGELOG.md +5 -0
- package/dist/skills/opengame-rhythm/SKILL.md +138 -0
- package/dist/skills/opengame-rhythm/references/charting.md +109 -0
- package/dist/skills/opengame-rhythm/references/timing.md +92 -0
- package/dist/skills/opengame-router/CHANGELOG.md +12 -0
- package/dist/skills/opengame-router/SKILL.md +14 -7
- package/dist/skills/opengame-scene-audit/CHANGELOG.md +15 -0
- package/dist/skills/opengame-scene-audit/SKILL.md +64 -47
- package/dist/skills/opengame-scene-audit/helpers/scene-audit.js +97 -19
- package/dist/skills/opengame-sdk/CHANGELOG.md +16 -0
- package/dist/skills/opengame-sdk/SKILL.md +27 -10
- package/dist/skills/opengame-sdk/references/engines.md +2 -2
- package/dist/skills/opengame-sdk/references/identity.md +10 -11
- package/dist/skills/opengame-sdk/references/leaderboards.md +11 -1
- package/dist/skills/opengame-sdk/references/radio.md +1 -1
- package/dist/skills/opengame-stability/CHANGELOG.md +4 -0
- package/dist/skills/opengame-stability/SKILL.md +59 -27
- package/dist/skills/opengame-trailer-make/CHANGELOG.md +5 -0
- package/dist/skills/opengame-trailer-make/SKILL.md +164 -0
- package/dist/skills/opengame-trailer-make/references/archetypes.md +113 -0
- package/dist/skills/opengame-trailer-make/references/briefs.md +273 -0
- package/dist/skills/opengame-trailer-make/references/core.md +129 -0
- package/dist/skills/opengame-trailer-make/references/craft.md +299 -0
- package/dist/skills/opengame-trailer-make/references/edge-cases.md +27 -0
- package/dist/skills/opengame-trailer-make/references/filmable.md +244 -0
- package/dist/skills/opengame-trailer-make/references/finishing.md +126 -0
- package/dist/skills/opengame-trailer-make/references/gates.md +184 -0
- package/dist/skills/opengame-trailer-make/references/measure.md +282 -0
- package/dist/skills/opengame-trailer-make/references/music.md +148 -0
- package/dist/skills/opengame-trailer-make/references/og-clash-record.md +133 -0
- package/dist/skills/opengame-trailer-make/references/ops.md +154 -0
- package/dist/skills/opengame-trailer-make/references/owner-eye.md +214 -0
- package/dist/skills/opengame-trailer-make/references/review.md +141 -0
- package/dist/skills/opengame-trailer-make/references/title.md +92 -0
- package/dist/skills/opengame-trailer-plan/CHANGELOG.md +5 -0
- package/dist/skills/opengame-trailer-plan/SKILL.md +176 -0
- package/dist/skills/opengame-trailer-plan/references/archetypes.md +113 -0
- package/dist/skills/opengame-trailer-plan/references/briefs.md +273 -0
- package/dist/skills/opengame-trailer-plan/references/core.md +129 -0
- package/dist/skills/opengame-trailer-plan/references/craft.md +299 -0
- package/dist/skills/opengame-trailer-plan/references/edge-cases.md +27 -0
- package/dist/skills/opengame-trailer-plan/references/filmable.md +244 -0
- package/dist/skills/opengame-trailer-plan/references/finishing.md +126 -0
- package/dist/skills/opengame-trailer-plan/references/gates.md +184 -0
- package/dist/skills/opengame-trailer-plan/references/measure.md +282 -0
- package/dist/skills/opengame-trailer-plan/references/music.md +148 -0
- package/dist/skills/opengame-trailer-plan/references/og-clash-record.md +133 -0
- package/dist/skills/opengame-trailer-plan/references/ops.md +154 -0
- package/dist/skills/opengame-trailer-plan/references/owner-eye.md +214 -0
- package/dist/skills/opengame-trailer-plan/references/reference-set.md +92 -0
- package/dist/skills/opengame-trailer-plan/references/review.md +141 -0
- package/dist/skills/opengame-trailer-plan/references/title.md +92 -0
- package/dist/skills/opengame-trailer-review-and-improve/CHANGELOG.md +5 -0
- package/dist/skills/opengame-trailer-review-and-improve/SKILL.md +145 -0
- package/dist/skills/opengame-trailer-review-and-improve/references/archetypes.md +113 -0
- package/dist/skills/opengame-trailer-review-and-improve/references/briefs.md +273 -0
- package/dist/skills/opengame-trailer-review-and-improve/references/core.md +129 -0
- package/dist/skills/opengame-trailer-review-and-improve/references/craft.md +299 -0
- package/dist/skills/opengame-trailer-review-and-improve/references/edge-cases.md +27 -0
- package/dist/skills/opengame-trailer-review-and-improve/references/filmable.md +244 -0
- package/dist/skills/opengame-trailer-review-and-improve/references/finishing.md +126 -0
- package/dist/skills/opengame-trailer-review-and-improve/references/gates.md +184 -0
- package/dist/skills/opengame-trailer-review-and-improve/references/measure.md +282 -0
- package/dist/skills/opengame-trailer-review-and-improve/references/music.md +148 -0
- package/dist/skills/opengame-trailer-review-and-improve/references/og-clash-record.md +133 -0
- package/dist/skills/opengame-trailer-review-and-improve/references/ops.md +154 -0
- package/dist/skills/opengame-trailer-review-and-improve/references/owner-eye.md +214 -0
- package/dist/skills/opengame-trailer-review-and-improve/references/review.md +141 -0
- package/dist/skills/opengame-trailer-review-and-improve/references/title.md +92 -0
- package/dist/skills/opengame-verification/CHANGELOG.md +11 -0
- package/dist/skills/opengame-verification/SKILL.md +113 -48
- package/dist/skills/opengame-weapons/CHANGELOG.md +11 -0
- package/dist/skills/opengame-weapons/SKILL.md +129 -55
- package/dist/skills/opengame-weapons/references/handling-numbers.md +42 -0
- package/dist/skills/opengame-weapons/references/reloads-and-hands.md +96 -0
- package/dist/skills/opengame-weapons/references/sights-and-scopes.md +98 -0
- package/dist/skills/opengame-weapons/references/threejs.md +71 -0
- package/dist/skills/opengame-world-physics/CHANGELOG.md +9 -0
- package/dist/skills/opengame-world-physics/SKILL.md +55 -24
- package/dist/skills/opengame-world-physics/references/measured-failures.md +28 -1
- package/package.json +1 -1
|
@@ -0,0 +1,126 @@
|
|
|
1
|
+
# Finishing: edit, grade, mix, render, QA, delivery
|
|
2
|
+
|
|
3
|
+
The OG Clash pipeline was ffmpeg and Python, engine-agnostic, and one command made a cut: shots and selects into an
|
|
4
|
+
edit list, a frame-exact conform, the grade, the mix, the render, then QA on the final file. Build yours the same way:
|
|
5
|
+
a script per stage, each restartable, and one command that runs them in order.
|
|
6
|
+
|
|
7
|
+
## 1. The edit
|
|
8
|
+
|
|
9
|
+
- **The edit list comes from the plan and the selects.** A select fixes a 1:1 mapping from take frame to timeline frame
|
|
10
|
+
(offset = timeline in − take in). The edit never changes speed: speed lives in the capture. A select may carry
|
|
11
|
+
segments (two takes in one slot).
