Chinese Version

Converting AFP to PDF with speed, accuracy, and robustness

Competitive advantages
  • Ultra-fast speed
  • Preservation of text, graphics, and images
  • Minimized data duplication
  • Sequential reading
  • Progressive writing
  • Uses .NET standard classes such as Image, Font, GraphicsPath
  • Uses Unicode for all text content
  • Standard Zip, Jpeg, CCITT, AHx compressions
  • Standard PDF features such as bookmarks and encryption
Outstanding features
  • Support for images of all pixel formats
  • Built-in, loss-less PDF direct import
  • Arbitrary mathematical transformation of contents
  • Subsetting of Type-1 and OpenType fonts
  • Downsampling of high-definition images
  • Coalesced Type-3 raster fonts
  • Outline-based Type-3 fonts
  • Pattern brushes from IOCA images and GOCA fills
  • Parsing and interpretation of outline fonts
  • Unicode values preserved for text as CMaps

The conversion from AFP to PDF has always been one of the most significant document transformations. Not only existing workflows involving AFP are healthily growing while print shops are welcoming PDFs then ever, but legacy archives are also being modernized for web access or long-term archiving in the form of PDF, which has been an ISO standard for multiple industries.

Our AFP to PDF conversion facility does the job with a top-notch speed, typically ranging between 3,000 to 10,000 pages per minute depending on the complexity of the AFP input. In addition to flexibility in output control, such as generating multiple PDFs in sequential or parallel, splitting the document at boundary of page group, etc., many advanced PDF features are also provided, such as encryption of file using information read from the page group properties, subsetting of fonts and downsampling of images for optimized PDF output size, and more.

Font handling has been a very strong aspects of our AFP conversion. Unlike most other similar products, there is no need for so-called font mapping, i.e. forcing the user to choose a font face and style for each and every input fonts, be it a bitmap font, Type-1 outline font, TrueType font, CID font, etc. In real life, an AFP font may also be embedded in an object container that is sent to printer hardware, or it can be referring to a disk file at a certain location. The whole thing about font mapping can be very unproductive and unfairly demands technical knowledge from a user.


Our program does all the tedious job automatically, typically requiring zero configuration. Fonts of different types are automatically converted into proper PDF fonts, and with some unique creativity. For example, raster CJK font is reorganized into dynamic Type-3 fonts that are effectively subsets; a CID Type-1 font family is converted into outlined Type-3 fonts; an external Unicode TrueType font is automatically subset to drastically reduce size. Similar expertise and innovation are also seen in image handling and graphics handling. As the system architecture is purely object-oriented, the AFP/PDF types are provided in their entirety as classes to ensure smooth and flexible integration with a calling environment, rather than a rigid black box to be invoked.

You can download a demo program (screenshots below) to experience the performance and output quality. The interface needs little explanation; you can add file or drop a file in the window, press start and watch the progresses change. One thing you can deduce from these images is: it is a multi-thread situation. Another thing is also prominent: that we do things quite unlike everyone else, yet it indeed makes sense.

There is usually not a premade software suitable for day-to-day use in all cases; in fact, each individual solution can be quite different than another, depending on the flavors — data composition style, usage of resources, specific requirements, etc. Please drop us a line to query about how a custom solution can be made. In some cases, you can consider the more controllable DocEvents Processing framework.



About the Author: Cyphia