Quality Printing on the Web

Steve Zilles
Manager of Standards
Adobe Systems Incorporated
April 23, 1996

Adobe Systems Incorporated is a company with a more than 10 year history of providing technology for high quality printing. The Adobe PostScript langauge has been the standard for high quality page descriptions, both for computer based printing and for newspaper, publication and commercial printer. Adobe has been a prime supplier of high quality, scalable fonts for publishing world wide. With the creation of the Portable Document Format (PDF) Adobe extended the power and quality of PostScript to displays as well as printing.

There are many issues involved with printing documents available on the Web. In this statement, I only wish to address two of the major issues: a media type for resolution independent graphcs and fonts. Adobe Systems believes HTML and the Portable Document Format (PDF) are complementary. A good example of the synergistic relationship possible between HTML and PDF is using PDF for graphical objects within HTML documents. The case for PDF as a media type for graphics in HTML is presented in the second section of this paper. Immediately below is Adobe's view of the goals for extensions to HTML to support a broad range of fonts.

Goals for a Web Font Solution

Rather than attempt to present a solution at this time, I believe that it is first important to agree upon what the goals of this solution should be. Much time could be wasted discsussing details if the particpants are trying to get to different final results. In that view, I present the following list as a (partial) set of goals that should be satisfied by a Web Font Mechanism/Format:
  1. [single doc] Support a single document with possibly many fonts used within it.
  2. [multi doc] Support a collections of documents/pages (both in HTML and in PDF) that use the same set of fonts. This may be a book, or a subsite or a collections of sites
  3. [rapid display] Provide rapid display of the text in a document even when a number of fonts are involved. (This could involve temporary font substitutions which are replaced when the real font data arrives. Adobe calls this progressive rendering. This could involve font subsetting and font embedding.)
  4. [Ease of Authoring] Creation of documents using Webfonts should be as easy as creating documents for any other environment.
  5. [authoring compatability] Document creators will not need to acquire new sets of fonts for Web documents.
  6. [Worldwide support] The Webfont technology should support glyphs for the full Unicode character code repetoire and should be able of supporting glyphs from any source.
  7. [Rendering quality] Provide a rendering quality that matchs current state-of-the-art font rendering (or better) and provides faithful reproduction of the user's fonts (font fidelity).
  8. [Better than GIF] The results of type produced with Webfonts and HTML should be as good as or better than type rendered into GIF files; that is, designers should not need to create a GIF file just to do good typography.
  9. [Request only missing data] Request fonts from a server only when the needed font data is not available local to the browser. (The font data may be installed on the platform on which the browser is running or the font data may be cached from some previous (partial) download.)
  10. [Client controllable] Much like the ability to turn off image retrieval when displaying HTML files, the user of the browser should be able to turn off font retreival and the browser will make best possible substitutions for the referenced fonts.
  11. [Minimal size] Reduce amount of font data that needs to be transfered for a given document/page to a small a size as possible.
  12. [Ease of use] The Webfont technology should not make excessive resource demands (such as large cache sizes) on the browser platform; it should be a lightweight solution.
  13. [IP rights] Protect the IP rights of the font vendors.
  14. [Time to market] The chosen Webfont technology should be deliverable within this year.
  15. [HTML syntax] Define how fonts are referenced from an HTML document. Where can these font references be placed. How do font references work with the <FONT> element and CSS-1 Style sheets?

Portable Document Format (PDF): A Media Type for Graphics

HTML is a powerful format for text that requires equally powerful formats for representing graphical and pictorial information. There is a positive, synergistic opportunity to combine the graphical representation capability of the Portable Document Format (PDF) with HTML to represent visually rich content. HTML provides flexible text; PDF provides powerful graphics.

PDF is an open, publicly available specification

The Portable Document Format (PDF) was developed by Adobe Systems Incorporated specifically to aid in the exchange and distribution of electronic or digital documents (or fragments thereof) across a wide range of platforms and distribution technologies. The PDF representation of a document is independent of application software, hardware, and operating system used to create that electronic document. Thus, PDF is intended to provide interchange of already formatted content, such as graphics, in the same environments in which HTML provides interchange of revisable, retargetable text.

