# Lumina Studio — Full Reference > Lumina Studio is an AI-powered creative production platform for design and marketing teams. ## Signal Test — Launch Trailer Lumina Studio's launch trailer, "Signal Test," is an 85-second cinematic product demonstration built entirely within the Workshop design language. The film opens cold — no title card, no voiceover — with a dark anodized-steel mainframe panel and a single boot sequence: `LMNA-OS MK-VI · INITIALIZING`. The framing is deliberate: this is hardware documentation, not a marketing video. The line that has resonated most with early viewers: *"This was made by people who get what serious creative work feels like."* That is the thesis. Every frame earns it. The Workshop design language governs every visual decision in the trailer. Klein Blue (#002FA7) carries signal state and action — it appears on armed module headers, CTA callouts, and the final savings reveal. IBM Plex Sans Condensed runs all display type at tight negative tracking (-1.6px to -2.4px). IBM Plex Mono handles status labels, counter readouts, and module IDs. The surface is anodized steel neutrals (Steel-900 #111213 to Steel-100 #F5F5F4) with no gradients except functional depth. Every border, LED pinpoint, VU meter, and rack screw is a real Workshop component — not a decorative approximation. If it appears in the trailer, it exists in the design system. The trailer runs six scenes with precise timing: S1 Mainframe Boot (0–8s, 480 frames) establishes the instrument-panel aesthetic with a cold power-on sequence; S2 Channel Rack Reveal (8–25s, 1020 frames) arms five modules — AI Stock Gen, Pro Photo Editor, Video Studio, PDF Suite, Canvas Editor — each with individual LED confirmation and VU meter activity; S3 Goal Set (25–32s, 420 frames) surfaces the audience's actual pain point using a typewriter LCD readout: "SUMMER LAUNCH CAMPAIGN · ASSETS NEEDED · 5 PLATFORMS · 72H DEADLINE"; S4 Signal Flow Proof (32–65s, 1980 frames) is the hero sequence — a complete multi-module workflow executing in real time, producing IG Feed (1080×1080), Story (1080×1920), LinkedIn (1200×628), X banner (1500×500), and display banner (728×90) assets for a single campaign brief; S5 Deprecated Revolt (65–78s, 780 frames) is the consolidation argument — Midjourney $30, Photoshop $23, Premiere $23, Acrobat $20, Canva Pro $15 each struck through one by one, with a final reveal: $87/mo eliminated, $1,044/yr reclaimed; S6 CTA Sign-Off (78–85s, 420 frames) delivers the line "One studio. Cancel the rest." with a STUDIO ONLINE LED indicator and the lumina-os.com URL. The audio design is intentional silence-plus-SFX. There is no voiceover — the on-screen LCD captions carry all narrative. The ambient bed is a low-frequency industrial hum (40–120Hz) that reads as "server room active." SFX layer above it: relay clicks on boot events, VU meter needle impacts, typewriter keystrokes on LCD text, and a clean rack-arm tone on each module confirmation. All audio is CC0-licensed and normalized to -14 LUFS with true peak -1 dBTP. Music integration is deferred to Phase 6 of the build plan. The technical signature: master render is 1920×1080 H.264 60fps AAC 48kHz stereo, 5.17 MB, 85.056 seconds (CRF 18, no perceptual quality loss). The social vertical variant is 1080×1920 at the same fps — a letterboxed 16:9 master centered in a 1080×1920 canvas with brand strips above and below. The top strip carries the "SIGNAL TEST" display headline and a pulsing green STUDIO ONLINE LED. The bottom strip carries "One studio. Cancel the rest." and the CTA button. Both formats render at 60fps for smooth motion on high-refresh displays. The Remotion build infrastructure uses 17 unit tests (all passing) and a TypeScript token bridge verifying design system parity at CI time. Watch the Signal Test trailer via the WATCH SIGNAL TEST button on the hero section of https://www.lumina-os.com/, or directly at https://www.lumina-os.com/signal-test/signal-test-16x9.mp4. The poster frame (frame 600, 10.0s, Channel Rack Reveal mid-sequence) is at https://www.lumina-os.com/signal-test/poster.jpg. ## Product Scope Lumina Studio combines design creation, brand management, content generation, and campaign workflow tooling in a browser-based environment. ## Functional Areas - AI image generation - AI image-to-image editing with region masking — users paint a binary mask over a source image (uploaded or chosen from generation history) with adjustable brush size and opacity, then apply a text-prompted edit scoped to the masked region; the unmasked area is preserved. Platform-credit AI calls route through the api.lumina-os.com Cloudflare Worker so no API key ever ships to the