Hello,
Most of my DT workflow involves images of physical media that I OCR. I noticed that since version 3.9.5. Apple Live Text is active when I convert images of documents to PDFs. I subsequently run OCR to make them indexed PDF+Text files.
However, I was experimenting and noticed that the ABBYY OCR was not better than the Live Text results. Live Text seems to recognize the order of lines of text better. So, I have a couple of questions about how to make the best use of the software.
Is there still any value in running the OCR plugin? Is the OCR from Live Text recognized in the database?
Live Text is not a replacement for proper OCR. It’s smoke and mirrors. There is no text layer applied to the document so logically it’s not searchable in DEVONthink.
1 Like
Thank you for the reply. Please make its limited utility more clear in the documentation. It’s listed as an improvement in the latest manual.
1 Like
Curious 
– Writing your request took longer than command>f in current documentation.
I won’t contribute to the rudeness here, but I want to communicate to the moderators that I find the DT community to be much less supportive of learning and less willing to understand the use cases of other users than other software that I have learned in the past few years.
I walked away from this discussion knowing almost nothing else about how DT encodes PDF layers, or how Live Text works, only that it’s bad and I was a fool to ask for a clarification. I used to pitch the software without hesitation but now it comes with a warning that you’re on your own.
I’m fairly certain that these questions have been discussed and clarified before in the forum. A search for “text layer” and “live text” should provide more detail.
Live Text, btw, is Apple’s thing. DT has nothing to do with it.
Also, there’s the post @BLUEFROG pointed out.
Is there a particular reason you want reiterated here what has already been explained elsewhere?
1 Like