|
|
12
|
+
- A slot with no select becomes a labelled slate ("SLATE, NOT FOR DELIVERY") so a watchable full-length cut exists from
|
|
13
|
+
hour one. A cut sent to the owner as a full cut never carries one.
|
|
14
|
+
- A fallback select lets a new slot show a neighbour's take until it is shot.
|
|
15
|
+
- Story beats are the only cuts allowed big luma jumps; list them.
|
|
16
|
+
- An overrides file holds per-shot exposure, temperature and tint trims, slips, the take-audio choice, gains, and the
|
|
17
|
+
title's frame range and stage.
|
|
18
|
+
- **Re-slotting:** move cuts 1-2 frames before beats or half-beats, keep payoffs inside their slots and covered by take
|
|
19
|
+
handles, write a reason per change, never more than 4 equal-length shots in a row.
|
|
20
|
+
- **A cue sheet keyed to payoffs:** sound cues reference the select's payoff frames (the nearest within 15 frames), so
|
|
21
|
+
sound follows re-shot picture. Re-key the cue sheet whenever selects or title strikes change.
|
|
22
|
+
|
|
23
|
+
## 2. Frame-exact conform
|
|
24
|
+
|
|
25
|
+
- Decode each event with its own ffmpeg, trimming by decoded frame index (`trim=start_frame=..:end_frame=..`), never by
|
|
26
|
+
time seeking; count the raw frames in your script and abort on any mismatch; feed one encoder at a constant 60 fps.
|
|
27
|
+
Self-test: sample frames must be bit-identical to the source.
|
|
28
|
+
- Frame-exact stills anywhere: `-vf "select=eq(n\,N)"`, never `-ss`.
|
|
29
|
+
- Plates for CG inserts: convert with `scale=in_color_matrix=bt709:in_range=tv:flags=accurate_rnd+full_chroma_int,format=rgb48be`
|
|
30
|
+
(a mixed extraction differed by up to 168 of 255).
|
|
31
|
+
|
|
32
|
+
## 3. Grade
|
|
33
|
+
|
|
34
|
+
- Per-act 3D LUTs (OG Clash built 65³ LUTs in Oklab on the engine's output) plus per-shot exposure, temperature and tint
|
|
35
|
+
trims. A colour script of one sentence per act.
|
|
36
|
+
- Light across cuts: median luma jump at or under 18, none over 40 except declared story beats; predict trims by
|
|
37
|
+
applying them to the two frames at the cut before rendering.
|
|
38
|
+
- No added grain, flares, chromatic aberration or vignette on top of game art; the game's own bloom stays.
|
|
39
|
+
- Titles and CG inserts go in **before** the grade; a studio mark, if the owner wants one, goes **after** it at encode.
|
|
40
|
+
- A montage in a named editor's style may have a finishing layer (levels, a curve, glow, punches); that is set by a look
|
|
41
|
+
test with the owner and runs once per cut (`opengame-montage-make` covers it).
|
|
42
|
+
|
|
43
|
+
## 4. Mix
|
|
44
|
+
|
|
45
|
+
- **The bed:** the music at one constant gain. Designed silences are edits of the track, not ducks.
|
|
46
|
+
- **The bus:** game sound, effects and voice peak-controlled as one bus keyed to the bed's headroom, big hits exempt.
|
|
47
|
+
Measure the music's level in the mix: a master limiter alone pulled OG Clash's music down 3.5-3.9 dB under every peak,
|
|
48
|
+
an accidental duck.
|
|
49
|
+
- **Game sound:** 0 to +6 LU over the bed in action (OG Clash needed +5 dB with per-shot trims).
|
|
50
|
+
- **Tails:** a 0.4 s raised-cosine continuation after each take group's last sound; slow-motion tails inside a group stop
|
|
51
|
+
5 ms before the next onset so no transient plays twice. No cut where the sound drops more than 20 dB within 2 frames.
|
|
52
|
+
- **Slow motion:** the real-time sound once, unstretched, anchored so the payoff sounds on its frame.
|
|
53
|
+
- **Never double a sound:** skip a designed cue where the take audio already carries the event (track each cue's source:
|
|
54
|
+
the take, the re-aligned take, a file, a generated sound). OG Clash's first round played 47 cues twice.
|
|
55
|
+
- **Designed hits:** on every big picture event, at least 35 % of the energy in 300 Hz-5 kHz, at least 6 dB over the
|
|
56
|
+
bed in that band, on the attack, in key, with a sub layer 14-21 dB under. Every flash gets a sound or is deliberately
|
|
57
|
+
silent.
|
|
58
|
+
- **Voice:** a static −3 dB presence dip on the bed around 2.2 kHz under each line, crossfaded over 8 frames; check each
|
|
59
|
+
line with a speech recogniser on the mixed master, four windows per line, three of four correct.
|
|
60
|
+
- **Electric effects:** a static −3 dB shelf at 9 kHz on your own layers, never on take audio.
|
|
61
|
+
- **Write every generated audio file atomically** (temporary name, then rename). A mix that read half-written files
|
|
62
|
+
produced a squashed master (−14.4 LUFS, 4.3 LU range).
|
|
63
|
+
- **Master:** −14 LUFS ±1 integrated and true peak at or under −1 dBTP on the AAC delivery; two limiters around a linear
|
|
64
|
+
gain with about 1 s of pre-roll; never ffmpeg's `loudnorm` filter, in one pass or two (its dynamic mode flattened drops by 3-4 dB).
|
|
65
|
+
The loudness range follows the track: document a waiver rather than leaving a gate red on every cut.
|
|
66
|
+
- A montage on the owner's song mixes differently (the song leads, untouched): `opengame-montage-make` covers it.
|
|
67
|
+
|
|
68
|
+
## 5. Render
|
|
69
|
+
|
|
70
|
+
- **Review cuts at 1080p at the master frame rate:** x264 `-preset slow -crf 18 -tune film` (about 17 Mbps; a fast
|
|
71
|
+
preset at CRF 15 was 29 % bigger). Under heavy load a fast preset is fine for review; the master path stays separate.
|
|
72
|
+
- **The master** at the plan's format, 1080p60 by default. A 4K master renders in 600-frame chunks, each reused when its
|
|
73
|
+
frame count matches (a later shot swap re-rendered one chunk in 344 s). Refuse to build a 4K master from 1080p
|
|
74
|
+
proxies.
|
|
75
|
+
- **Unique work folders and a lock** before cutting (process id plus seconds, never a minute stamp): two director
|
|
76
|
+
processes started in the same minute shared a work folder on OG Clash and produced a cut of 3,319 of 3,720 frames with
|
|
77
|
+
a frozen opening. Keep the work folder until QA passes on the final file.
|
|
78
|
+
- **Cap threads** (`ffmpeg -threads 6`) and send heavy jobs through your machine's job queue if it has one; an uncapped
|
|
79
|
+
encode at 880 % CPU froze the game team's work.
|
|
80
|
+
- **Long jobs survive session restarts** under a service manager (a user launchd agent on macOS, a systemd user unit on
|
|
81
|
+
Linux) rather than an agent's background shell. On macOS use a LaunchAgent with an interactive process type and
|
|
82
|
+
`launchctl bootstrap`; a job started with `launchctl submit` runs at background priority and crawls. On completion, stop it **and** delete its definition: OG Clash's jobs
|
|
83
|
+
set to run at load re-ran five times after delivery, rewriting the edit list and the mix.
