Experimental interface engine for private spatial memory.
Presence OS / Memory Atlas is an advanced prototype exploring how a private archive can become a living spatial memory field instead of a static gallery.
The project is built around a simple idea:
Drop your archive. Watch it respond to your presence.
Memory Atlas transforms archive fragments into a responsive interface shaped by stillness, return, attention, hesitation and synthetic weather. It combines local archive intake, spatial memory fields, cinematic inspection, artifact export and an experimental WebXR Memory Room.
Advanced prototype / in active development
This project is not positioned as a finished commercial product yet. It is a flagship R&D prototype for an experimental interface system.
Current state:
- Core web flow is implemented.
- Demo Memory Field is working.
- Local archive intake and reconstruction flow are implemented.
- Cinematic Inspect is implemented.
- Memory Reel export exists at prototype level.
- XR Memory Room is implemented and tested on Meta Quest 3.
- VR session trace, replay and local artifact export pipeline exist.
- Quest hand, ray, photo selection and zoom interaction is still being stabilized.
- Mobile and final production polish are still in progress.
Presence OS is the interface engine. Memory Atlas is the first experience.
Presence OS = engine
Memory Atlas = first use case
Vanishing UI = behavior language
Synthetic Weather = atmospheric state system
Memory Artifact System = export / commercial layer
The system does not treat images as passive files. Instead, each memory fragment can carry role, signal, stability, visibility, return count and weather response.
Interaction logic:
stillness -> clarity
return -> memory density
attention -> stabilization
speed -> disappearance
hesitation -> hidden context
The result is not a dashboard, not a lightbox and not a gallery. It is a spatial memory interface.
The current prototype includes the following experience flow:
Entry Ritual
-> Demo Memory Field
-> Local Archive Intake
-> Reconstruction Sequence
-> Cinematic Inspect
-> Human Result
-> Memory Reel
-> XR Memory Room
-> VR Replay / Artifact Export
A minimal signal-based entrance introducing the core premise:
Drop your archive.
Watch it respond to your presence.
Users can upload a local archive for browser-side processing. The intake flow analyses basic media properties such as format, aspect ratio, brightness, warmth, saturation, contrast and fragment role.
The system does not claim to understand the personal meaning of the image. It works with visual, behavioral and structural signals.
Uploaded fragments pass through a reconstruction ritual:
Reading light
Finding repeated signals
Detecting silence
Mapping fragments
Building archive weather
Opening field
The archive becomes a spatial field of fragments. Each fragment can wake, stabilize, vanish, deepen or respond to return and attention.
Inspect is not a standard modal or lightbox. It is designed as a memory recovery sequence:
fragment selected
-> image expands
-> background dims
-> context forms
-> memory stabilizes
The session produces a readable result:
What your archive became
Memory climate
Readable memories
Guarded traces
Export options
The session can become a vertical cinematic artifact with a phone-style preview and WebM export flow.
The same session can be explored in a WebXR spatial room. The XR layer includes gaze focus, command rail, atmosphere, memory cluster travel, hand presence experiments and two-hand inspect research.
The XR passage can be recorded as structured session trace data and transformed into replay artifacts, rather than relying only on raw headset recording.
- Entry ritual
- Demo Memory Field
- Local archive upload / intake
- Reconstruction sequence
- Algorithmic fragment classification
- Format-aware media frames
- Synthetic Weather layer
- Cinematic Inspect sequence
- Human Result / Summary
- Memory Reel preview
- WebM export prototype
- Session artifact bridge
- WebXR Memory Room route
- Quest-oriented room layout
- Gaze focus
- Command rail
- Memory distance bands
- Smooth cluster travel
- Atmospheric room layer
- Constellation / archive depth field
- Hand presence experiments
- Hand ray navigation experiments
- Two-hand distance inspect experiments
- Surface reading experiments
- VR session trace logger
- Replay timeline builder
- Replay preview
- Cinematic replay export
- Weather-based ambient sound layer
- Local artifact ZIP bundle prototype
This is an in-development prototype. The following areas are not final:
- Quest hand interaction stability
- Hand ray photo focus / selection
- Selected-only two-hand zoom gating
- Swipe gesture tuning
- Mobile and tablet layouts
- Final Memory Reel visual polish
- PDF / booklet artifact output
- Artifact Theme System
- Production-level performance optimization
- Final accessibility and reduced-motion pass
A future research layer explores selected recovered memories as controlled spatial depth objects using single-image 3D reconstruction / Gaussian Splat workflows.
This is not intended to replace the core archive experience. The correct role is selective:
selected important memory
-> recovered through attention
-> opens as Spatial Trace
This layer is planned as a separate R&D path before full Memory Room integration.
