1Photograph every page
Flat, evenly lit, camera square over the page. Take them in order — the filenames keep that order all the way to the zip.
2Drop the set in
Up to twenty pages at a time. Every page gets the same treatment, so the document stays consistent from cover to signature.
3Take back a PDF or a zip
One PDF with the pages in the order you named them, or a zip of separate images — whichever the other end will accept. A handful of pages typically lands somewhere around a megabyte, against the thirty they started at.
Questions
How do I compress a scanned document without losing text quality?
Leave the width where it is and use the quality setting. The two controls fail in different ways on a page of text: quality softens the picture gradually and evenly, while narrowing the page destroys the strokes of the letters, and the small print is the first thing to go. This page already comes set up that way — enough compression to make the file manageable, at a width that still lets somebody read a clause or copy a reference number off the page. If it is still too large, take the quality down a step before you touch the width.
Will the small print still be readable?
Yes, and that is the reason this page exists separately. A photograph survives heavy compression because nobody is inspecting individual leaves; a page of text does not, because somebody is going to read a clause or copy a reference number off it. The quality here is deliberately set higher than it would be for holiday photos. A page of ordinary body text comes back around 150 to 200 KB and reads comfortably on screen.
Why are the photos so enormous to begin with?
A phone camera does not know it is looking at paper. It captures twelve megapixels of mostly white, at a level of detail meant for faces and landscapes, and writes three or four megabytes per page. Nearly all of that is describing the texture of the paper.
Can I get a PDF instead of a zip of images?
Yes. Once the pages are done, the results offer Download as one PDF beside the zip, with the pages in the order the files were named. That offer needs the pages to be coming back as JPEG, which is what this page is set to produce. It works the other way round too: if what you already have is a PDF, drop that in instead of photographing it and every page comes back as an image.
Is it sensible to put a contract or an ID through a website?
Judge it for yourself rather than taking a reassurance. Each upload sits at a random, unguessable address and is deleted automatically. Nothing reads the contents, and nothing is done with them but making the smaller versions you asked for. The camera and location data are stripped from what you download, so the copy you send on does not record where you photographed it. The Privacy Notice sets out the whole of it. For genuinely sensitive originals, an offline tool on your own machine is still the more cautious choice, and that is a reasonable thing to prefer.
I have more than twenty pages.
Do them in batches of twenty. The settings stay put between batches, so page twenty-one onwards is treated exactly like the first twenty, and you can unzip the lot into one folder afterwards.