Typographic Composition - Compose Yourself

[Colour Management]
[Free Stock Photos]
[InDesign CS2]
[QuarkXPress 6.5]

[Acrobat 7]
[Creative Suite 2]
[Fontographer 4.1]
[FrameMaker 7]
[InDesign CS2]
[PageMaker 7]
[PagePlus 10]
[Publisher 2003]
[QuarkXPress 6.5]
[Ventura 10]

[Colour Management]
[Free Stock Photos]
[CD Authoring]
[Mac/PC History]
[Creative Suite]
[Acrobat Directions]
[Font Formats]
[DTP Tagging ]
[Data-Driven Design]
[Windows XP]
[Paper to PDF]
[Page Imposition]
[eBook / ePaper]
[Removable Storage]
[Understand Halftones]
[XML Revolution]
[PDF Workflows]
[Text Typography]
[Spot Colour]
[Database Publishing]
[Layout Compositing]
[Design Automation]
[Stock Photography]
[Quark v Adobe]
[Asymmetric Grids]
[Commercial Print]
[Colour Print]
[Preparing Print]
[Understanding EPS]
[Layout Grids]
[Beyond CYMK]
[Acrobat v Immedia]
[Postscript 3]

[Acrobat Tutorial]
[PageMaker Tutorial]


[Home / What's New]
[DTP / Publishing]
[Vector Drawing]
[Bitmap / Photo Editing]
[Web Design]
[3D Software]
[All Reviews]
[All Articles / Tutorials]
[Book Search / Shop]
[Site Map / Search]

you can help support the site with a direct donation or by shopping via the following links:


Thank you!



Good Typographic Composition

Tom Arah shows you the secrets of good typographic composition.


The foundation of good design is the underlying typography which determines the appearance of your text. Much of the art of typography is concerned with typefaces and the shape of the individual characters of your fonts. Less recognized but just as important is the way that these characters are put together to produce the words and lines that make up your text. In fact this text composition is in many ways more important as it affects readability directly and so can make the difference between text that gets read and text that doesn't.

Composition is crucial, but the basics are very straightforward. As you type on a typewriter, for example, the platen simply moves a fixed distance to make room for the next letter. Each character in the font is the same width, either 10-pitch or 12-pitch which means ten or twelve characters per inch, so there are a set number of characters possible per line. As you reach the end of the line you work out whether the next word will fit on the line and, if not, hit carriage return to begin a new line.

It might be simple but there's a major problem -- the results are typographically horrendous. To begin with there are no options for vertically varying point size, but far worse is the fact that each letter, or glyph, must be the same width horizontally so that an "i" is forced to take up the same width as an "m". The typewriter fonts, such as Courier and American Typewriter, have been specially designed to minimize this manhandling but the results are still unappealing. Worse, the monospacing messes up the natural shape of the words and it is these that the eye recognizes when reading. The end result is that typewritten text is limited, ugly and, worst of all, difficult to read - the cardinal sin of good design.

Typewriter v typeset composition

There is a huge difference between the typographical quality and complexity of ragged-right monospaced text and justified proportional text.

The obvious solution is to use proportional fonts where each character can be a different width to the next. This immediately opens up the vast range of available typefaces and point sizes to designers and frees them to make their designs as attractive and readable as possible. However, using proportional fonts creates a new problem. Without monospacing the number of characters that can fit on a line is no longer fixed but depends on exactly which letters are used. This makes it more difficult to judge whether the last word will fit on the line.

In fact the problem is more complicated than this. With a traditional typewriter there is no option but to have "ragged-right" text where line lengths vary depending on the number of characters in the last word of one line and the first word of the next. Unfortunately this can look messy and, particularly in multi-column layouts, the professional designer needs clearly defined text blocks where all lines are exactly the same length to produce a neat, finished appearance. But how can you produce such "justified" text when every line is an arbitrary width dependent on the characters that make it up?

The answer is variable spacing. Because the eye is actually reading word shapes it is comparatively uninterested in the space between words. By expanding or shrinking these spaces it is possible to spread out or compress a line to fit the desired width without it disturbing the reading flow. If the spaces do become too big, creating distracting "rivers" in the text, there is also the second option of slightly increasing the spacing between letters - though as this affects word shape it is far less desirable.

Justification by spacing

Lines of text are justified by varying word and letter spacing according to customizable parameters.

Such variable spacing is the only way in which justified paragraphs can be produced and although it might seem like a simple extension of ragged-right alignment it actually moves text composition into a completely new arena. To begin with typewriter-style composition is essentially word-focused whereas justified text demands a wider line-based approach. More importantly, where ragged-right monospaced composition is purely sequential and pre-determined, justified proportional composition introduces variables and decision-making. In particular there are two key alternatives to consider: to squeeze on another word and tighten spacing or to drop the word and spread spacing. In other words justification introduces the whole question of aesthetics to composition.