|
|
84
|
+
- Copy from slow external drives into internal disk before encoding (a 4K encode reading from a spinning drive took 54
|
|
85
|
+
minutes against about 40 s locally), then copy the result out and verify it.
|
|
86
|
+
|
|
87
|
+
## 6. QA on every cut, on the final muxed file
|
|
88
|
+
|
|
89
|
+
| Check | Test |
|
|
90
|
+
|---|---|
|
|
91
|
+
| stream | frame count against the edit list; real and average rate exactly the film rate; BT.709 tags; audio and video within a frame |
|
|
92
|
+
| frames | black frames; frozen runs of 24 frames or more; exact duplicates; unique frames equal frames |
|
|
93
|
+
| light | the luma jump at each cut (the 160x90 grey mean of the cut frame against the frame two before it): median ≤ 18, none over 40 unless a story beat |
|
|
94
|
+
| loudness | −14 ±1 LUFS (a trailer), true peak ≤ −1 dBTP on the file's own AAC, loudness range (waiver documented) |
|
|
95
|
+
| silence | the quietest 25 ms RMS window stays above −26 dBFS, the first 150 ms and the final 0.8 s excluded (document a waiver for an intended quiet outro) |
|
|
96
|
+
| clicks | at every clip edge: no first-difference peak over 8x the local median within ±10 ms of the edge while the window is above −20 dBFS |
|
|
97
|
+
| moved cuts | strips at every cut that moved since the last cut; every changed shot's frame on a sheet |
|
|
98
|
+
|
|
99
|
+
The picture check inside OG Clash's render step passed on the broken cut with the frozen opening; only QA on the final
|
|
100
|
+
file caught it.
|
|
101
|
+
|
|
102
|
+
## 7. Delivery
|
|
103
|
+
|
|
104
|
+
The master format is chosen in the plan (default 1080p60). OG Clash was mastered at 4K60:
|
|
105
|
+
|
|
106
|
+
| Deliverable | Settings | OG Clash |
|
|
107
|
+
|---|---|---|
|
|
108
|
+
| 1080p60 review | libx264 slow CRF 18 tune film, High profile, yuv420p, GOP 120, BT.709, AAC 320k, `+faststart` | 134 MB, about 17 Mbps |
|
|
109
|
+
| 4K60 archive master | libx265 medium CRF 12, Main10, yuv420p10le, keyint 60, `hvc1` tag | 1.23 GB, about 158 Mbps |
|
|
110
|
+
| 4K60 delivery | hardware HEVC Main10 (on macOS `hevc_videotoolbox -profile:v main10 -pix_fmt p010le -b:v 90M -maxrate 140M -bufsize 180M -tag:v hvc1`) with BT.709 tags, or x265 CRF 16 with `-tag:v hvc1` | 700.6 MB, 90 Mbps |
|
|
111
|
+
| web | 1080p60 x264 CRF 20-23, max rate 7 Mbps; 720p CRF 26 for an archive page | 56 MB, 8-18 MB |
|
|
112
|
+
| inline chat preview | 720p, under 50 MB | |
|
|
113
|
+
| social | 30 s and 15 s in 9:16 with burned-in captions | |
|
|
114
|
+
|
|
115
|
+
**Measure "visually lossless" before claiming it:** PSNR and SSIM of the delivery against the master. Hardware HEVC
|
|
116
|
+
was about 100x faster than x265 under load. HEVC in an MP4 needs the `hvc1` tag or QuickTime, Safari and iOS will not
|
|
117
|
+
play it. Settle a target size or bitrate per deliverable in the plan ("optimize it
|
|
118
|
+
... so its not a crazy file size").
|
|
119
|
+
|
|
120
|
+
## 8. A studio mark (only if the owner wants one)
|
|
121
|
+
|
|
122
|
+
If the owner asks for a studio or publisher mark on every frame: draw it once at the master resolution and downsample
|
|
123
|
+
it, overlay it as a full-frame straight-alpha image at encode (after the grade, never graded), one flat white with no
|
|
124
|
+
shadow, small (OG Clash's avatar was about 3.5 % of frame height, with margins of about 1.9 % of the width and 3 % of the height), with a
|
|
125
|
+
switch that renders without it. A blind critic flagged the mark sitting over the title and the end card; offer to hide
|
|
126
|
+
it there.
|
|
@@ -0,0 +1,282 @@
|
|
|
1
|
+
# Measuring a cut: the recipes behind every gate
|
|
2
|
+
|
|
3
|
+
Every gate in these skills is a number you can get from the rendered file, the take's own capture log, or the
|
|
4
|
+
audio. None of it needs a special tool: ffmpeg, ffprobe, Python with numpy and scipy, OpenCV for optical flow, and
|
|
5
|
+
librosa for beats and onsets cover all of it. Install them in a private virtual environment for the film. Never play
|
|
6
|
+
audio aloud to "check" it: an agent measures sound, it does not listen.
|
|
7
|
+
|
|
8
|
+
Rules that apply to every recipe:
|
|
9
|
+
|
|
10
|
+
- Measure the full-resolution review render (or the delivered file), never a proxy or an older cut. A gate passed on
|
|
11
|
+
the wrong file is a gate that did not run.
|
|
12
|
+
- Frame numbers are 0-based at the film's rate. At 60 fps, seconds = frame / 60.
|
|
13
|
+
- Grab exact frames by index with `select=eq(n\,N)`, never by `-ss` seeking, which lands on the wrong frame.
|
|
14
|
+
- Prefer the capture's own logs (camera, speed, probe) over picture analysis when you have them. A picture measure
|
|
15
|
+
guesses at what the log already knows. Use picture measures for files you did not capture (references, a closed
|
|
16
|
+
game, a critic's review).
|
|
17
|
+
- Run heavy decodes politely: `nice -n 10`, `ffmpeg -threads 6`, through your machine's job queue if it has one.
|
|
18
|
+
|
|
19
|
+
## 1. Stream facts, frames and stills
|
|
20
|
+
|
|
21
|
+
```bash
|
|
22
|
+
# size, rate, frame count, colour tags, duration of each stream
|
|
23
|
+
ffprobe -v error -count_frames -show_entries \
|
|
24
|
+
stream=index,codec_type,width,height,r_frame_rate,avg_frame_rate,nb_read_frames,pix_fmt,color_primaries,color_transfer,color_space,duration \
|
|
25
|
+
-of json cut.mp4
|
|
26
|
+
|
|
27
|
+
# one exact frame, and a strip of consecutive frames (8 per row) for motion checks
|
|
28
|
+
ffmpeg -v error -i cut.mp4 -vf "select=eq(n\,1234)" -vframes 1 f1234.png
|
|
29
|
+
ffmpeg -v error -i cut.mp4 -vf "select=between(n\,1200\,1263),scale=480:-1,tile=8x8" -vframes 1 strip_1200.jpg
|
|
30
|
+
```
|
|
31
|
+
|
|
32
|
+
- Expected: the planned size, `r_frame_rate` and `avg_frame_rate` both exactly the film rate (60/1), BT.709 primaries,
|
|
33
|
+
transfer and matrix, and audio and video durations within one frame.
|
|
34
|
+
- If a re-encode leaves the colour tags `unknown`, write them into the stream: for H.264,
|
|
35
|
+
`-bsf:v h264_metadata=colour_primaries=1:transfer_characteristics=1:matrix_coefficients=1`.
|
|
36
|
+
- **Every frame, never every second frame,** for any shot with a body, a fall, a hero move or a slow-motion span.
