Screenshot to OCR PDF

In DTPO I have an issue where in a screenshot to PDF conversion the size (dimensions in metadata) are “wrong” thus making the PDF of the image too small to read (before manually zooming well beyond 100%). I would like the PDF of the image to remain the same size as the original screenshot when opened.


I captured a screenshot to the clipboard (command control shift 4), pressed Command-N to paste it into my Inbox. Info shows PNG with resolution 1478x585 and dpi 72 (note screen is ~109 dpi). I then right click select Convert -> to Searchable PDF. The info for the PDF says 4.9" x 1.9" which works out to taking original resolution and assuming it was 300 dpi = 1478 / 4.9.

The problem is that at default 100% zoom the PDF is only 1/7 the width of my screen (unreadable) whereas the original image was more than 1/2 the width of screen.

In preferences I have resolution set to “Same as scan” and quality 100%, manually setting resolution to other values has no discernible effect.

The current OCR engine is optimized for scans but a future upgrade will support this. In the meantime the only workaround is probably to resize the images using a third-party software (e.g. or Automator) before OCRing them.

Thank you for the suggestion. I tried this by resizing the screenshot image by a factor of 300/72=4.16x with Preview and this did produce a dimensionally larger output PDF via DTPO OCR but the PDF file was also 10x larger and the resize was not without artifacts.

Is there a timeline for the DTPO update you refer to regarding OCR? Would this imply an update to the ancient ABBYY FineReader v8.x engine reported by DTPO About window (I see the FineReader Engine is available as version 11.x per Any approximate timeline for this?

This will also update the OCR engine and will be available next year.