PDF XChange Editor OCR error

Sham · Post by **Sham** » Fri Feb 25, 2022 7:27 am

Hello,

I heavily use PDF XChange Editor which is the best PDF editing software I've ever used.

There is an error when I perform OCR especially a PDF document which contains some pictures.

It does not happens irregularly, at specific point in some PDF documents happens this error.

For example, I say there are 800 pages in a one single PDF file and I selected page range "All". Then at 27 page happens this error, no matter how many times I perform OCR on the same range.

I may not enable "Ignore text in graphics" option because I need some texts in graphics to be searchable.

I need a solution. Thank you.

Fri Feb 25, 2022 2:53 pm

Hello Sham,

Welcome to our Forum.

Could you please tell me what version is your PDF Editor?

You can see this information in the PDF Editor's Help -> About menu.

Also, if it is possible, could you please send us one of the documents you have this issue with?

Regards.

Sham · Post by **Sham** » Sat Feb 26, 2022 12:33 am

Hello,

Thank you for your reply.

My current version is the latest version which is 9.2.359.0.

And I can send the PDF document sample about which we are talking, and it is about 35MB, 40 pages. Please let me know where I can send it.

And for what it's worth, I had noticed some unusual activities of my computer when the OCR error happens. CPU usage decreasingly goes down from 100% to 1%, and RAM usage follows the pattern of CPU usage. Here are two pics about it.

: proof 1.JPG (28.32 KiB) Viewed 1454 times

: proof 2.JPG (25.9 KiB) Viewed 1454 times

Thank you.

Mon Feb 28, 2022 10:44 pm

Hi Sham.

I think 35 MB will be accepted on this forum. It restricts to just PDF files.

If it doesn't let you, or if you prefer not to make the PDF public you can upload it to our file service as detailed here: https://www.pdf-xchange.com/knowle ... le-service

Just let us know the file name when the upload is complete and we'll move it off of there.

Sham · Post by **Sham** » Wed Mar 02, 2022 7:24 am

Hello,

I checked the link you provided here and followed the instructions.

I've sent an email to support@pdf-xchange.com so please check.

Thank you.

Wed Mar 02, 2022 12:05 pm

Hello Sham,

Thanks! I can confirm we've received the e-mail and Paul will have access to your file a bit later today when he gets to the office (it's currently 4:05 AM where he is).

Kind regards,
Stefan

Post by **Paul - Tracker Supp** » Wed Mar 02, 2022 3:07 pm

Hi Sham,

I tested OCR with your file and it is taking a long time and eventually produces a "Catestrophic failure". Thanks for the file I have passed it on o the team for analysis nd we will post back here what we find.

regards

Wed Mar 02, 2022 7:49 pm

Hi Sham.

We investigating this case and at the moment you may use a workaround:

1. Use "Desckew Pages" feature
2. Use EOCR with Accuracy=Low

Seems the pattern-like background causing the issue:

We will look deeply into it..

Cheers.

Post by **Paul - Tracker Supp** » Wed Mar 02, 2022 9:23 pm

There is a support ticket associated with this issue now. Wile for internal use only, referring to RT#5974: catastrophic failure" message from the Editor on running OCR here will help support get you a status report.

I hope that helps.

Sham · Post by **Sham** » Thu Mar 03, 2022 2:08 am

Hi all,

Thank you for your attention to this matter.

I appreciate your response and look forward to hearing the good news or the status report from you again soon.

Best Regards.

Thu Mar 03, 2022 7:23 am

foxbud · Post by **foxbud** » Mon Mar 21, 2022 1:50 pm

Hello,

I've been experiencing this issue too, recently. However, for me, the OCR functionality doesn't work at all.

Regardless of which PDF document I'm viewing or which combination of OCR settings I'm using, I will always get the same error mentioned at the start of this thread. I can even create a new PDF document using "File -> New Document -> With Blank Pages", and I still get the catastrophic failure error if I attempt to run OCR.

In addition, when I try to select a region using the snapshot tool and then select "OCR Selected Region..." in the right click context menu, it appears to run successfully, but then when I check, it hasn't actually generated any text (not even in the PDF content tab).

All of this functionality used to work like a charm for me, but it suddenly stopped working a week or so ago. Unfortunately, I don't know what might have caused it.

System Information

Software: PDF-XChange Editor Plus Version 9.2, build 359.0 (Enhanced OCR) (Nov 23 2021; 14:32:29)
Operating System: Microsoft Windows 11 Pro Version 10.0.22000, build 22000

I'd be happy to provide any other diagnostic information you may need.

Thank you,

Post by **TrackerSupp-Daniel** » Mon Mar 21, 2022 4:14 pm

Hello, foxbud

Your issue, though similar, sounds a bit different from what is being reported here. It seems more likely that there was an installation issue in your case (if you can choose to OCR a region without getting the error, but nothing happens, that supports this).

My first recommendation, before you try anything else, would be to uninstall the software, restart the PC, and then reinstall the software directly from your account page, and try again.

Kind regards,

foxbud · Post by **foxbud** » Tue Mar 22, 2022 12:46 pm

Hello Daniel,

I gave that a try, but I'm still running into the same problem. I also tried removing the cache and configuration data from "C:\Users\x\AppData\Local\Tracker Software" and "C:\Users\x\AppData\Roaming\Tracker Software", but still no luck.

Just for completeness, I also tried running the editor from the command line to see if it printed any sort of error message to stdout or stderr, but it did not.

I also tried using the installer's "repair" tool, but that didn't seem to fix it, either.

Post by **Paul - Tracker Supp** » Tue Mar 22, 2022 5:18 pm

Hi foxbud,

As you know, we do have a ticket in the system to investigate this, if you want to add your details it may help.

We are not sure at this juncture whether you will be able to generate a crash dump as it does not appear to be the Editor process itself that fails but one called PDFX.ProcessPool.Worker.x64.exe

Do you want to try using the instructions here: https://www.pdf-xchange.com/knowle ... on-method1 but for PDFX.ProcessPool.Worker.x64.exe and see if it triggers a dump file?

If it generates a dump file it will be too large for the forum and is not something you should publicly post anyway, so please, if you get one upload it here as per this KB: https://www.pdf-xchange.com/knowle ... le-service

foxbud · Post by **foxbud** » Tue Apr 05, 2022 1:45 pm

Good morning,

I've discovered that it is only the "Enhanced (FineReader)" OCR-Engine that causes this error. The default options seems to work correctly.

Regards,

Tue Apr 05, 2022 1:51 pm

Hello foxbud,

Thank you for the additional information.

We do appreciate your effort.

Regards.

PHK · Post by **PHK** » Mon May 23, 2022 2:10 pm

foxbud wrote: ↑Tue Apr 05, 2022 1:45 pm Good morning,

I've discovered that it is only the "Enhanced (FineReader)" OCR-Engine that causes this error. The default options seems to work correctly.

Regards,

I am having the same experience (361/Win11). Why is that?

Post by **TrackerSupp-Daniel** » Wed May 25, 2022 7:14 pm

Hello, PHK

I cannot tell you why right now, but Paul and the Devs are looking into the logs which have been sent to us thus far, hopefully they can provide us some answers soon.

Kind regards,