A PDF is a PDF is a PDF... Right?

[#1]

Not quite...

There are not only different levels or versions of PDF, but there are also standards of PDF which are required for different consumers of the document. Each have their own specific requirements and each present their own challenges. I'm going to have a look at some of the requirements of PDF here and how using Arbortext products allow you to meet them.

Lets start by looking at why you want to create a PDF in the first place.

Some people are happy with just a plain old vanilla PDF. All they want is something that looks like a printed page so they can read it on their computer, stick it on a website or make paper copies. They're not too bothered about colour reproduction, download speeds, storage or accessibility. As long as it doesn't look rubbish, they're happy. That is fine. Nothing wrong with it at all. PDF is more than happy to give you this.

But...

• What if you are bothered that the expensive glossy catalogue you've blown your marketing budget on looks as good as it can?
• What if you want to make sure people with disabilities can read and appreciate your legislative information as much as anyone else?
• What if you want to make sure the PDF you've created can be opened again in 5, 10 or 20 years time and still be readable?
• What if you don't want your readers to wait 3-4 minutes to start reading a PDF online while the whole thing downloads?

These are some of the reasons why there are different standards of PDF and why we're working hard to allow you to create them. This is also why PDF has a collection of features which allow you to create output which suits your consumers' needs. Features such as:

• Font embedding - If you include the font in a PDF, you can ensure that it can be reproduced accurately later on even if the font doesn't exist on the output system. If you include the whole font, then you can allow the PDF to be edited too (not everyone's cup of tea, but still a useful tool)
• XMP metadata - Effectively embeds an XML instance containing metadata into the PDF
• Colour - Using RGB colour definitions for print, which uses (at least) four colours (CMYK) leads to problems translating colours from one colour space to another. Providing colour profiles (for RGB and CMYK), which give the colour values context, also ensures accurate colour reproduction
• Tagging - As I discussed in my Styler 6.0 M020 Accessibility posting, PDF tagging gives the content of the PDF structure and ensures readability by assistive technologies.
• Content boxes - PDF has a set of boxes which define different dimensions of the document. These boxes allow accurate placement of the page on high-end output devices and provide information on how much bleed to allow off the edge of the page.
• Web-optimised PDF front-loads the PDF with the information it needs to display the first page so it can show this while the rest is downloading. You can also reduce the resolution of images to ensure the PDF downloads quicker.

As PDF has evolved, different versions of PDF have been released, each typically tied to a version of Adobe Acrobat. The versions start at PDF 1.0 and progress through PDF 1.7. Different features are added as you progress through the versions so you need to be sure that you're outputting to the right version to include the feature you need. For example, you can't embed XMP metadata before version 1.4 and 3D graphics aren't supported before version 1.6.

To help people work with with these features and to ensure customer and supplier are talking the same language, a number of PDF standards are available. The two main ones are PDF/X and PDF/A, both ISO standards. There are other standards, but these are the two key ones to be aware of:

• PDF/X is there to allow exchange of graphical information in PDF, mostly for professional reproduction. There are several versions of PDF/X, each with different levels of support for handling colour information, fonts and so on.
• PDF/A is mostly an archiving standard which requires that the PDF is self-contained and doesn't require additional files or resources to be readable in years to come.

Each of these standards use (or actively disallow) different PDF features to ensure they work well in their intended use. It is always worth talking with those who will receive your PDFs to establish whether they'd like to receive it in one of these standards.

So what are we, PTC, doing about it?

APP Version 11 (which arrived in June) provided a significant upgrade to its PDF output capabilities and included:

• Source TrueType font embedding
• Tagged PDF including custom tagging and role maps. PDF tagging has already been exposed in Styler's 6.0 M020 release.
• Content bounding boxes boxes output by default
• Web-optimised PDF
• XMP metadata. We're also planning to allow custom XMP rather than just document property based XMP
• The ability to select different versions of PDF from 1.3 to 1.7

The majority of these will be applied by default to PDF output from APP, including from Styler. The M010 release of APP's Version 11 (currently scheduled for December 2011) will also provide support for colour profiles which will then allow us, in the same release, to allow users to select the PDF/X or PDF/A standard they wish to output to. Once the APP functionality is in place, we will work on exposing it through Styler and PE.

Long-term, the APP engine will provide us with a great platform to extend our PDF support to include multimedia, 3D objects, interactivity, layers and so on. I would be very interested to hear your opinions on these features.

I hope the information here whets your appetite to learn more about PDF and encourages you to explore how you can produce better output using Arbortext tools. If you have any questions or have other PDF requirements, please feel free to get in touch.

Obviously, any forwards looking information here may change without notice.

[DmitryTsarev]

Great information, Simon and just in right time. I was recently looking for the short and simple overview of different PDF standards.

And what's about embedding 3D in PDF published with Publishing Engine? Can you share any information on the functionality? Is it planned for the upcoming releases?

[SimonTaylor]

3D publishing in PDF is on the plan. Our PDF library can accept U3D objects so we will build on that. The objective is to provide a similar experience to HTML.