Opening PDF in Publisher Alters Text

[TomHayes]

Program: Affinity Publisher 

Version: 1.10.5

Font type: Calibri - Regular

When opening a PDF in Affinity Publisher, that was created in Affinity Publisher, some of the text is altered. The text is not altered when the PDF is opened with Adobe Acrobat Reader. This only happens with Calibri font type.

See attached imaged for before and after export. It seems to convert "ti" to "Ï " or "Á"

Is there a was to fix the way Publisher reads the text? Is this a common issue with Publisher?

 

Regards

Tom

[Callum]

Hi Tom Hayes,

Welcome to the forums
This is quite an odd one! To get your document exported correctly you could go to File > Export select PDF and then hit the more button and then towards the bottom of the window that appears select Text as Cruves from the embed fonts dropdown. Which operating system are you using?

Thanks
C

[RM f/g]

I think it's an issue with Calibri.

Back in my Illustrator days, when I openend a pdf with Calibri, text in this font was completely illegible.

https://community.adobe.com/t5/illustrator-discussions/illustrator-opens-pdf-with-corrupted-characters/td-p/4851202

[walt.farrell]

1 hour ago, Callum said:

This is quite an odd one! To get your document exported correctly you could go to File > Export select PDF and then hit the more button and then towards the bottom of the window that appears select Text as Cruves from the embed fonts dropdown.

It seems like the usual problem with PDF files created in the Affinity applications, when they are viewed subsequently in Affinity applications. 

It is an issue with the ligatures, and the other fixes a user can perform are:

• turn off Standard Ligatures in the Typography options (Typography panel, Character panel, Text Style).
  or
• In the PDF Export options, under More, in Font Embedding, turn off Subset Fonts.

Another user has suggested somewhere else that Affinity is failing to put entries for the ligature glyphs into one of the tables in the PDF file, but I don't remember where that was suggested. It seems to me that this would be the best fix  

[MickRose]

For what it's worth Xara Designer seems to open the Publisher created PDF correctly. That suggests to me it's the reading of the PDF which has an issue rather than the exporting. I think @walt.farrell is on the right track. Turning off Subset Fonts when making the PDF works here as a workaround.

[Alfred]

8 hours ago, TomHayes said:

The text is not altered when the PDF is opened with Adobe Acrobat Reader.

 

3 minutes ago, MickRose said:

For what it's worth Xara Designer seems to open the Publisher created PDF correctly. That suggests to me it's the reading of the PDF which has an issue rather than the exporting.

I had always assumed that the ‘Subsetting’ option resulted in ligatures not being included in the exported file, but if Adobe Reader and Xara Designer open the PDF file correctly then they must be in there somewhere!

[Alfred]

2 minutes ago, Hens said:

Isn't subsetting only including the base font with the extra used glyphs?

That’s what I would expect, but the ligatures aren’t correctly displayed when the PDF is reopened in the Affinity apps. This makes it look as though those glyphs are missing from the file.

[walt.farrell]

5 hours ago, walt.farrell said:

Another user has suggested somewhere else that Affinity is failing to put entries for the ligature glyphs into one of the tables in the PDF file, but I don't remember where that was suggested. It seems to me that this would be the best fix  

@Callum: The other Affinity user I mentioned was @LibreTraining, and here is one post they made which discusses this issue with how the Affinity applications (or the PDFLib library, perhaps) handles the ToUnicode table in PDF files when the fonts are subset:

 

[TomHayes]

Thank you all very much for the prompt support!

13 hours ago, walt.farrell said:

It seems like the usual problem with PDF files created in the Affinity applications, when they are viewed subsequently in Affinity applications. 

It is an issue with the ligatures, and the other fixes a user can perform are:

• turn off Standard Ligatures in the Typography options (Typography panel, Character panel, Text Style).
  or
• In the PDF Export options, under More, in Font Embedding, turn off Subset Fonts.

Another user has suggested somewhere else that Affinity is failing to put entries for the ligature glyphs into one of the tables in the PDF file, but I don't remember where that was suggested. It seems to me that this would be the best fix  

I used this fix (turn off Standard Ligatures) and I do not have any issues now.

[TomHayes]

15 hours ago, Callum said:

Hi Tom Hayes,

Welcome to the forums
This is quite an odd one! To get your document exported correctly you could go to File > Export select PDF and then hit the more button and then towards the bottom of the window that appears select Text as Cruves from the embed fonts dropdown. Which operating system are you using?

Thanks
C

Thanks for the suggestion Callum,

I would prefer save with PDF with text rather than curves but that would be on workaround.

I am using Windows 10.

[TomHayes]

15 hours ago, RM f/g said:

I think it's an issue with Calibri.