PDF has been an open, published, publicly available specification from its original annoucement. Both PDF producers and PDF consumers (viewers or printers) have been implemented by multiple, independent organizations. The ANSI Committee on Graphics Arts Technologies Standards (CGATS) has announced the creation of an ANSI (and likely ISO) standard for interchanging digital pages using PDF. Thus PDF is a suitable format for Web; it already has a MIME type, application/pdf.

PDF provides visually rich, design controlled content

Controlling the appearance of your documents on the Web is a challenge. With PDF, however, you can acheive the highest quality and greatest control to preserve the integrity of your design -- including scalable typefaces and graphics. PDF is designed to capture the final formatting decisions, the positioning of elements, and faithfully reproduce them, both when displayed on a screen and when printed.

Print Ready content

PDF is derived from the well established PostScript (tm) Language. It has (or will have with the release of the final Adobe Acrobat (tm) Amber products) the full power of the level 2 PostScript Langauge. Therefore, there is a straightforward translation from a PDF file into a PostScript file providing access to both quality computer printers and to reproduction using the imagesetters to prepare plates for printing presses. (Acrobat products also are capable of printing on a range of non-PostScript printers.)

Ease of Authoring

PDF files are easy to create. Your existing tools for document creation already support the creation of PDF files. For many Windows and MacIntosh applications, creating a PDF file is as simple as printing the document. Other applications provide "SaveAs" PDF. For applications that produce their output in the PostScript Language, there is the Adobe Acrobat Distiller that converts the PostScript to PDF, compressing it and allowing image downsampling to further reduce the resulting file size. Finally, for documents available only in paper form, the so called legacy documents, there is Adobe Acrobat Capture to do text recognition and to create a PDF form of the document that looks just like the scanned page. Since text is represented in a form that allows extracting the character content, these PDF documents, no matter how created, can also be indexed. In short, if you can (or did) print it, you can PDF it.

PDF is an object-oriented, device independent representation of the document. Text is represented as strings that are rendered using high quality, typograhic scalable fonts. These fonts are identified with enough information to allow quality substitutions to be performed if the font is not available at the receiving site or the real font data can be embeded in the font for printing and viewing. Both European and non-European (Asian) fonts are supported. The Adobe Acrobat Amber viewers provide anti-aliased display of these fonts.

Graphics are respresented using the time proven PostScript imaging model. The PDF language has the full capability of PostScript, already used to represent most sophisticated printed graphics. PDF also has the wide range of PostScript Level 2 image formats. This includes JPEG and CCITT compression with accurate, device independent colors. This allows the same image to be used for display on the screen (by subsampling) and for printing (because the image data is scaled as needed).

PDF goes beyond PostScript in a number of ways. The most important of these for the Web is the inclusion of URL (and non-URL) based hyperlinks which allow PDF files to intimately linked with HTML files and other Web document. In addition, PDF files can be optimized for rapid page by page display in a Web browser.

Ease of use

PDF support on the Web is easily available. Adobe provides a free Acrobat Reader on a number of platforms (Mac's, Windows, a set of Unix platforms, and more). This reader can be downloaded from www.adobe.com. The reader is enable to work with standard web browsers to allow display of PDF documents in the browsers window or within frames within that window. Thus PDF is a natural format to combine with HTML to handle the graphics/image intensive or controlled design portions of a document that HTML cannot do.

Summary: HTML plus PDF is more powerful than either alone

HTML was designed to allow optimal formatting for each different presentation enviroment; for example, different formatting for display windows and print files. PDF was designed to faithfully capture the result of a formatting process for both display and printing. PDF was also designed to have the proven power of the PostScript Language graphics capability.Together, HTML and PDF allow a document designer to produce document designs that include powerful graphics and make appropriate trade-offs between flexibility and design control. Since PDF is device and resolution independent, the same file can be suitable for both screen display and printing. It is an ideal choice for a media type to embed in and augment HTML documents.

Adobe, Acrobat and PostScript are trademarks of Adobe Systems Incorporated.