|
|
37
|
+
An every-2nd-frame sheet hides a head flipping between two poses on alternate frames; on the OG Snipers montage that
|
|
38
|
+
defect passed the director and a gate and was found after the cut was called final.
|
|
39
|
+
|
|
40
|
+
## 2. Frozen, duplicate, black and unique frames
|
|
41
|
+
|
|
42
|
+
```bash
|
|
43
|
+
ffmpeg -v error -i cut.mp4 -f framemd5 - > frames.md5 # one hash per frame
|
|
44
|
+
ffmpeg -v info -i cut.mp4 -vf freezedetect=n=0.001:d=0.4 -f null - 2>&1 | grep freeze_
|
|
45
|
+
```
|
|
46
|
+
|
|
47
|
+
- Unique frames: count distinct hashes in `frames.md5`. For footage you captured frame by frame, unique frames must
|
|
48
|
+
equal frames. A frame-count check alone passed an OG Clash cut whose first 402 frames were one frozen picture.
|
|
49
|
+
- Duplicate share in slow motion: identical consecutive hashes divided by frames in the slow span. Real simulated
|
|
50
|
+
slow motion has none; cheap interpolated or frame-doubled edits measured 8-10 %.
|
|
51
|
+
- Black: decode to 160x90 grey and call a frame black when 98 % of its pixels sit below luma code 26. A run outside a
|
|
52
|
+
designed black slug fails. A film's own "black" check that only looks for fully black frames missed a set that read
|
|
53
|
+
as black on a phone.
|
|
54
|
+
|
|
55
|
+
```python
|
|
56
|
+
import subprocess, numpy as np
|
|
57
|
+
w, h = 160, 90
|
|
58
|
+
raw = subprocess.run(["ffmpeg", "-v", "error", "-i", "cut.mp4", "-vf", f"scale={w}:{h},format=gray",
|
|
59
|
+
"-f", "rawvideo", "-"], capture_output=True, check=True).stdout
|
|
60
|
+
frames = np.frombuffer(raw, np.uint8).reshape(-1, h, w)
|
|
61
|
+
black = [i for i, f in enumerate(frames) if (f < 26).mean() >= 0.98]
|
|
62
|
+
```
|
|
63
|
+
|
|
64
|
+
## 3. Cuts and shot lengths
|
|
65
|
+
|
|
66
|
+
- If you built the cut from an edit list, take the cut frames from it. That is the truth.
|
|
67
|
+
- For a file you did not cut, detect cuts adaptively: per frame, the mean absolute difference of a 160x90 grey frame
|
|
68
|
+
against the previous one, compared with a rolling median of the last 30 differences; a cut is a difference over
|
|
69
|
+
about 4x the median that does not return to the previous picture within 2 frames (that return is a flash or a strobe,
|
|
70
|
+
not a cut). ffmpeg's `select='gt(scene,0.3)'` alone misses 30-70 % of cuts on dark footage.
|
|
71
|
+
- Report: cut count, cuts per second, mean and median shot length, share of shots under 0.5 s and under 1 s, the
|
|
72
|
+
longest run of equal-length shots (in beats), and cuts per 10 s window across the film.
|
|
73
|
+
|
|
74
|
+
## 4. Motion: camera speed, seams, ramps
|
|
75
|
+
|
|
76
|
+
Camera speed is apparent motion in pixels per frame at 1080p. From a capture log, project the camera's motion. From
|
|
77
|
+
pictures, use dense optical flow:
|
|
78
|
+
|
|
79
|
+
```python
|
|
80
|
+
import cv2, numpy as np, json, sys
|
|
81
|
+
cap = cv2.VideoCapture(sys.argv[1]); prev = None; rows = []
|
|
82
|
+
while True:
|
|
83
|
+
ok, frame = cap.read()
|
|
84
|
+
if not ok: break
|
|
85
|
+
g = cv2.cvtColor(cv2.resize(frame, (480, 270)), cv2.COLOR_BGR2GRAY)
|
|
86
|
+
if prev is not None:
|
|
87
|
+
flow = cv2.calcOpticalFlowFarneback(prev, g, None, 0.5, 3, 15, 3, 5, 1.2, 0)
|
|
88
|
+
mag = np.linalg.norm(flow, axis=2) * 4 # 480 wide -> 1080p pixels
|
|
89
|
+
ang = np.degrees(np.arctan2(np.median(flow[..., 1]), np.median(flow[..., 0])))
|
|
90
|
+
rows.append({"px": float(np.median(mag)), "p90": float(np.percentile(mag, 90)), "deg": float(ang)})
|
|
91
|
+
prev = g
|
|
92
|
+
json.dump(rows, open("flow.json", "w"))
|
|
93
|
+
```
|
|
94
|
+
|
|
95
|
+
- Every frame of every shot moves ≥ 1 px/f unless the plan scopes a hold that pays off within 0.3 s.
|
|
96
|
+
- Seams: the last 3 frames out and the first 4 frames in. Measure the first frame after the cut too: a rig that starts
|
|
97
|
+
its camera on output frame 0 makes every first frame a still, and critics read those as stutters.
|
|
98
|
+
- Direction (a trailer): compare the median flow angle on each side; fail a seam slower than 0.6 px/f on either side or
|
|
99
|
+
more than 90° apart unless the plan states a cause. A montage carries speed across the cut and lets direction differ
|
|
100
|
+
(FAINT's seams sit a median 115° apart); there, fail only a still-to-still seam.
|
|
101
|
+
- Speed at a montage cut: over the last 3 frames out and the first 3 in, fail when the faster side is more than 30 %
|
|
102
|
+
above the slower one (or the directions part by more than 30°) while both move over 8 px/f; a cut between two still
|
|
103
|
+
pictures needs no match. Report whether each side's cut frame sits within 2 frames of its own speed peak over ±12
|
|
104
|
+
frames.
|
|
105
|
+
- Constant velocity: a third-person shot whose smoothed flow has a coefficient of variation under 0.15 is a robotic
|
|
106
|
+
dolly. No more than a quarter of them.
|
|
107
|
+
- Flow around hero cuts: frames −1 and +1 each at least 1.4x the mean of frames −8 to −5.
|
|
108
|
+
- Beat-phase motion swing: bin every frame's flow by its phase within the beat (20 bins); swing = the largest bin over
|
|
109
|
+
the smallest. Kaiser's montage edits measured 1.6-3.3x with the peak on the beat; a capture reel cut to a click track
|
|
110
|
+
measured 1.37x with the peak mid-beat.
|
|
111
|
+
- **Zoom blindness:** a phase-correlation motion measure on a small decode reads in coarse steps and cannot see a
|
|
112
|
+
zoom. Take camera speed from the capture's camera log before calling a shot dead or a seam regressed.
|
|
113
|
+
|
|
114
|
+
Ramps:
|
|
115
|
+
|
|
116
|
+
- From the capture log (best): the simulation speed per output frame. Report the slowest speed, where it sits in the
|
|
117
|
+
shot, the largest one-frame change as a ratio, |s(t+1) / s(t) − 1| (fail over 15 % on a cinematic pair of frames),
|
|
118
|
+
and any run of 20 or more cinematic frames whose speed never moves (maximum over minimum − 1 under 2 %), which fails
|
|
119
|
+
when it follows a payoff.
|
|
120
|
+
- From pictures: a ramp is legible when the flow inside one shot changes by 3x or more and the content does not explain
|
|
121
|
+
it (the camera did not simply stop). On the OG Clash trailer six shots had real quarter-speed spans and a blind
|
|
122
|
+
critic's flow analysis still found zero ramps, because the camera moved through them at one speed.
