Dangerous Edit feature in Acrobat DC

In Acrobat DC Adobe wants to facilitate editing texts in scanned documents. For this Acrobat tries to convert text defined in pixels or even vectors into editable live text using text recognition (OCR). This works quite well if Acrobat finds a suitable font on the system. If not a substitution font is used or an artificial font (ClearScan) is generated which often doesn’t look that good. Interestingly there are differences between the Mac and the Windows version of Acrobat DC. This example with vectorized text looks different after text recognition with option “Use available system fonts” and without option “Use available system fonts”.

Text recognistion is started automatically when calling the Edit feature if no text objects are recognized on the page. Pages which contains only images and/or vectors are processed automatically. In the first version of Acrobat DC this conversion removed embedded ICC profiles from the the images. This bug is gone in the first update of Acrobat DC.

You can revert the text recognition by clicking on Revert to Image on the bottom of the right hand page or by using the menu Edit > Undo ClearScan OCR (CMD+Z). However all editing already done will be lost. Therefore it’s recommended to perform the revert/undo of the text recognition as the first step after calling Edit.

This is an other typical attempt of Adobe to make life of office users easier. The fact that this might ruin the work of professional users in print production doesn’t matter! Especially in packaging it is still common to use only vectorized texts. This can have fatal consequences.

Adobe should introduce a preference to disable the automatic text recognition during Edit. I have already send a feature request to Adobe.


Leave a Comment