- Vite
- React
- TypeScript
- Three.js
- React Three Fiber
- Drei
- WebXR
- Canvas rendering
- MediaRecorder / WebM
- WebAudio
- JSZip
- sessionStorage
- local object URLs
- custom CSS / design system layers
src/app/
App.tsx
routes.tsx
layout/
src/features/presence/
PresenceProvider.tsx
usePresenceSignals.ts
presenceResponseMap.ts
src/features/archive/
ArchiveIntake.tsx
UploadDropzone.tsx
createLocalArchive.ts
src/features/reconstruction/
ReconstructionSequence.tsx
classifyFragments.ts
analyzeImageFile.ts
buildReconstructionBridge.ts
src/features/weather/
ArchiveWeatherProvider.tsx
SyntheticWeatherLayer.tsx
generateArchiveWeather.ts
src/features/memory-field/
MemoryField.tsx
applyPresenceToFragment.ts
src/features/inspect/
FragmentInspect.tsx
usePresenceInspectSequence.ts
src/features/summary/
AtlasSummary.tsx
src/features/reel/
MemoryReelPreview.tsx
memoryReelExportMetadata.ts
src/features/session/
buildMemorySessionArtifact.ts
memorySessionArtifactStorage.ts
saveLiveMemorySessionArtifact.ts
src/features/xr/
MemoryRoomThreePreview.tsx
readMemoryRoomSession.ts
src/features/xr/hands/
hand presence / gesture / ray / inspect systems
src/features/xr/room/
atmosphere / guide / tuning / constellation systems
/ Entry
/demo Demo Memory Field
/atlas Local Archive Intake
/summary Human Result
/reel Memory Reel
/memory-room XR Memory Room
npm installnpm run devnpm run dev -- --host 0.0.0.0 --port 5173 --httpsThen open on Meta Quest Browser:
https://YOUR_LOCAL_IP:5173/memory-room
npm run buildThe XR Memory Room is tested locally on Meta Quest 3 using HTTPS.
Recommended test sequence:
1. Open /memory-room
2. Enter XR Room
3. Check gaze focus
4. Check command rail
5. Check memory cluster travel
6. Check hand visibility
7. Check hand ray navigation
8. Check photo selection
9. Check two-hand inspect / zoom
10. Check reset / return / exit
Known active testing areas:
hand ray photo targeting
pinch/select behavior
two-hand zoom activation threshold
swipe travel skip behavior
Quest performance tuning
Memory Atlas is designed around local-first trust.
During local preview:
Processed locally.
Nothing is uploaded.
Nothing is stored.
Uploaded files are handled with local object URLs and browser session state.
Any future cloud rendering, hosted sharing, print ordering or Spatial Trace processing would require explicit user consent.
The visual language is:
dark
cinematic
slow
quiet
museum-tech
atmospheric
signal-based
restrained
The project intentionally avoids:
generic dashboard UI
cyberpunk neon
FX-showcase visuals
game HUD language
standard photo gallery patterns
Light, fog, motion and disappearance are used as part of the interface grammar.
Memory Atlas is designed to evolve into a Memory Artifact System.
Potential future outputs:
- cinematic Memory Reel
- printable Memory Booklet
- cover frame
- VR replay artifact
- premium local artifact pack
- print-ready memory products
- creator / studio workflows
- B2B / exhibition / archive installations
The commercial direction is not photo storage. It is the transformation of a private archive into beautiful interactive, cinematic, printable and spatial artifacts.
Immediate priorities:
1. Stabilize Quest hand / ray / photo selection interaction
2. Finalize selected-only two-hand zoom behavior
3. Complete XR usability tuning
4. Add Artifact Theme System
5. Polish Memory Reel visual output
6. Add PDF / Memory Booklet artifact
7. Complete mobile / tablet pass
8. Prepare public case study and demo capture
Future R&D:
Spatial Trace / SHARP feasibility
controlled single-image 3D memory objects
Artifact Pack monetization test
hosted share pages with explicit consent
B2B / installation mode
This project demonstrates:
- original product concept development
- experimental interface design
- React / TypeScript architecture
- local-first media workflow
- spatial UI thinking
- Three.js / React Three Fiber prototyping
- WebXR experimentation
- Quest-oriented testing
- cinematic interaction design
- artifact export architecture
- product and monetization strategy
Presence OS / Memory Atlas is a prototype for a broader direction: interfaces that behave less like static screens and more like living systems shaped by attention, atmosphere and presence.
Concept, design direction, product architecture and development:
brenychstudio
Project status:
Advanced prototype / in development
- GitHub: https://github.com/brenychstudio/Presence_OS_Memory_Atlas_-
- Preview image:
./public/memory-atlas/interface/interface_01_main_mvp_hero.webp - Public demo: not published yet
Not specified yet.