|
|
123
|
+
- Slow spans: camera median ≥ 2 px/f, never under 1. Character angular rate (from the probe: the angle each joint chain
|
|
124
|
+
turns per output frame) on no frame under 40 % of the span's median. Under that, the viewer sees a freeze, usually a
|
|
125
|
+
hitstop timer counted on raw time.
|
|
126
|
+
|
|
127
|
+
## 5. Sharpness, light, colour, flashes
|
|
128
|
+
|
|
129
|
+
```bash
|
|
130
|
+
ffmpeg -v error -i cut.mp4 -vf "signalstats,metadata=print:file=stats.txt" -f null - # YAVG, YMIN, YMAX, SATAVG per frame
|
|
131
|
+
ffmpeg -v error -i cut.mp4 -vf "blurdetect,metadata=print:file=blur.txt" -f null - # per-frame blur (FFmpeg 5.1 or newer)
|
|
132
|
+
```
|
|
133
|
+
|
|
134
|
+
- Luma at cuts: decode to 160x90 grey as in section 2 and take the absolute difference of the mean of the cut frame c
|
|
135
|
+
and the frame c − 2. Median ≤ 18, none over 40 except cuts the plan declares as story beats. (`signalstats` YAVG is
|
|
136
|
+
in limited range; if you use it, scale the thresholds to about 15.5 and 34.) Predict a fix as an exposure trim on the
|
|
137
|
+
incoming shot. A montage whose approved look alternates dark gameplay and bright cinematics (FAINT's median jump is
|
|
138
|
+
37, with 29 of 68 cuts over 40) waives the gate at gameplay-to-cinematic cuts.
|
|
139
|
+
- Sharp subject: every shot's blur within the film's own norm (median plus a margin you set from your sharpest shots).
|
|
140
|
+
A frame with nothing sharp, or a defocused foreground over 10 % of the frame, fails. A Laplacian variance per shot
|
|
141
|
+
(`cv2.Laplacian(gray, cv2.CV_64F).var()`) is the fallback.
|
|
142
|
+
- Flat white on effects: decode to RGB and count pixels whose R, G and B are all at or above 225 in limited-range code
|
|
143
|
+
values. ffmpeg's usual conversion to full-range RGB moves 225 to about 244, so threshold the converted frame at 244.
|
|
144
|
+
A hero effect held on screen stays at or under 0.5 % of the frame; measure held effect frames, not the single impact
|
|
145
|
+
flash frame.
|
|
146
|
+
|
|
147
|
+
```python
|
|
148
|
+
img = cv2.cvtColor(cv2.imread("f1234.png"), cv2.COLOR_BGR2RGB)
|
|
149
|
+
flat_white = (img >= 244).all(axis=2).mean() * 100 # percent of frame
|
|
150
|
+
```
|
|
151
|
+
|
|
152
|
+
- Grade: content luma Y 40-70 (limited range) for a dark genre, clipping under 3 %. Per act, the mean chroma (U, V)
|
|
153
|
+
vector should move by 4 units or more between acts when the plan has a colour script.
|
|
154
|
+
- Flashes: a luma rise of 12 or more over 2 frames away from a cut. Every one gets a designed sound or is deliberately
|
|
155
|
+
silent. Photosensitivity: no more than 3 general or red flashes in any 1 s (WCAG 2.3.1 counts a flash over more than 25 % of
|
|
156
|
+
any 10° field of view, a much smaller area than a quarter of the frame); run an open-source analyser such as EA's IRIS
|
|
157
|
+
on every cut sent to anyone.
|
|
158
|
+
- Banding: look at a 4x-contrast crop of every dark gradient (night sky, fog). If it bands, raise the capture bitrate
|
|
159
|
+
(HEVC at 300 Mbps banded least on the OG Clash storm sky) before touching the grade.
|
|
160
|
+
|
|
161
|
+
## 6. Loudness, holes, the loudest moment, tails
|
|
162
|
+
|
|
163
|
+
```bash
|
|
164
|
+
ffmpeg -nostats -i cut.mp4 -filter_complex ebur128=peak=true -f null - 2>&1 | tail -15 # I, LRA, true peak
|
|
165
|
+
ffmpeg -v error -i cut.mp4 -af "ebur128=metadata=1,ametadata=mode=print:key=lavfi.r128.S:file=short.txt" -f null -
|
|
166
|
+
ffmpeg -v error -i cut.mp4 -ac 1 -ar 48000 -f f32le mono.f32 # samples for numpy
|
|
167
|
+
```
|
|
168
|
+
|
|
169
|
+
- Integrated loudness and true peak on the delivered AAC, not on the WAV you mixed: the encode moves the peak.
|
|
170
|
+
- Short-term spread across the body (a trailer): at least 8 LU from the quietest to the loudest section. The OG Clash
|
|
171
|
+
trailer held 2.3 LU for 32 s and read as a flat wall.
|
|
172
|
+
- RMS in 50 ms windows: the loudest window's time is the loudest moment. In a trailer it belongs on the boss impact,
|
|
173
|
+
an explosion or the title, never on a mid-film feature. A montage on the owner's song has no slam: there the check is
|
|
174
|
+
that nothing but a payoff outranks the payoffs (section 8).
|
|
175
|
+
- A designed hole (trailers only): at least 20 dB under the previous 2 s, 0.3-2.4 s long, followed within 0.25 s by a hit at least
|
|
176
|
+
15 dB up. Inside it, no 25 ms window under −90 dBFS (digital zero reads as a dropout), and the 10 frames after its cut at
|
|
177
|
+
least 20 dB under the 10 frames before.
|
|
178
|
+
- Tails: at every cut, the RMS of the 2 frames after against the 2 frames before; a drop over 20 dB is a sound cut dead
|
|
179
|
+
unless it is a designed silence.
|
|
180
|
+
- Clicks: at every clip edge, fail a first-difference peak of the samples over 8x the local median within ±10 ms of the
|
|
181
|
+
edge while the window is above −20 dBFS.
|
|
182
|
+
- Silence: the quietest 25 ms RMS window of the film stays above −26 dBFS, excluding the first 150 ms and the final
|
|
183
|
+
0.8 s; write a waiver for an intended quiet outro.
|
|
184
|
+
|
|
185
|
+
## 7. Beats, onsets and sync
|
|
186
|
+
|
|
187
|
+
```python
|
|
188
|
+
import librosa, numpy as np
|
|
189
|
+
y, sr = librosa.load("track.wav", sr=48000, mono=True)
|
|
190
|
+
onsets = librosa.onset.onset_detect(y=y, sr=sr, units="time")
|
|
191
|
+
tempo, beats = librosa.beat.beat_track(y=y, sr=sr, units="time")
|
|
192
|
+
```
|
|
193
|
+
|
|
194
|
+
- Never trust one beat tracker. Refine the tempo from attacks in one band (a hi-hat or snare band), fit beat index
|
|
195
|
+
against attack time by least squares, and check half and double tempo with onset autocorrelation. On the OG Clash
|
|
196
|
+
track 32 high-band attacks gave 105.0001 BPM with a 0.08 ms residual, where two trackers said 105.04 and 105.47.
|
|
197
|
+
- Click check: mix 2 kHz clicks on downbeats and 1 kHz on beats at −6 dB over the track, then measure the track's own
|
|
198
|
+
attacks within ±15 ms of each grid beat. Pass when every downbeat you cut on sits within one frame.
