Help, privacy & practical limits
DocScan processes page images locally. Ordinary app, font and worker assets still load from the host. There are no analytics, uploads, remote processing fallbacks, or document persistence.
Session limits
Up to 20 pages, 60 MP total working pixels, and 100 MB total normalized page images. Each source is at most 25 MB / 60 MP / 15,000 px per edge; each working page is at most 12 MP / 5,000 px per edge. Output is capped at 150 MB. These provisional guardrails are not a guarantee of device capacity. Downsampled working images preserve no original file metadata.
Compatibility and cancellation
Camera requires HTTPS or localhost and getUserMedia; image import requires createImageBitmap and Canvas. Worker/OffscreenCanvas processing is used when available, with a local Canvas fallback. Cancel stops a worker immediately; fallback cancellation takes effect between row batches or after a native encoding step. No percentage is shown within a page when it cannot be measured.
Reset and resource cleanup
Reset page restores its corners, crop, rotation and color settings. Clear workspace drops all page/output references and stops the camera. The camera also stops when the document is hidden; pagehide clears the session, including on back/forward cache navigation. This is not secure erasure of browser/OS memory. Original files remain unchanged.
Known verification limits
Automated checks use desktop Chrome with synthetic images and camera streams. Physical front/rear selection, mobile cameras, Firefox and Safari require separate real-device checks. PDF contains images only; there is no OCR, automatic edge detection, certification, storage, or account. If processing fails, use fewer or smaller images and re-prepare. Downloads can be affected by browser settings; confirm the file was saved before clearing.