Serious OCR problems with DTP Office

I’m considering upgrading to DTPOffice from DTP, so am trying it out. Email import is as sweet as could be, but scanning in files via OCR is a nightmare. First, and less importantly, DTPO won’t work directly with my Brother scanner, whether via Image Capture (though this normally works perfectly well) nor TWAIN. I understand from Bill there is a problem with various Brother models, and it is certainly possible to work round by using Brother’s own Control Center app. However, this doesn’t allow multi-page scans (or using the scanner’s document feeder). The only alternative for multipage docs, then, is to scan separately, then import into DTPO.

This is where the trouble really starts. It is fantastically slow. Reading a six-page PDF, scanned at 300 dpi, into DTPO via OCR takes ten and a half minutes! I kid you not - I have timed it twice, watch in hand. This brings the working day to a grinding halt.

What could be the problem here? The docs I’ve tried so far are standard bank statements, nothing psychedelic, hardly a graphic in sight. My system may be a bit long-in-the-tooth: it’s a non-Intel 1.8 GHz G5 PowerPC with 1.5 GB of memory and running OS 10.5.5. But surely this isn’t decrepit enough to cause such a slowdown.

I should add that single sheets scanned with Control Center are OCRed in a few seconds.

This seriously makes me wonder whether the upgrade is worth the seventy bucks. After all PDFs are searchable in DTP without the OCR hassle and emails can be imported by AppleScript, even if in nothing like the style of DTPO.

(Incidentally, it’s frustrating to a longtime fan of all things DEVON to find that DTPO‘s scan feature is so clunky, never mind the OCR problem, when the notoriously buggy and infuriating SOHONotes does this from my Brother as smooth as silk just by clicking a single icon. It makes multi-page, searchable PDFs via Image Capture, and does so in the blink of an eye… I refuse to believe that DEVON can’t match Chronos, may heaven shine upon them, at something like this.)

Hi, I have problems with my brother MFC 9840CDW too. Image Capture works only when the printer is connected via USB (seems to be common issue with 10.5.5). Import to Devon Pro via ExactScan does not (no picture …). I downloaded Exact Scan 2.1 demo. This seems to work. Are there plans to update the Devon Pro Exact Scan module to 2.1? My brother has a multifeed tray and duplex options. It would be excellent to use both for OCR of multipage documents. So is there any other option?

Whenever you have a problem with ExactScan, please report it using support@devon-technologies.com together with the following information:

  • Scanner model
  • Driver version
  • OS version
  • Architecture (PowerPC, Intel)
  • Connection: USB or network
  • Detailled problem description
  • if available: Crash report

Please send us these so that we can forward it to the ExactCode team.

As to the OCR: depending on the complexity of the page it can take a long time. And I have no idea of the definition of “complexity” here, this is something to be taken up with IRIS. But if you do not set the PDF attributes, this conversion takes place and finishes in the background and you can concentrate on other work.

Thanks, Annard. I’ll send the information re ExactScan. So, re the OCR delay, there’s not much to be done. Sure, it can be done in the background, but if you’re feeding the database with a lot of material, this isn’t much comfort. I’m a novice at this, but it seems from other posts that there may be apps out there that do the job more efficiently than IRIS (Acrobat Pro is one obvious example, but clearly out of the question just for this due to price). I hesitate to get into this and would hope that IRIS will improve, if this is the engine DEVON is committed to. But is there any reason why an OCR or PDF + text scan done in another app and imported into DTPO should cause any problems?

AFAIK Adobe also uses part of the IRIS engine (or used to) for some of their products. There is not a lot of competition on the Mac, this leads to stagnation. Nothing we can do about that at the moment.

I would suggest you read our online help regarding OCR, it lists a number of programs that we support for this purpose. That is: you can send the output from them directly to DTPO and it will start the OCR process automatically. Image Capture is one of them.

Wow - thanks for the super-rapid response! I see the problem. I have read Help. It seems that most appropriate apps are those that come with the various scanners. I can use Brother Control Center, which works fine with DTPO - the problem is that it only does single pages.

My scanner (a Brother MFC-240C) appears to be one of the models by that manufacturer that DTPO can’t make work with Image Capture. It hangs while scanning. This is weird, as normally Image Capture works perfectly well with it. I use Image Capture for the workround for multipage docs (which it won’t do alone either, as far as I’ve been able to discover, though the dreaded SOHONotes, for one, does enhance it to allow this). Generally this means scanning individual pages, stitching them together in Preview, Skim or whatever, then importing them via OCR into DTPO (which is where the delays happen). Laborious and clunky, but it works, at the cost of some gnashing of teeth.

ExactScan also hangs, but while initializing the scanner. I’ve just left it doing this for ten minutes with no joy, and then had to Force Quit it. I’ll let Support know this.

Cheers and thanks again.