|
|
199
|
+
- Cuts married to sound: the share of cuts within ±50 ms of an onset of the final mix, against a null made by shifting
|
|
200
|
+
the cut list by random offsets 1,000 times. Aim for 30 % or more and at least 2x the null on the hero section.
|
|
201
|
+
- Performed on the beat: the share of action peaks (muzzle flash, impact, jump apex, from the probe or the luma) within
|
|
202
|
+
±2 frames of a beat. Cuts on the beat are not this.
|
|
203
|
+
- Beat lock of a reference: convert each cut time to a phase angle within its beat and run a Rayleigh test (the mean
|
|
204
|
+
resultant length R and its p-value); most official trailers do not lock (p from 0.93 down to 0.003 across the
|
|
205
|
+
measured set), Kaiser's edits do. Also report the share of cuts in the last eighth of a beat.
|
|
206
|
+
- Best audio/video shift: slide the cut list against the onset list in whole frames from −6 to +6 and keep the shift
|
|
207
|
+
with the most cuts within a frame of an onset; Kaiser's edits peak at the picture 1-2 frames early.
|
|
208
|
+
- A sound's time is its audible attack: the rise into the peak of its 200 Hz-12 kHz envelope, not its broadband peak.
|
|
209
|
+
Dull reports placed by peak started 4-10 frames early on the OG Snipers montage.
|
|
210
|
+
- Slow-motion hits: the flash's visual peak can land many output frames after its engine event (about 10 frames at
|
|
211
|
+
0.3x). Measure the flash frame on the picture and put the sound there.
|
|
212
|
+
|
|
213
|
+
## 8. Sound as heard
|
|
214
|
+
|
|
215
|
+
- Payoff level as heard: band-pass the delivered mix (or the game bus, then the mix) to 150 Hz-8 kHz and take the RMS
|
|
216
|
+
over 300 ms from each payoff's attack. Table every payoff with the film's median and spread. A gate that measured
|
|
217
|
+
each gunshot asset at a fixed level said "spread 0.1 dB" while the hits as heard spread 6-15 dB, because impacts,
|
|
218
|
+
confirms and medals piled onto some shots and not others.
|
|
219
|
+
- Rank every transient in the film; nothing that is not a payoff may be louder than the loudest payoff.
|
|
220
|
+
- Designed hits: at least 35 % of their energy in 300 Hz-5 kHz (sub-only hits are silent on a phone) and at least
|
|
221
|
+
6 dB over the music in that band over their first 50 ms, for a trailer. A montage on the owner's song sits payoffs
|
|
222
|
+
about 3 dB over the song in 1-5 kHz instead.
|
|
223
|
+
- Music level in the mix (accidental ducks): fit the music stem to the mix by least squares in 100 ms windows. The
|
|
224
|
+
fitted gain should stay flat within 1.5 dB under voice and effects; a dip is a limiter pulling the bed down.
|
|
225
|
+
- Song untouched (a montage): align the song file to the music bus by cross-correlation and fit one gain over 1 s
|
|
226
|
+
windows across the film. The gain must hold within 0.5 dB everywhere except an agreed fade at the end; a jump is an
|
|
227
|
+
edit or a duck.
|
|
228
|
+
- Intelligibility: run an open-source speech recogniser (Whisper or faster-whisper) on the mixed master, four
|
|
229
|
+
windows per line; at least three of four must read the line correctly. Test callouts over the real song at several
|
|
230
|
+
offsets with two different recognisers before mixing; treat the result as a screen, and the owner's ear as the
|
|
231
|
+
decision.
|
|
232
|
+
|
|
233
|
+
## 9. Characters, subjects and recurring elements
|
|
234
|
+
|
|
235
|
+
- Subject in frame: project the probe's subject point (or box) through the logged camera on every frame. A follow
|
|
236
|
+
camera keeps the hero's centre inside 12-85 % of frame height and 5-95 % of width; a drift keeps at least 60 % of the
|
|
237
|
+
subject's box in frame, with no run of 30 frames or more that is 70 % sky or empty ground.
|
|
238
|
+
- Gap to a title or insert: project the insert's box and a conservative 3D box around every character and prop on
|
|
239
|
+
every frame; the smallest pixel gap must stay above 0.
|
|
240
|
+
- Pose pops: on head, hand and weapon crops, flag frame n when it differs strongly from both n−1 and n+1 while n−1 and
|
|
241
|
+
n+1 match each other (an A-B-A flip). Run it on every body in every slow span, not only on the hips.
|
|
242
|
+
- Surface flicker (z-fighting): warp both neighbours onto each frame with optical flow and flag 32x32 blocks that
|
|
243
|
+
differ from both warped neighbours while those agree. Leave out whip frames over 40 px/f, 2 frames either side of a
|
|
244
|
+
cut, HUD zones and close occluders. Scan only the changed spans (diff the frozen edit list against the last checked
|
|
245
|
+
one frame by frame and list the ranges whose take or source frame changed), print the verdict before drawing strips,
|
|
246
|
+
keep a cleared list keyed by take, frames and box, and calibrate the threshold on a frame the owner flagged; an
|
|
247
|
+
uncalibrated scan ranked the one real defect 658th under camera motion.
|
|
248
|
+
- Hero jitter: project the hero's position through the camera on every frame, fit a quadratic in game time over 9
|
|
249
|
+
frames, and take the RMS of the residual; pass at 1.5-2.0 px. On OG Snipers it read 51-68 px before the draw-sync hook
|
|
250
|
+
and 0.2-0.4 px after.
|
|
251
|
+
- Cross-clip sheets: crop the same region (kill feed, hit marker, score bug, position badge, the hero's weapon, the
|
|
252
|
+
title) from every clip onto one sheet and find each element's anchor by template matching. Report the anchors in
|
|
253
|
+
pixels. One such sheet found in minutes a feed-spacing defect that fifteen deliveries had missed.
|
|
254
|
+
- Bot tells: compare the scripted player's view log with 3-5 minutes of a good human playing the same game: time to
|
|
255
|
+
aim, flick size and duration, overshoot, how often the view sits dead still, hit rate. A screen, not a verdict.
|
|
256
|
+
|
|
257
|
+
## 10. Titles and inserts
|
|
258
|
+
|
|
259
|
+
- Stillness at rest: render the title shot with the camera locked and its energy off; every frame's hash must match.
|
|
260
|
+
- Camera match: reproject markers placed on logged joints through your rebuilt camera; mean error ≤ 0.1 px, maximum
|
|
261
|
+
≤ 0.25 px, and a background feature track ≤ 0.25 px RMS with no drift.
|
|
262
|
+
- Drift of the inserted object ≤ 0.3 px/f against its anchor; crispness (Laplacian variance) at least 0.9 of its hold
|
|
263
|
+
median; size at rest as the plan says.
|
|
264
|
+
- Words: crop every word the plan puts on screen from the full-resolution render and read it. The scene graph is not
|
|
265
|
+
evidence: a call to action once faced away from the camera for a whole round while the handoff called it built.
