Opening large PDF-files creates application crash

[Oliver S.]

After update to V2.x.x large PDF's over 1GB file size cannot be opened in publisher or designer.
Trying to place them into a document also fails and causes programm crashes.

File sizes are from 3 to 5,4 GB. My target was to modify them to a document you can work with.

System settings like RAM were correct according to the available RAM on the machine.
Example: 16GB system RAM available -> max 10GB for Publisher App

[Hangman]

Hi @Oliver S.,

The first obvious question is going to be could you upload perhaps one of the smaller files so we can take a look... I'm unsure what the file size limit for uploads to the forum is but I know files in excess of 1Gb have been successfully uploaded in the past...

Without seeing a sample file I think it's going to be difficult to diagnose potential causes of the crash though a crash report may also help...

[Oliver S.]

Hi Hangman,
uploading this files in the forum should be a little nightmare.
What i can do is opening one of this files again with publisher and paste the crash report from apple here.

Here the link to the files from German Federal Archive (Website operated by the Federal Republic of Germany).
In my case it was the 'BD2' on the right side wich i use for a research work.
Take care it takes a lot of time due very lousy download rates.
https://www.bundesarchiv.de/DE/Content/Artikel/Benutzen/Hinweise-zur-Benutzung/Unterseiten-Militaer/Militaerische-Verbaende-und-Einheiten/benutzen-speziell-milit-verbaende-einheiten-tessin.html

The PDF-files are ocr scanned from books from the German Military Archieve. The idiot who scanned it did it in colour instead of b/w and
includes a lot of space wich did not belongs to the book.

I found a workaround for the problem. Opening the 350 pages files with "Preview" delete 300 pages and rename it "***1-50.pdf"
Copy the complete file in the same folder again, delete the next pages  and rename to "***51-100.pdf".
Each file is then around 750mb and can be opened and it's possible to modify the file to b/w, remove the incorrect OCR data and resize it.

regards
Oliver

 

 

[walt.farrell]

14 minutes ago, Oliver S. said:

I found a workaround for the problem. Opening the 350 pages files with "Preview" delete 300 pages and rename it "***1-50.pdf"

Have you tried Opening the full file in Publisher, but when prompted for what to load, telling it to load pages 1-50?

[Oliver S.]

Hello Walt,

unfortunately i tried to open the full file. As described i also started to place it into a document with bad result.

Where do i find these 'PDF Options' ..... never seen this in Publisher, looks that this can help also 😀.
Tomorrow i will set up a windows 11 machine, then we'll have the same platform. Normally you find me in the Apple corner.

regards
Oliver

[walt.farrell]

54 minutes ago, Oliver S. said:

Where do i find these 'PDF Options' ..... never seen this in Publisher, looks that this can help also

They're from File > Open.

[Oliver S.]

Hello Walt,

no such options in "open file" see attached screenshot.
Maybe a special option for Windows based systems or had to be activated in the preferneces.

MacOS crash report from the opening of the PDF-file is attached.

Affinity Publisher crash report - large files.txt

[Hangman]

Hi @Oliver S.,

I see an instant crash when attempting to open Bd_2_ocr.pdf in Publisher v1.10.8, v2.0.4 and v2.3.0. This is also why you're not seeing the options @walt.farrell shows in his screengrab, the file is crashing Publisher before the PDF Options window appears.

Crash Reports

Affinity Publisher-2023-12-07-094845.ips

Affinity Publisher 2-2023-12-07-094804.ips

Affinity Publisher 2 Affinity Store-2023-12-07-094736.ips

16 hours ago, Oliver S. said:

The idiot who scanned it did it in colour instead of b/w and includes a lot of space which did not belongs to the book.

Adopting a different approach I converted the PDF to a Greyscale version using Ghostscript which reduces the file size from 5.37 GB to 235 MB. It's one to set running while going off to make a cup of coffee but you will see the progress page by page...

Quote

gs -sOutputFile=Bd_2_ocr_Greyscale.pdf -sDEVICE=pdfwrite -sColorConversionStrategy=Gray -dProcessColorModel=/DeviceGray -dAutoRotatePages=/None -dCompatibilityLevel=1.7 -dNOPAUSE -dBATCH Bd_2_ocr.pdf

The scanning is a bit silly as the pages could have been cropped though that is easy enough to do in Publisher or Apple Preview.

Here is the converted version which opens quite happily in Publisher... If you need it in a different PDF version then simply set -dCompatibilityLevel=1.7 accordingly, e.g., to -dCompatibilityLevel=1.4

You can likewise remove redundant pages using the command line rather than doing so in Publisher which would reduce the file size further.

On a side note, you will notice that because the book wasn't perfectly straight when scanned the invisible OCR text layer is broken into small chunks on some pages meaning it's difficult to copy and paste the text if that is the intention...

If you happen to have an iPhone or iPad running iOS 15 or later you will find Live Text is extremely helpful as it allows you to copy the text for entire pages and paste it as live, fully editable text in a new document.

Let us know if this helps at all...

Bd_2_ocr_Greyscale.pdf

 

[Oliver S.]

Hi Hangman,

thanks for your hint to "Ghostscript" wich i didn't know. This will help a lot.
It seems that there was also an issue with the PDF-version. Now it is possible tho choose the options from Walt during opening the file.
The work on the document like removing the bad OCR data and resizing or cropping the pages is now much easier to do.

Many thanks 🙂

regards
Oliver

[Hangman]

Hi @Oliver S.,

That's no problem at all, happy to help...

The original file is a PDF 1.7 file as is the version I've converted to Greyscale...

It would still be helpful to understand why the file causes Publisher to instantly crash so I'm hoping someone in the moderation team will be able to shed some light on that...

[Dan C]

Thanks for your report @Oliver S.!

I can confirm I've been able to replicate this crashing issue and I've reported it to our development team for further investigation, as I cannot see a clear reason for this crash to occur currently.

Interestingly on Windows, the Affinity app shows a 'File type not supported' error when trying to open this PDF.

I've been able to import a different large PDF document (3.8GB) without this crash occurring, so I don't believe it's specifically related to the size of the PDF file itself - though it's interesting to know that after resaving a reduced version of this PDF though Preview, Affinity was able to import the document & I'll be sure to include this in the development log.

Many thanks to @Hangman for the additional information and workaround provided above!