I am reading a long Stanford University research paper on AI. I am attempting to highlight text using DEVONthink mobile. The problem that I’m running into is that the highlighter is too “tall.” I cannot find a way to adjust it like one can in program such as PDF Expert. I have two questions
Thanks in advance for the help.
I think you will be able to extract the highlights… but you will not be exactly sure that you highlighted what you wanted to highlight. I think perhaps the issue here is that the highlighter tries to highlight what it thinks is a “line height” but it cannot extract the proper height from the font definition. Does it also happen in DT desktop and/or Preview?
You are using the ‘free form’ annotation tool that is a translucent marker style – it writes all over the page rather than follow the line height of the text. You can’t extract hightlights this way, no – it’s essentially a large paint brush.
For extraction, you need to work from an OCRd PDF and use the regular ‘text highlighter’ tool which is on the far left of your annotation tool bar (letter T in a coloured box).
But you can, in fact, adjust the thickness of the free-form highlighter too, if you press the coloured dot as shown here (and adjust opacity and thickness):
This response is correct in every way!
Also, in case you did not already know, the DEVONthink forum is an excellent place to seek support and learn more about the app.
And their support documents are thorough too.
And the Text Highlight tool produces broadly compatible annotations, useful for DEVONthink, Preview, etc.
The Freeform Highlight tool essentially produces a shape, not associated with any particular text.
PS I am never a fan of the two column layout for the purpose of annotation. In my experience it’s more hit and miss since the OCR engine needs to deal with two independent text blocks on one page, and sometimes runs these together (ditto for on-page shoulder notes 
). Particularly when I know I want reliable highlighting and the ability to extract quotations, I re-OCR the PDF with my stand-alone ABBYY Finereader (DT3’s inbuilt OCR engine also does a good job).
Lots of the automated OCRing you get on e.g. Google Books, or Archive.org materials, or some older academic journals, leaves a lot to be desired in terms of getting accurate OCR (you might find that the text layer isn’t exactly level with the printed text, which can also give you the sense that you are highlighting ‘between the lines’).
Anyway – happy highlighting!
Thanks, everyone, problem solved! I did not realize I could hold the colored circle to bring up Apple Pencil options. Very much appreciate it. You folks are why this forum is so valuable and a pleasant place to hang out.
