(Beginner) Publisher: if the end result (PDF) looks decent, the process involved doesn't matter?

[Amy Choue]

Learning Affinity has been (still is) quite a struggle to me. Before Affinity, MS Word and the Internet were just about the only things I needed a computer for. I am now beginning to see how I might create a decent-looking PDF document ready for print, but not fully understanding the process involved. I am trying and doing a lot of things manually, hoping that I too will start to work smarter and not harder sometime in the future. 

 

Attached is an image from this book: https://www.amazon.com/Appearance-Reality-F-H-BRADLEY/dp/B000UO6N20/ref=sr_1_3?keywords=appearance+and+reality+bradley&qid=1693195180&sprefix=appearance+and+reality%2Caps%2C246&sr=8-3 

 

Customers have uploaded images that are filled with textual garbage. I don't know how such a thing can happen. And I know PDF files do not simply go haywire upoing printing like that. But part of me is unsure. So my question is, once an Affinity file is exported into PDF, there is nothing that can change the PDF (unless of course it is opened in a program and edited)? As long as the end result (the PDF) looks decent and professional, I can rest assured that it will stay the same? I think I am worried about this especially because I feel the way I am now working with Publisher is not really the best way. 

[firstdefence]

Hi @Amy Choue 

Are all the images that are causing problems in the SVG format?

...and can you upload a few SVG images for us to look at and also a sample PDF file and/or a sample AFPub file with a few of the problem pages included?

The reviews pretty much speak for themselves and are justifiable. There are some good hints to improve the books formatting.

[firstdefence]

After doing a bit of digging the SVGs all relate to Italicised Capital letters, these should be text not images

If you want a full transcript you can access it from: https://archive.org/stream/in.ernet.dli.2015.179941/2015.179941.Appearance-And-Reality_djvu.txt There are textual errors such as misinterpretation from the OCR scanner but I would say 99.99% of the text is correct.

[Amy Choue]

34 minutes ago, firstdefence said:

Hi @Amy Choue 

Are all the images that are causing problems in the SVG format?

...and can you upload a few SVG images for us to look at and also a sample PDF file and/or a sample AFPub file with a few of the problem pages included?

The reviews pretty much speak for themselves and there are some good hints to improve the books formatting.

Oh that book is not what I created. I have not created a PDF yet! (haha). I am really a beginner and learning Affinity with much difficulty. What I knew about PDF was that it is itself a sort of an image file and and that it retains the original formatting, that it is not susceptible to distortion, the way Word document is? But what I thought I knew about things of this sort so often turned out to be not quite true. And I was getting concerned about whether PDFs I create might have problems of the kind the image shows. 

[Amy Choue]

7 minutes ago, firstdefence said:

After doing a bit of digging the SVGs all relate to Italicised Capital letters, these should be text not images

If you want a full transcript you can access it from: https://archive.org/stream/in.ernet.dli.2015.179941/2015.179941.Appearance-And-Reality_djvu.txt

The person who created the book may not have uploaded a PDF, I guess. He might have used a Word file, or some other form of file that can easily be broken down like that. 

[firstdefence]

MS Word formatted text comes with it's own issues and the general rule of thumb is to process text coming from MS Word through a text editor and save it as plain text to remove any formatting word has applied. At least with the link to the text you have a good chance to eliminate all of the SVG garbage. 

I would certainly advocate reading through the feedback/reviews on Amazon and actioning the noted errors and lack of formatting. It will take a good while to do this but better good reviews than poor.

Effort = Reward.

[firstdefence]

I would break the book up into chapters, so each chapter is a separate afpub file, then use the Book option in Affinity Publisher to compile a book using the chapter files.

https://affinity.help/publisher2/English.lproj/pages/Advanced/aboutBooks.html

[thomaso]

4 hours ago, Amy Choue said:

I am trying and doing a lot of things manually, (…)

But part of me is unsure. So my question is, once an Affinity file is exported into PDF, there is nothing that can change the PDF (unless of course it is opened in a program and edited)? As long as the end result (the PDF) looks decent and professional, I can rest assured that it will stay the same?

Different to your and usually every manual approach in Affinity this book seems to be a result of a rather automatic process (starting with OCR), not human checked before upload … and before purchase, too, as the available preview already shows the issues.

But on the other hand, even with a manually created layout in Affinity unexpected text results may happen on export or print, then usually as an issue caused by certain font files. There are some threads in this forum reporting such results for PDF export, showing kind of text 'garbage' / odd characters. Many years ago (before Affinity) I experienced myself a print result of a 'healthy' PDF that showed certain characters replaced by unwanted glyphs, as far I remember all "r" and "i", while the according font was a client's corporate font file I had used without issues before and after that experience, so I assume an issue on the printer side in their prepress process, maybe caused by an outdated RIP.

1 hour ago, firstdefence said:

the SVGs all relate to Italicised Capital letters.

Interestingly your link shows issues in its plain text that don't exist in the OP's example at the exactly same position within the text, and vice versa. For instance:

[Amy Choue]

3 minutes ago, thomaso said:

But on the other hand, even with a manually created layout in Affinity unexpected text results may happen on export or print, then usually as an issue caused by certain font files.

