Why does this PDF look odd after Opening?

[walt.farrell]

I downloaded a PDF from the Internet Archive this morning, and was surprised that the pages appear blank in Affinity applications, but OK in other PDF viewers.

For example, here are pages 6-7 as shown in Firefox (and similarly in Acrobat Reader):

but in Publisher each appears like this, with just a gray image:

What has happened to the actual image? That is page 6, which should be a picture of the ballpark as shown in Firefox. Page 7 looks the same, where in Firefox (and Reader) the text block on the page is shown. (The actual text in that Artistic Text layer will not be shown, as it has no text fill. But the image of it should be there.)

This seems to happen for other (all?) PDF files at The Internet Archive, too.

I would attach the PDF if the forum would let me. Instead, you can get it from https://archive.org/download/ramblerclubsball00shep/ramblerclubsball00shep.pdf

 

Edit: Even though the forum says the upload failed, it seems to have attached it successfully anyway. Weird forum bug.

nine.pdf

[v_kyr]

That's a PDF/A document, don't know if Apub does support this fully (aka read/write).

[Alfred]

15 minutes ago, v_kyr said:

That's a PDF/A document, don't know if Apub does support this fully (aka read/write).

I’m pretty sure I saw something in another thread about PDF/A not being supported.

Edit: Here’s the thread:

 

[PaulEC]

I just tried opening it in Foxit Reader, then printing it as a PDF using Primo. I was then able to open the new file in APub. (All the pages were images, so if you want editable text you'd have to OCR it.)

[walt.farrell]

Thanks, everyone.

[v_kyr]

54 minutes ago, LondonSquirrel said:

...if you open the PDF in Preview and then export it to PDF, it will open in APub. The size difference is not much either - the exported PDF is about 3MB larger.

Good joke, not sure how you managed that then, since here it (re)exports via Preview.app as 3.76 GB PDF then!

[Old Bruce]

16 minutes ago, v_kyr said:

Good joke, not sure how you managed that then, since here it (re)exports via Preview.app as 3.76 GB PDF then!

Version 2 is the exported one.

44 minutes ago, PaulEC said:

(All the pages were images, so if you want editable text you'd have to OCR it.)

Here on Mac using Preview to export a PDF gives me text as text in the exported version.

[v_kyr]

19 minutes ago, LondonSquirrel said:

Works for me:

You should provide that "nine-ex.pdf then here for walt.farrell, as far as that opens correctly in APub. - Here under El Capitan the Preview app will always generate a >= 3.xx GB PDF file, there is no OCR option on this older one.

[walt.farrell]

Just now, v_kyr said:

You should provide that "nine-ex.pdf then here for walt.farrell, as far as that opens correctly in APub. - Here under El Capitan the Preview app will always generate a >= 3.xx GB PDF file.

Thanks, but I don't really need it. It is more a curiosity question at the moment, and an attempt to see if there is a bug in the Affinity processing, and if so whether it is already known.

(Ideally, of course, Affinity processing would be able to handle PDF/A files. But if not, it would be better that they refuse to Open the file and tell the user why.)

[v_kyr]

1 minute ago, LondonSquirrel said:

I'm using the same.

El Capitan? - Interesting, since I don't have any settable OCR recognition & conversion options in my MacOS 10.11.6 based Preview app here.

[v_kyr]

3 minutes ago, LondonSquirrel said:

Apologies. No, I am not using El Capitan. I'm using Catalina.

Ah Ok, you've all (Old Bruce & you) are using newer MacOS versions, where things like Preview.app have been enhanced accordingly to support more PDF conversion/output generation features.

[Hangman]

4 hours ago, walt.farrell said:

I downloaded a PDF from the Internet Archive this morning, and was surprised that the pages appear blank in Affinity applications, but OK in other PDF viewers.

