Why Your Photos Never Leave Your Computer
An engineering transparency breakdown of LuminaCull's local-first browser architecture, the W3C File System Access API, and zero-telemetry design.
1. The Problem With "Cloud Photo Cullers"
Most web-based photography tools follow a traditional client-server architecture: they instruct users to upload gigabytes of personal or commercial photos to an external cloud server, process them remotely, and prompt downloads later.
For photo culling, this model fails on multiple fronts:
- Privacy Vulnerability: Personal family memories or commercial client shoots are transmitted to third-party databases.
- Bandwidth Bottlenecks: Uploading 10,000 photos (often 50GB–100GB) takes hours on typical home or studio connections.
- Cost Surcharges: Cloud storage servers require recurring monthly subscriptions to offset bandwidth expenses.
2. How LuminaCull Works 100% Locally
LuminaCull uses the W3C File System Access API, a modern web standard implemented in Chromium-based desktop browsers (Google Chrome, Microsoft Edge, Brave, and Opera).
When you click "Select Photos Folder", the browser invokes window.showDirectoryPicker(). This presents a native operating system folder dialog. Once selected, your browser grants LuminaCull an isolated FileSystemDirectoryHandle pointing directly to that folder on your local storage drive.
At no point is any file streamed across the network. All file system handles, image streams, and binary slices stay inside your browser's local sandbox memory.
3. The Zero-Telemetry Commitment
Zero External Requests
LuminaCull makes zero network API calls during your session. No photos, thumbnails, or filenames are ever sent outside your machine.
Zero Analytics or Tracking
There are no tracking pixels, Google Analytics scripts, Facebook beacons, or session recording libraries running in LuminaCull.
Zero Account Requirement
We do not collect names, email addresses, passwords, or credit cards. You are completely anonymous.
Full Offline Operation
Once loaded, you can disconnect your Wi-Fi or unplug your Ethernet cable. LuminaCull will continue to cull photos seamlessly.
4. Client-Side Memory & Performance Safety
Handling 10,000 high-resolution photos in a single browser tab requires strict memory discipline:
- Binary Slice EXIF Parsing: LuminaCull's zero-dependency EXIF reader only slices the first 64KB of each image to extract camera capture timestamps and exposure specs without decoding the full multi-megabyte image buffer.
- Sliding-Window Blob Preloading: Active photos are converted to ephemeral object URLs (
URL.createObjectURL()). As you advance through photos, previous URLs are immediately revoked withURL.revokeObjectURL()to trigger browser garbage collection. - Recycled DOM Virtual Filmstrip: The bottom filmstrip renders only the thumbnails currently visible in the viewport plus a 2-item buffer, keeping DOM node count constant regardless of whether your folder has 100 or 20,000 photos.
5. How to Verify LuminaCull's Privacy Yourself
We encourage every user and technical auditor to verify our privacy claims directly:
- Open LuminaCull in Google Chrome, Edge, or Brave.
- Press F12 or Ctrl+Shift+I to open Developer Tools.
- Click on the Network tab.
- Select a folder with 5,000 photos, cull through several dozen images, and click "Finish Culling".
- Observe the Network tab: 0 requests are made after initial static asset delivery. Zero bytes of image data leave your machine.
Experience 100% Private Photo Culling
No upload, no account, no installation. Pure local performance.
Launch LuminaCull Now