Built for True Photo Privacy
Who created CleanSlate, why it exists, and how it strips hidden metadata without degrading your files.
Created by Octobridge
Software Engineering
Hi, We are Octobridge. A small team of developers passionate about privacy and security. With a background in software engineering, we care deeply about building dependable, transparent tools that solve real-world privacy problems without unnecessary complexity or server-side data collection.
Why It Was Built
We built this utility to help us make our photos safe before sharing them online. Modern smartphones and digital cameras silently embed a surprising amount of sensitive data into every shot: exact GPS coordinates of your home or travels, timestamps, camera and lens serial numbers, owner names, and editing software history.
Most existing “EXIF remover” websites require uploading your private photos to someone else’s remote server—defeating the entire point of privacy—or they re-encode the image, causing generation loss and muddying colors. CleanSlate was engineered so my photos (and yours) never leave the device and never lose a single pixel of quality.
Methodology: How It Works
Rather than re-compressing media through an HTML5 canvas or uploading files to a backend API, CleanSlate performs deterministic, lossless binary surgery directly in your browser using JavaScript typed arrays (ArrayBuffer, Uint8Array, and DataView).
1. Strict Magic-Byte Verification
CleanSlate never blindly trusts file extensions. Every file is inspected at byte zero for authentic binary signatures (such as FF D8 FF for JPEG, 89 50 4E 47 for PNG, RIFF/WEBP, %PDF-, and ISOBMFF box headers) before any byte is modified.
2. Lossless Image Scrubbing (JPEG, PNG, WebP)
For JPEGs, the parser walks marker segments and excises APP1 (EXIF, GPS IFDs, and XMP packets), APP12, APP13 (IPTC/Photoshop), vendor markers, and COM comments while preserving ICC_PROFILE color data and the compressed SOS pixel bitstream untouched. For PNG and WebP, ancillary metadata chunks (eXIf, tEXt, zTXt, iTXt, tIME, XMP) are removed and VP8X header flags are updated cleanly.
3. Offset-Safe Document & Media Sanitization
Deleting bytes from PDFs or MP4/MOV videos can corrupt internal byte-offset lookup tables. In PDFs, CleanSlate wipes /Info dictionary tags (/Author, /Creator, /Producer, timestamps) and <x:xmpmeta> streams in-place so the Cross-Reference (xref) table stays 100% valid. In MP4/MOV/M4A containers, metadata atoms (udta, meta, uuid, and ISO 6709 GPS coordinates) are converted in-place into zeroed free padding boxes so media sample offsets (stco/co64) never break.
4. Transparent Pre-Removal Audit
Before stripping metadata segments, the engine decodes TIFF/EXIF IFD pointers, rational GPS coordinates, XMP attributes, and ID3 frames in memory so you can see every hidden tag that was found and removed.