Almost every request for a "scanned copy" can be satisfied with a phone camera, provided the photos are taken and processed with a little care. The difference between a usable scan and an obviously photographed page is mostly lighting and geometry, neither of which needs an app.
Photographing the Pages Properly
Most bad phone scans fail at this step, and nothing downstream fixes them. Five things matter:
- Put the page on a flat, dark surface. Contrast at the edges makes cropping far easier and more accurate.
- Use even, indirect light. A window works well. Overhead light directly above you puts your own shadow across the page.
- Hold the camera parallel to the page, not at an angle. A tilted shot gives you a trapezoid instead of a rectangle, and correcting that always costs some sharpness.
- Fill the frame with the page and leave a small border. Zooming in later throws away resolution you cannot get back.
- Turn the flash off. It produces a bright hotspot in the middle and darkens the corners.
Photograph every page before processing any of them. Keeping the camera at one height and one lighting setup makes the finished PDF look consistent, which is most of what separates a scan from a set of snapshots.
Crop and Straighten Each Shot
Crop away the surface around the page and correct any slight rotation. This is worth the two minutes: a document whose pages are all square and consistently cropped reads as a scan, and one where each page sits at a slightly different angle does not.
Straightening also matters for the OCR step later. Text recognition works line by line, and a page rotated even a couple of degrees measurably reduces how much of it is read correctly.
Combine the Photos into One PDF
Add the photos in page order and each becomes one page of a new PDF. Three settings decide whether the result looks like a document:
- Page size: choose A4 or US Letter rather than the image's own size. A phone photo's natural dimensions produce an oddly shaped page that prints badly.
- Orientation: following the image handles a document with a mix of portrait and landscape pages without rotating anything.
- Margin: a 6 mm or 13 mm margin looks correct in print and keeps content clear of a binding.
If the photos came off an iPhone as .heic rather than .jpg, convert them to JPG first. Also worth knowing: this produces a PDF containing pictures of text, not text. Selecting or searching a word in it will not work yet — that is the next step.
Make the Text Searchable
OCR reads the letter shapes in the images and recovers the text. You can either extract that text on its own, or export a searchable PDF — which keeps the original scan exactly as it looks and lays an invisible text layer over it, so the page appears unchanged while the words underneath can be selected, copied and searched.
The searchable version is almost always what you want for a document you are filing or sending. It looks like a scan and behaves like a document.
- Run the combined PDF through the Scanned PDF to Text tool.
- Choose Searchable PDF as the output rather than plain text.
- Open the result and try searching for a word you know is in it, to confirm the layer took.
OCR accuracy depends almost entirely on the quality of the photographs. Clean, straight, evenly lit pages read close to perfectly; blurred or skewed ones produce text with errors scattered through it. If the results are poor, retaking the photos will help more than any setting.
Shrink It Before Sending
Photo-based PDFs get large quickly — a ten-page scan from a modern phone camera can easily exceed 30 MB, and most email systems still reject attachments over about 25 MB.
Compress the file before sending. For a document that only needs to be read rather than printed, a substantial size reduction is usually invisible. If it contains anything confidential, add a password before you send it rather than afterwards.
Summary
A phone camera replaces a scanner as long as the pages are shot flat in even light, then cropped and straightened. Combine the photos into one PDF at A4 or US Letter with a small margin, and remember that this alone gives you pictures of text rather than text. Running OCR and exporting a searchable PDF keeps the scan looking identical while making every word selectable and searchable. Compress before emailing, and password-protect before sharing anything sensitive.