Problems with new OCR

the old OCR worked great. The new one? Not so much. Here’s what I’m seeing:

when I first launch a scan (from my Fujitsu ScanSnap), nothing happens. When I scan a second time, it’s like the app “wakes up” and begins to process the scan, but I get an error about something called Scribbler, that I’ve never seen before. It says it can’t install it or something, but then it processes the scan anyway and puts it into my inbox.

Then what I’m seeing is that some of the text in the PDF is found upon searching, but most isn’t. It’s a mess.

How do I fix this? I loved the old OCR engine. Hopefully this one will get dialed in but right now I need some help.

This is one of the messages in Console, if that helps:

3/8/09 8:24:31 PM DEVONthink Pro[390] ERROR: problem occurred during DEVONscribbler startup: Server com.devon-technologies.DEVONscribbler not available!

then there’s this:

3/8/09 8:20:33 PM DEVONthink Pro[390] ExactScan application not found.

The first message will be generated only the first time OCR is attempted after installing a new plugin.

The second message is nothing to worry about. DTPO2 will look for the full version of ExactScan, which is marketed by ExactCode. That message was also generated in DTPO 1.5.4.

well, I’m getting the first message every time I attempt OCR with DTP.

And again, I have a HUGE problem here in that not all text on the PDF page appears to be searchable, whereas it was with the prior engine.

I’ve got a library where I scan in all our “explanation of benefits” forms from our health insurer. then I have smart groups setup to show by year and family member. I just scanned one in last night and saw it wasn’t being picked up by the expected smart group. So I did manual searching on content and found that some content was found, most not.

Which means if I can’t fix it, the entire use of that database becomes worthless. There’s no reason to scan stuff in if I can’t find it.

I NEED this to work, and right now, it isn’t.

Update: When I said I was getting that Scribbler message every time, that was true, but I’d only run the OCR 2 separate times (different days). Each day, I got the same message, so it didn’t happen just the first time after loading the new plugin.

But, I decided to try again this morning so I could recreate the problem and it didn’t generate the message and the resulting PDF was full searchable. Then I noticed that the one from last night was not imported as a PDF + Text, but just as a PDF. Don’t know why that happened, but now they appear to be importing properly.

Weird, but at least it’s working now, thankfully.