|
|
266
|
+
|
|
267
|
+
## 11. Delivery checks
|
|
268
|
+
|
|
269
|
+
```bash
|
|
270
|
+
ffmpeg -i delivery.mp4 -i master.mov -lavfi "[0:v][1:v]psnr" -f null - # PSNR against the master
|
|
271
|
+
ffmpeg -i delivery.mp4 -i master.mov -lavfi "[0:v][1:v]ssim" -f null - # SSIM against the master
|
|
272
|
+
```
|
|
273
|
+
|
|
274
|
+
- Measure PSNR and SSIM before calling a compressed delivery "visually lossless". Both inputs must have the same size
|
|
275
|
+
and frame count: scale one first (`[1:v]scale=1920:1080[m];[0:v][m]psnr`) when the delivery is smaller than the
|
|
276
|
+
master.
|
|
277
|
+
- A mix belongs to one picture: hash the edit list (every slot's take, in and out, source in and payoff frames) into a
|
|
278
|
+
signature, store it with the mix, and mux only when they match. A stale mix muxed onto new picture put two clips of
|
|
279
|
+
the OG Snipers montage 21-27 frames off.
|
|
280
|
+
- Copying to the owner's drive: write to a plain temporary name (`cut.mp4.part-12345`, never a name starting with a
|
|
281
|
+
dot, which some network mounts refuse), rename, then read the file back and compare size and checksum. Report "not
|
|
282
|
+
delivered" until the read-back matches.
|
|
@@ -0,0 +1,151 @@
|
|
|
1
|
+
# Capturing a montage: tiers, film-build hooks, and shooting without source
|
|
2
|
+
|
|
3
|
+
Check the route with the capability checklist C1-C19 in the filmable chapter first. Then pick a tier, and say its limits
|
|
4
|
+
in the plan before anything is shot.
|
|
5
|
+
|
|
6
|
+
## 1. The tiers
|
|
7
|
+
|
|
8
|
+
| Tier | The game has | Shoot with |
|
|
9
|
+
|---|---|---|
|
|
10
|
+
| A | source or a film build, a stepped clock, a free camera | the default stepped capture (the filmable chapter); OG Snipers |
|
|
11
|
+
| B | deterministic input replay | record a human live, re-run the input in A's rig, verify per-frame state hashes |
|
|
12
|
+
| C | a replay or demo with a free camera, no source | play live, shoot each moment in passes (as played, outside, the receiver's side), render offline if possible |
|
|
13
|
+
| D | a replay without a free camera, or photo mode only | the game's own cameras plus live capture at a high frame rate |
|
|
14
|
+
| E | live play only | high-frame-rate lossless capture; a slow-motion budget of capture fps ÷ 60; events mined from the feed; no second angle |
|
|
15
|
+
| F | no GPU, or a software renderer only | stop and say so |
|
|
16
|
+
|
|
17
|
+
- **Check replay slow motion on frames before planning ramps on it.** Most replays store snapshots and interpolate
|
|
18
|
+
between them: gliding limbs and missing particles at 0.1x. If it interpolates, use it only down to 0.5x. Version-lock
|
|
19
|
+
replays (some stop loading after a game patch): capture each the day it is recorded, or freeze the build.
|
|
20
|
+
- **Decide the master resolution before the first select.** Every OG Snipers take was captured at 1080p, so its 4K60
|
|
21
|
+
delivery was an upscale; the seeded 4K re-shoot route was proven on day one and never used.
|
|
22
|
+
- **Tiers B-E have no film build,** so staging, seeds and probes become recorded human play, replay passes and event
|
|
23
|
+
mining. At C and D slow motion is only as fine as the replay re-simulates (interpolated replays to 0.5x); at D and E it
|
|
24
|
+
is capped at capture fps ÷ 60; at E there is no second angle, so plan a lower cinematic share.
|
|
25
|
+
|
|
26
|
+
## 2. Film-build hooks (tier A)
|
|
27
|
+
|
|
28
|
+
`opengame-capture` lists the in-game rules for each hook; the filmable chapter has the general catalogue. What a
|
|
29
|
+
montage needs, from OG Snipers:
|
|
30
|
+
|
|
31
|
+
- **The minimum:** a frozen build with byte-identical parity (every switch off equals shipped); a per-frame command bridge
|
|
32
|
+
with a far keepalive (the engine quit 10 s after its last command); a stepped clock with any step and a slow floor of
|
|
33
|
+
0.05x or lower; HUD per element and event filters (a hero-only feed); a stage camera with a cut that resets temporal
|
|
34
|
+
history, and attach spaces; a per-frame probe of the player, the camera, every actor, every payoff event (receiver,
|
|
35
|
+
cause, zone) and audio cues; **a deterministic seed over every random generator** (the hook that paid off most:
|
|
36
|
+
inserts, re-shoots and two-pass lenses all depend on it); opponent staging (place, act, path, loadout, name, health or
|
|
37
|
+
pace); a renderer check that refuses software rendering.
|
|
38
|
+
- **What paid off:** drawing every body in sync every frame under slow motion (the hero's residual jitter went from 32-60
|
|
39
|
+
px to 0.2-0.4 px); an aim trim inside the game's rules; death and ragdoll controls (speed, variety, no pops, dropped
|
|
40
|
+
props as objects); **drawn body points and a camera look-at** (they came on day 3 and should come on day 1, after two
|
|
41
|
+
guessed framings failed); penetration and scene-ray queries; overlay switches; a hero avatar swap for outside shots;
|
|
42
|
+
payoff events on their own frame; prop landing events; a shutter in game time.
|
|
43
|
+
- **What did not pay:** an outside kill-cam hook (written, never run); a film-only damage rule without the owner's yes
|
|
44
|
+
(built, never used); a small hand prop for the ending (it took builds 4 to 45 before it read on camera); a backdrop fix that did not cure a
|
|
45
|
+
flicker (the clip moved to another map instead).
|
|
46
|
+
- **Traps:** a field added to a struct that is serialised into a cooked scene made the loader drop it, and two builds
|
|
47
|
+
never started a match (put film switches on run-time structs); staging switches acted on a change of value, so send a
|
|
48
|
+
reset 6 frames before; one replay per page load; a camera that follows the probe lags a frame (use attach spaces); the
|
|
49
|
+
world replays under a seed but a closed-loop scripted hand does not (re-shoots replay the recorded input open loop).
|
|
50
|
+
- **On day one, also expose:** the drawn points of whatever the engine animates worst (a skeleton, wheels, board pieces,
|
|
51
|
+
units), the payoff point, the line of the attack, ground height and a camera look-at.
|
|
52
|
+
- **Probe translations by genre:** racing has poses, wheel contacts, collisions, position changes and laps; fighting has
|
|
53
|
+
hitboxes, hit events with damage and frame advantage, and the KO frame; sports has the ball, players and score events;
|
|
54
|
+
strategy has units, deaths and buildings; rhythm has judgements and combo; VR has head and hand poses, grab and hit
|
|
55
|
+
events.
|
|
56
|
+
- **Reaction controls by genre:** ragdoll for shooters, crash for racing, KO for fighting, and a hitstop multiplier
|
|
57
|
+
wherever hitstop exists.
|
|
58
|
+
- **One shared clock library** with a unit test against the engine's own time: a clock-sum off-by-one between the payoff
|
|
59
|
+
solver and the engine failed one take three times.
|
|
60
|
+
|
|
61
|
+
## 3. Shooting without source
|
|
62
|
+
|
|
63
|
+
### A replay or demo system (tiers C and D)
|
|
64
|
+
|
|
65
|
+
1. **Play live, shoot later.** Play, or have players play, normally (real human feel, no bot tells), recording every
|
|
66
|
+
replay. Index events from the replay data, or from a pass that reads the kill feed or score pops.