browser. - AI video and storyboard support - AI image-to-video (Veo) — animate a still source image (uploaded or chosen from generation history) into a short clip by describing the motion and choosing a camera move (static, pan left/right, zoom in/out, orbit), aspect ratio, and a 2–8 second duration. Generation polls the operation to completion and drops the resulting clip into the gallery. Platform-credit calls route through the api.lumina-os.com Cloudflare Worker (the conditioning image is forwarded server-side) so no API key ever ships to the browser; users on BYOK call Veo directly with their own key. - AI Brand Studio — five brand-locked generation signatures that share one engine: every signature inherits the brand-locking spine (saved palette, tone, characters, and products), the same reference-priority model (an identity lock that is never dropped, up to 5 character references within a 14-reference budget), and the same BrandGuardian consistency scoring on every output (a 0–100 on-brand score with an itemized violation list and one-click auto-fix). The five signatures are: (1) Brand-Locked Generation (character + style lock) — produces a single brand-consistent image conditioned on saved characters and style so the same face and the same product appear across every render; (2) Virtual Photoshoot — takes one product photo and expands it into a full series of on-brand shots across scene packs (studio, lifestyle, seasonal), generating roughly five shots per pack with a per-shot BrandGuardian score; (3) Brand Spokesperson — turns a script into a talking-presenter video using a consistent presenter, last-frame continuity chaining (each clip seeds the next from its final frame), WebVTT-synced captions, and a one-click MP4 export stitched in-browser via a lazily-loaded ffmpeg.wasm core; (4) Brand Campaign Agent — a two-phase, human-in-the-loop agent where one brief first yields a copy kit plus three BrandGuardian-scored hero concepts, the user picks a hero, and the agent then produces per-channel, ready-to-post variants (Instagram, Facebook, LinkedIn, X) via Magic Resize plus an optional image-to-video campaign clip; it runs under a dedicated top-level Campaign studio mode. Platform-credit generations route through the api.lumina-os.com Cloudflare Worker (the Gemini/Veo key is a Worker secret, model names regex-pinned) so no API key ever ships to the browser; BYOK users call Gemini/Veo directly with their own key held only in local storage. - Video Studio Pro — a DaVinci Resolve-style suite with five pages (Edit, Cut, Color, Fusion, Fairlight). The program monitor renders the real frame under the playhead — a pure compositing core resolves the topmost active video clip and computes its draw rectangle from the clip transform, and a renderer hook seeks a hidden media element and paints it to the canvas with opacity and rotation. Timeline export records the program canvas to a downloadable video via the browser's MediaRecorder over captureStream, choosing the best supported container per preset (WebM/VP9 by default, MP4/H.264 where the browser supports it) and disabling export with an explanation when no format can be recorded; a progress meter runs during export and a download link appears on completion. The Color page scopes — waveform, RGB parade, vectorscope, and histogram — are computed from real program-frame pixels (Rec.709 luma, 256-bin per-channel histograms, per-column luma/parade statistics, and Rec.601 chroma), not a simulation, re-sampling the current frame as the playhead moves. Projects autosave to the browser's IndexedDB on a debounced interval and restore automatically on reload; users can also save, restore, and delete named snapshots. Media file blobs are never persisted — a restored project prompts the user to re-import its media. Degrades to localStorage when IndexedDB is unavailable. - Design canvas workflows - Brand kit and reusable assets — beyond colors, typography, and logos, each brand kit has an asset library: a per-user repository of reusable images, icons, graphics, and logos that users upload, browse in a tile grid, insert onto the canvas in one click, and delete. Assets live in a per-user folder of a public Storage bucket with folder-scoped owner-only row-level security. Brand kits also support custom font upload and type pairing: users upload their own typefaces (woff2/ttf/otf), which are registered at runtime via the FontFace API and previewed live, then pick a heading + body family pairing that persists on the kit and is enforced by the brand-compliance checker. Uploaded fonts live in the same per-user folder-scoped public Storage model as assets. - Campaign planning and content