Skip to content

Proxy flow for large images

  1. The browser reads the image header. Below limits.maxClientMegapixels (16 by default) it decodes and keeps the full image; above it, the server produces a deterministic proxy (mode: 'proxy') or the browser decodes a proxy natively.
  2. Every edit previews on the proxy through the same WASM core.
  3. Export or download sends the original (inline, or as an S3 or URL reference you provide) plus the edit log to the server. The server renders the full resolution with the same binary and returns the file or stores it.
  4. Both sides can verify the result by hash.
ts
const editor = document.querySelector('sealpixel-editor');
editor.server = { endpoint: 'https://xyz.lambda-url.eu-central-1.on.aws/render' };

The @sealpixel/server handler enforces byte and megapixel limits before decoding a single pixel, and estimateMemoryMB(width, height) tells you the Lambda size to configure.