|
|
67
|
+
2. **Shoot each moment as several passes:** the as-played view, then outside cameras (the clip's cinematic), then the
|
|
68
|
+
receiver's side. That is "show it twice" for free.
|
|
69
|
+
3. **Check replay fidelity on frames before planning.** Replays are often low-rate state snapshots; checkpoints can hold
|
|
70
|
+
actors from different frames. Look for gliding limbs, missing particles, frozen cloth, popping ragdolls and missing
|
|
71
|
+
sound.
|
|
72
|
+
4. **Check replay slow motion.** If it interpolates, use it only to 0.5x. If the replay re-simulates from inputs, slow
|
|
73
|
+
motion can be real. Measure the duplicate-frame share and ragdoll stepping on a test clip.
|
|
74
|
+
5. **Render offline if you can** (some replay tools render through ffmpeg; Unreal can re-render a replay through Movie
|
|
75
|
+
Render Queue). If the replay only plays in real time, record at the highest stable frame rate, play at 0.25x and
|
|
76
|
+
record, then conform.
|
|
77
|
+
6. **Version-lock replays,** and capture each the day it is recorded.
|
|
78
|
+
7. **Design the speed curve before rendering,** then render the replay through it so slow spans have real frames.
|
|
79
|
+
|
|
80
|
+
### Live play only (tier E)
|
|
81
|
+
|
|
82
|
+
1. Record everything, lossless or near-lossless, at the highest constant frame rate the machine holds; game audio on its
|
|
83
|
+
own track, in-game music muted.
|
|
84
|
+
2. The slow-motion budget is capture fps ÷ 60: at 240 fps, 0.25x is clean. Below that, optical-flow retiming is the last
|
|
85
|
+
resort; say so to the owner and ask.
|
|
86
|
+
3. No second angle exists. The show-it-twice layer has to come from staged private matches with a spectator, a photo
|
|
87
|
+
mode, the kill cam, or a separately built cinematic. If none exists, lean on transitions caused by the HUD and the
|
|
88
|
+
weapon (scope irises, tint flashes), and accept a lower cinematic share. Report it as a known limit.
|
|
89
|
+
4. Mine footage at scale with an event detector on the kill feed or score pops (template matching), then pick by hand.
|
|
90
|
+
5. Fake camera moves only as slow 5-10 % pushes on 4K captures, never upscaled beyond 1.25x.
|
|
91
|
+
|
|
92
|
+
### Deterministic input replay (tier B, the best case without a film build)
|
|
93
|
+
|
|
94
|
+
Record a human performance once, live, then re-run it as often as needed with a stepped clock, a free camera and any time
|
|
95
|
+
scale. Fighting games with rollback netcode, speedrun tools, many rhythm games and any engine where you own the loop can
|
|
96
|
+
do this.
|
|
97
|
+
|
|
98
|
+
1. Record input against frame numbers, not wall-clock timestamps.
|
|
99
|
+
2. Input replay reproduces state only if the simulation is a pure function of state and input: seed every random
|
|
100
|
+
generator, fix the timestep, remove wall-clock reads (weather, sky and voice seeds on the wall clock are the usual
|
|
101
|
+
offenders).
|
|
102
|
+
3. **Verify per frame:** hash the player's and key entities' state every frame on the live run and on the replay. The
|
|
103
|
+
first frame they differ is where determinism broke.
|
|
104
|
+
4. If determinism cannot be fixed, fall back to state recording: write every relevant transform and animation state per
|
|
105
|
+
frame during live play and replay it as a puppet for camera passes. Physics and particles then need re-simulation or
|
|
106
|
+
are approximate.
|
|
107
|
+
5. Slow motion through the stepped clock changes nothing in the outcome, because inputs are keyed to simulation frames.
|
|
108
|
+
Hitstop timers on raw time are the trap.
|
|
109
|
+
6. Human feel stays human because a human played it. Scripted input still needs the bot-tell checks.
|
|
110
|
+
|
|
111
|
+
### No free camera
|
|
112
|
+
|
|
113
|
+
Use every camera the game ships (photo mode, spectator, kill cam, death cam, replay TV cameras, the menu character
|
|
114
|
+
viewer, cutscene cameras); look for mods and console variables; if you have source, add a stage camera with a cut that
|
|
115
|
+
resets temporal history (usually an afternoon). Without any of these, gameplay carries the film: transitions come from
|
|
116
|
+
objects in the gameplay frame, slow pushes on 4K only, a lower cinematic share said out loud, and no faked third person
|
|
117
|
+
made of motion graphics.
|
|
118
|
+
|
|
119
|
+
### 2D only
|
|
120
|
+
|
|
121
|
+
Capture at native resolution and upscale by an integer factor with nearest-neighbour; whole-pixel camera positions. The
|
|
122
|
+
cinematic layer is scale and time: integer zoom steps or in-engine zoom, slow motion through the stepped clock, the event
|
|
123
|
+
replayed at 2x or 4x zoom on the consequence; parallax layers pulled apart for depth moves. Keep the art's own animation
|
|
124
|
+
frames in slow motion (a 12 fps walk at 0.25x shows each drawing for 20 output frames; never interpolate drawings), and
|
|
125
|
+
snap sprites to the pixel grid at every step. Text in the world uses the game's pixel font on its grid. Screen shake and
|
|
126
|
+
palette flashes are the game's own grammar: keep them where they mark the event.
|
|
127
|
+
|
|
128
|
+
### No characters
|
|
129
|
+
|
|
130
|
+
The hero is the system's visible agent (a cursor, a piece, a ball, a vehicle, a structure). The money event is the
|
|
131
|
+
cascade; the receiver's side is the consequence up close. Macro lenses and slow orbits do the cinematic work, and depth of
|
|
132
|
+
field is safer here than anywhere. Repetition is the risk: vary scale and colour per section.
|
|
133
|
+
|
|
134
|
+
## 4. Take checks before a take is offered
|
|
135
|
+
|
|
136
|
+
A take is offered only after the checks that need no GPU (about 20 s): payoffs on grid frames; no dead time; only the
|
|
137
|
+
high-skill form of the core action; opponents moving; HUD and feed names right; cause before effect; the cinematic's
|
|
138
|
+
receiver matching the payoff; smoothness on every frame. And the rig's own checks: frame count, duplicate signatures,
|
|
139
|
+
clock skips, the quality preset, and every payoff the spec asked for actually happened.
|
|
140
|
+
|
|
141
|
+
**Preflight offline before a take takes a slot.** Every way the scripted run can fail, checked without a GPU. On OG
|
|
142
|
+
Snipes each of these lost real takes: lines of fire, thin cover, shields, lobby bots, pace marks on another payoff's line,
|
|
143
|
+
clock sums. Racing: the line, AI pace, contacts. Fighting: inputs against frame data.
|
|
144
|
+
|
|
145
|
+
**Clock-easing traps:** changing the clock moves where an action lands (an eased clock after a hop lost a kill); at 0.3x
|
|
146
|
+
one simulation frame spans 3.3 output frames, so a key may be unreachable; a payoff cinematic under 28 frames cannot fit
|
|
147
|
+
the V ramp. Check every take's speed log: fail any one-frame change over 15 % on a cinematic and any run of 20 or more
|
|
148
|
+
frames at constant speed after a payoff.
|
|
149
|
+
|
|
150
|
+
**The master format:** capture at it, or keep every select seeded and re-shoot it at the master size, verified (frame
|
|
151
|
+
count, speed curve, payoff frames, receivers, the camera within 0.15 m and 4 % field of view) before the master.
|