Image limits, storage and result recovery

Browser processing keeps files local, but it still needs memory and device storage.

Compressed size is not memory use

A small compressed file can decode into millions of pixels. A 6,000 × 4,000 image needs roughly 96 MB for one four-byte pixel buffer, before canvas, decoding and export overhead. Purow accepts up to 20 MB per file and 24 megapixels, processes files sequentially and limits a batch to ten.

Input type and image dimensions are checked before bitmap decoding. A corrupt file, unsupported animation or oversized image is rejected with a useful error. Browser and device limits can be lower than the advertised maximum.

Results belong to this browser

Each successful batch has a random result URL. Its output files are stored in this device’s IndexedDB for one hour so a refresh can recover them. Opening the same URL in another browser or device cannot recover the images. Clearing site data also removes the results.

Expired records are cleaned on the next process. Delete a result immediately with its delete button if you use a shared device. The original file is never deleted by Purow.

Recover from an error

If storage is full, free device space or permit site storage, then process again. If a file is unreadable, open it in a trusted image application and export a supported static image. Use fewer or smaller images on a device with limited memory. Cancel stops the batch without uploading files or publishing a partial result.