Back in my Illustrator days, when I openend a pdf with Calibri, text in this font was completely illegible.

https://community.adobe.com/t5/illustrator-discussions/illustrator-opens-pdf-with-corrupted-characters/td-p/4851202

Interesting. That is quite frustrating as calibri is the most commonly used font type (MS Word Defult). I don't think I could get my work colleges to consistently use a different font.

 

I did try export and open the same document above with a arial font type, no issues.

[TomHayes]

15 hours ago, Hens said:

Next question would be;
Is it opened on the same pc or another one?
Or even a different OS?
There have been reports that if the fonts are different versions per pc or OS it may give different results.

Same PC but when I perform the same operations on a different PC samer result.

Both Windows.

[kenmcd]

First, there is nothing wrong with the Calibri font. It is a 100% valid to the OpenType spec font. But it (and a few other MS fonts) is constructed differently than the typical Adopy fonts (which also have their own quirks). Based on the link above, IL appears to have its own import faults - like the APub import faults we are discussing here.

When exporting to PDF from APub with font sub-setting On, the ToUnicode table is not created correctly at all (total mess). So importing does not work at all. Note: some posters have said this worked in their PDF but I do not know why - and the ones I have looked at are bad. 

When exporting to PDF from APub with font sub-setting Off, the ToUnicode table is mostly correct - except for things like OpenType ligatures where the replacement glyph does not have a Unicode code point (such as ti). That glyph will not get a correct code(s) in the ToUnicode table. Based on some PDFs posted here by other users, Adopy apps will put the codes for both characters (t and i) in the ToUnicode table. So then the editing app has both characters and the font may again substitute the ligature glyph. 

When APub opens a PDF for editing it appears it is looking at the ToUnicode table to get the character codes. And since some of those are bad or missing in a PDF created in APub - you get the odd characters. 

Important points - PDF is really old, supported all kinds of devices, and predates Unicode. The ToUnicode table is actually optional. Inside the PDF the characters are kept track of using the character ID (CID). With the encoding info, and known character maps of that info, and the CID ... you can translate the characters into Unicode code points.

I assume this is what Acrobat (and Xara) are doing. That is how they are getting the correct characters from an APub created PDF which has bad/missing info in the ToUnicode table. Assuming the text is actually editable text and it is not just displaying the glyph (can you copy-and-paste into say a text editor?).

EDIT to be clear: PDF reader apps are just going to display what is there visually. That is what they are supposed to do. And when you open the PDF in a PDF editor, it too will just display what is there (the shape objects which look like characters). But in APub the text is imported as a text object, not a shape, and so the character codes attached to those objects become an issue.
Note: it is my understanding (from other posts here in the forum) that when a PDF is opened in Adopey Illustrator or Photoshop what you see is the shapes, not editable text. Being Adobe-Free for a few years now I cannot check this.

[TomHayes]

Thank you Libre for the comprehensive explanation. I am no expert, so much of the more technical side of things will go straight over my head. Dealing with fonts in the Affinity suit is one hiccup that I keep encountering. I have downloaded dozens of fonts from Google Fonts to be able to open PDF files that have embedded fonts that I do not have installed.

For example if I want to print a brochure from a PDF file, Adobe will always open the fonts without an issue but APub will almost always require me to download a new font. Is there a way to convert text to curves when opening a PDF in APub?

I realise this is not necessarily an issue with Affinity Publisher, but it is not as user friendly to open/merge/print PDF's as Adobe Acrobat etc. Affinity Publisher is probably "over powered" for the function I am using it for, but I can't justify purchasing Adobe Acrobat Reader for over $200 a year. 

 

[Catshill]

13 minutes ago, TomHayes said:

Thank you Libre for the comprehensive explanation. I am no expert, so much of the more technical side of things will go straight over my head. Dealing with fonts in the Affinity suit is one hiccup that I keep encountering. I have downloaded dozens of fonts from Google Fonts to be able to open PDF files that have embedded fonts that I do not have installed.

For example if I want to print a brochure from a PDF file, Adobe will always open the fonts without an issue but APub will almost always require me to download a new font. Is there a way to convert text to curves when opening a PDF in APub?

I realise this is not necessarily an issue with Affinity Publisher, but it is not as user friendly to open/merge/print PDF's as Adobe Acrobat etc. Affinity Publisher is probably "over powered" for the function I am using it for, but I can't justify purchasing Adobe Acrobat Reader for over $200 a year. 

 

Why don’t you use the passthrough option if you are looking to have a pdf within your document but unedited?

[TomHayes]

14 minutes ago, Catshill said:

Why don’t you use the passthrough option if you are looking to have a pdf within your document but unedited?

That is a good option, thank you for the suggestion.