calendar - Team collaboration — asynchronous, workspace-based. Team plans provide shared workspaces with role-based permissions (viewer/editor/admin) and a review queue where approvers decide on requests with per-request status. Three enterprise capabilities are not finished: submitting a design for approval is not available in the studio, Lumina does not record audit events, and saved white-label branding is not applied. Single sign-on (SAML/OIDC) is on the roadmap and not available on any plan today. Canvas edit history is saved to the cloud and restorable from any device. Lumina does not support simultaneous multi-user editing of a single canvas or live cursors; teams divide work per member rather than co-editing one document. - Template marketplace — a template is a canvas design. Users browse by category, open a free template or buy a paid one through Stripe Checkout, and it opens on the Canvas at its own artboard size. Creators submit the design on their canvas through a submission portal (validated listing intake: title, description, category, tags, price; the thumbnail is rendered from the design), and a moderation queue — open to marketplace moderators only — approves or rejects pending submissions with an optional reason recorded back to the creator. Creators can onboard for payouts via Stripe Connect (beta, test-mode): a hosted onboarding flow creates an Express account, a live status poll reflects payout eligibility, and template sales split with a platform fee routed through a stateless payments Worker — the Stripe secret never reaches the browser. ## AI Usage Allowances Each plan includes a monthly allowance of AI generations, counted separately per AI feature type (image, video, text, audio). Allowances are measured over the calendar month in UTC and reset at the start of each new calendar month; they are not pro-rated and unused generations do not carry over. A generation counts against the allowance once the underlying AI provider has processed the request. This includes a request the provider's own content-safety systems decline to complete, and a response cut short because the output limit was reached — provider compute is spent reaching either outcome. A generation does not count when the request fails before or during processing: a provider outage, a network error, a timeout, or a request Lumina rejects for an invalid or unsupported model. Lumina does not issue refunds, credits, or allowance restorations for a generation the provider declined; the guidance is to reword the input rather than resubmit it unchanged. Users who supply their own AI provider API key ("bring your own key") generate directly between their browser and their provider, are billed by that provider at its own rates, and do not draw on the allowance included with their plan. The binding statement is Terms section 4.6; the plain-language version is in the user guide under Billing & Plans. ## Audience - Designers - Marketing teams - Content creators - Small and mid-size creative teams ## URLs - Website: https://www.lumina-os.com/ - App: https://app.lumina-os.com - Pricing: https://www.lumina-os.com/pricing/ - User Guide: https://www.lumina-os.com/guide/ - Blog: https://www.lumina-os.com/blog/ ### Feature Pages - AI Image Generator: https://www.lumina-os.com/features/ai-image-generator/ - Video Studio: https://www.lumina-os.com/features/video-studio/ - Photo Editor: https://www.lumina-os.com/features/photo-editor/ - PDF Suite: https://www.lumina-os.com/features/pdf-suite/ - Brand Kit: https://www.lumina-os.com/features/brand-kit/ ### Comparisons - Lumina Studio vs Canva: https://www.lumina-os.com/compare/lumina-vs-canva/ - Lumina Studio vs Adobe Creative Cloud: https://www.lumina-os.com/compare/lumina-vs-adobe/ - Lumina Studio vs Figma: https://www.lumina-os.com/compare/lumina-vs-figma/ - Lumina Studio vs Photoshop: https://www.lumina-os.com/compare/lumina-vs-photoshop/ ### Legal - Privacy Policy: https://www.lumina-os.com/privacy/ - Terms and Conditions: https://www.lumina-os.com/terms/ ## Crawl Signals - Sitemap: https://www.lumina-os.com/sitemap.xml - Robots: https://www.lumina-os.com/robots.txt - LLM summary: https://www.lumina-os.com/llms.txt - Blog archive for LLMs (one entry per post): https://www.lumina-os.com/long-llms.txt - Machine-readable pricing: https://www.lumina-os.com/pricing.md ## Publisher - Parent company: Strategia-X - Company URL: https://www.strategia-x.com - Support: support@strategia-x.com - YouTube: https://www.youtube.com/@StrategiaX ## Blog The blog has 47 posts, indexed in one place — the archive at https://www.lumina-os.com/long-llms.txt, one entry per post. Blog: https://www.lumina-os.com/blog/