I think it is likely down to a lack of support for JPEG 2000, in particular the .jpx file format in Publisher which is how the scanned pages are formatted in the PDF file. It would also explain why PDF Readers such as Apple Preview and Adobe Acrobat can open the file since they do support JPEG 2000.

Publisher is also treating the images strangely...

[lacerto]

Obsolete.

[walt.farrell]

2 hours ago, lacerto said:

EDIT2: When Affinity apps directly open the document, they actually open the OCR content of the document.]

They open both the image layer, and the OCR layer, in my experiments. However, the OCR layer is unusable without adjustments, as the text in it has no fill color, and is thus transparent. (Besides being relatively raw OCR, and containing OCR errors that need to be corrected.)

 

2 hours ago, lacerto said:

EDIT3: ...and below two views of the document when its OCR content is opened and exposed in Photo:

I presume you must have adjusted the text fill to "expose" it? But I'm also curious that you also still have the image layer showing, and as it's above the text layers, did you do something else to let the text show through?

[lacerto]

Obsolete.

[lacerto]

Obsolete.

[Dan C]

Hi @walt.farrell!

I concur that this document is being displayed as such in Publisher due to the combination of PDF/A compression and JPEG 2000 being used.

As I understand it, we do not currently support importing/exporting to the PDF/A compression standard - equally we don't currently support the JPEG 2000 file format.

I'm going to be logging this specific file with our developers now, to see if we can better handle these types of documents in the future, or perhaps show a warning that due to compression the document may not look as expected.

I hope this helps

[thomaso]

4 minutes ago, Dan C said:

I concur that this document is being displayed as such in Publisher due to the combination of PDF/A compression and JPEG 2000 being used.

But why this specific page 6 in particular? – It appears all images in this PDF are in a PDF/A and compressed as JPEG 2000. Also, page 6 contains additional OCR text like other pages do. Can you / someone detect what makes page 6 to be different from others?

[walt.farrell]

15 minutes ago, thomaso said:

Can you / someone detect what makes page 6 to be different from others?

For me, all pages are handled the same, and there is nothing unique about page 6. I just chose that one as an example.

[walt.farrell]

26 minutes ago, Dan C said:

equally we don't currently support the JPEG 2000 file format.

Thanks, Dan.

It actually seems that Affinity does handle some JPEG2000 image files, from my experimenting. But perhaps not within PDF files?

Also, @lacerto mentioned above that the image layers can be rasterized, which makes them visible. I have not tried that yet, but that would also seen to suggest that JPEG2000 is not the issue here.

[Dan C]

49 minutes ago, walt.farrell said:

It actually seems that Affinity does handle some JPEG2000 image files, from my experimenting. But perhaps not within PDF files?

This is certainly a possibility - I've looked further into JPEG2000 images and it appears as though in a previous update we internally added OpenJPEG to try and handle JPEG 2000 files but within the bug report confirming this, we noted it does not work 'across the board' with all JPEG 2000 files - hence we don't offer 'official' support for these filetypes currently. 

I could imagine when going through the PDF library, these files are handled differently (perhaps not using OpenJPEG), but I would need the developers to confirm this.

I'll be updating the file log with this info now, as it may help them to identify the cause of this file acting this way 

[lacerto]

Obsolete.

 

[lacerto]

Obsolete.

[R C-R]

2 hours ago, lacerto said:

Here is also a Photoshop 2021 saved JPG2000 image that Affinity Photo can open...

FWIW, all 3 Affinity apps can open it, but in the Mac version of APub it can't be opened from File > Open (it is greyed out) but like with other raster image formats APub will happily open it via the Finder right-click "Open with" option.

The same is true for any file I could open in the Mac Preview app & from it used Export to the JPEG-2000 format, even GIFs, PSD's, anything else Preview app can open. This gives the export a jp2 file extension. Also seems to work with JPEG2000 files with the j2k or jpf extension & maybe others -- basically anything I tested that on the Mac has a "Kind" of "JPEG 2000 image."

[lacerto]

Obsolete.