Ah yes, I remember hearing about this from several different places (I have finished 3 Publisher-related courses so far, at least one of them must have mentioned about this). I'll definitely have to check on this again (and again). There are so many things to consider I, a beginner, feel like screaming. 

[PaulEC]

Just a couple of thoughts: Firstly, there are a lot of rubbish books on KDP and Kindle that range from "just" badly formatted to almost unreadable, because whoever created them did no manual editing or correcting whatsoever. Secondly, always get an author copy of your own books, and if there is a problem update the book ASAP.

[firstdefence]

33 minutes ago, PaulEC said:

Just a couple of thoughts: Firstly, there are a lot of rubbish books on KDP and Kindle that range from "just" badly formatted to almost unreadable, because whoever created them did no manual editing or correcting whatsoever. Secondly, always get an author copy of your own books, and if there is a problem update the book ASAP.

I don't know how they get away with it, surely there are standards on KDP or have they just gone?

[PaulEC]

20 minutes ago, firstdefence said:

I don't know how they get away with it, surely there are standards on KDP or have they just gone?

I think the problem is that there are so many KDP books published that there is no way for Amazon to quality check them all. As long as they meet the technical specs (size, bleed etc) they get published. (Even if they are virtually unreadable!)

[firstdefence]

5 minutes ago, PaulEC said:

I think the problem is that there are so many KDP books published that there is no way for Amazon to quality check them all. As long as they meet the technical specs (size, bleed etc) they get published. (Even if they are virtually unreadable!)

That's madness, can you imagine after another 100 years having to sift through all that junk to get to a coherent version. I hate to say this but, it looks like AI will have to be utilised to scan through this digital publishing junkyard and cull everything below a certain standard.

[thomaso]

41 minutes ago, firstdefence said:

I don't know how they get away with it, surely there are standards on KDP or have they just gone?

8 minutes ago, firstdefence said:

it looks like AI will have to be utilised to scan through this digital publishing junkyard and cull everything below a certain standard.

Hardly. How should a preflight / algorithm detect text like "SPECIAL_IMAGE" or "….svg-REPLACE_ME" as an unwanted issue, not as parts in a coding manual for instance?

Not to mention content, grammar or typo check that might fail with slang and poetry or books like Finnegan's Wake.

[firstdefence]

16 minutes ago, thomaso said:

How should a preflight / algorithm detect text like "SPECIAL_IMAGE" or "….svg.REPLACE_ME" as an unwanted issue, not as parts in a coding manual for instance?

Not to mention content, grammar or typo check that might fail with slang and poetry or books like Finnegans Wake.

I'm not talking about trivial typo's, not even grammatical errors nor slang, all of that could be dealt with by Ai learning what this text is and that they have alternative versions for the purpose of accent. Most typo's and grammar to a lesser extent can be picked up by the apps or add-ons/plugins that are used to format the text, but  "SPECIAL_IMAGE" or "….svg.REPLACE_ME" is just laziness, exporting to a PDF would surely have exposed this or checking formatting for the first chapter. The problem is this is a big book and going through it page by page would be a tedious task but looking for a decent OCR of the text is that daunting as this proves: https://archive.org/stream/in.ernet.dli.2015.6649/2015.6649.Appearance-And-Reality_djvu.txt 

It's just breaking it down into it's constituent prefaces, forewords, dedications, chapters and references etc. No offence to Affinity Publisher but I'd probably use Scrivener for this task. 

I really can't believe someone even thought they could sell a book with so much corruption in it and it ain't cheap either.

[thomaso]

44 minutes ago, firstdefence said:

all of that could be dealt with by Ai learning what this text is and that they have alternative versions for the purpose of accent.

No. Since human language may be used with intensive flexibility there is no way for an AI to distinguish errors from purpose, it just can guess or estimate but not judge reliable. Hw shd n knw, prf r jdg f ths sntnc ws tpd n prps? Or how about the well known fish | ghoti example?

As long applications used by humans require options like "Ignore", "Learn", "Define" or "Guess" an AI can't do any better work and not filter coincidence from desire.

There are many books of various generations and decades about absurdities in human language – but they don't lead to less absurdities but rather to less rules in using the language … and thus, from their author's perspective, rather lead to more absurdities.

Interestingly till nowadays there is no human or AI able to explain or translate the text in the ancient "Voynich manuscript" though it is investigated since ages by hundreds of researchers.

[firstdefence]

Yes I've seen the Voynich Manuscript.

So for now, we have an extensive library of mediocre > poor literature even when that literature is rehashed older work, the quality of storage is at best lacking.

[R C-R]

3 hours ago, firstdefence said:

So for now, we have an extensive library of mediocre > poor literature even when that literature is rehashed older work, the quality of storage is at best lacking.

From https://en.wikipedia.org/wiki/Sturgeon's_law & related topics it would appear this is nothing new, nor does it have much to do with the language used.