The process of justifying text produces far more complexity but it is thanks to its inherently non-sequential nature that publishing as we know it became possible. Johann Gutenberg's solution to the problem back in the mid 15th century was to cast letters and varying sizes of spaces - most famously the "em" and "en" spaces we still refer to today - as individual pieces of metal that could be arranged to exactly fit the desired line length. Each line's worth of characters was arranged in a "stick" and multiple lines gathered into a "galley" separated by strips of "leading" and locked into a "chase" from which multiple pages could be printed.

Gutenberg's movable type system was the basis for all print for centuries and, in the hands of skilled craftsmen, the resulting typography became a true artform. Eventually though the process succumbed to computerization and the human aesthetic judgment gave way to number-crunching. The key to computerized justification is the dual use of font metrics, where the width of each letter in the typeface is built into the font, and spacing preferences, where the user can set the desired word and letter spacing and the acceptable maximum and minimum. Essentially the computer then adds up the character widths of each line plus and minus one word. It then works out how it would have to justify each option and makes its decision by choosing word spacing before letter spacing and compression before expansion. Having made the choice it then moves onto the next line.

This is the essential core engine that underlies computer based text composition, but for professional DTP work there are two other important complicating factors. To begin with there is a serious problem with default font metrics as these treat the characters in a font as if they were regular squares. In fact letterforms are very different and, as such, their ideal spacing isn't fixed. Instead the ideal spacing depends on what character the current character follows. With common combinations such as To, Tr, Ta, Yo, Ya, Wo and Wa, for example, the second letter should be tighter to the first and actually lie under the capital letter's overhang.


Kerning effectively varies a character's width depending on its shape and the following character's.

The solution is "kerning" where the second letter can be moved marginally nearer to the first, slightly overlapping its in-built metrics. This can usually be done manually, but clearly that's not practical for long sections of text. The real answer is to have the kerning information built into the font and nowadays all professional fonts contain a huge number of kerning pairs to control exact character placement depending on which letter follows which. Clearly the fact that each character width can effectively vary depending on its context has consequences for text composition, but looking up the kerning pairs and working out the new equivalent character and word widths is a one-off process and, once determined, is fixed.

The second factor that has to be taken into consideration is both more subtle and more problematic. Sometimes, especially when dealing with short lines containing long words, it's not possible to justify the line while keeping within the maximum and minimum acceptable word and letter spacing. If the word that straddles the end of the line is "supercalifragilisticexpialidocious", for example, we're in trouble. Squeezing the word in would demand tightening spacing until words and letters overlap, while leaving it to the next line will lead to huge gaps opening up. Both options are less easy to read and disrupt the "typographic colour" or density of the paragraph and impact on the texture of the overall design.

The obvious solution is to split the problem word, but how? You could split the word manually by picking a natural break and inserting a hyphen and a space but, if the paragraph then recomposes, your word could end up unnecessarily split in the middle of a line. Clearly the process must be left to the text composition engine and handled automatically - but how can a dumb computer know where to hyphenate? After all, as far as the computer is concerned each word is only a combination of character codes.

The solution engineered in the earliest DTP packages was a hyphenation algorithm. Essentially this attempted to break down words into recognizable syllables based on consonant and vowel combinations that could then be hyphenated. The problem is that English wasn't actually devised on this basis so plenty of the resulting hyphenation choices were wrong, while others such as "the-rapist" were legitimate but clearly undesirable. In the end it became apparent that hyphenation was actually more intelligently dealt with by a dumb look-up table and nowadays all DTP applications come with extensive hyphenation dictionaries, most of which can be customized by users.

The end result is far more complexity for the composition engine. Now when a line falls outside the optimal justification parameters the program must check the straddling word for possible hyphenation points, first against its user-defined dictionary, then against its main hyphenation dictionary and then, if all else fails, it must attempt to determine them algorithmically. It must then choose the best option from all possibilities based on the effect on spacing. With advanced hyphenation dictionaries it must also factor in preferred hyphenation points so that it's better to hyphenate supercallifragilisticexpialidocious after super- rather than supercalli- for example.

The introduction of hyphenation immediately complicates the composition process drastically but there's more to come. Rules have to be set within which the hyphenation process operates. Quark's H & J dialog is typical offering precise control over both desired spacing and hyphenation. In particular you can set the size of the smallest word eligible for hyphenation, the minimum number of letters that should be left before and after the hyphen, an option to ignore all capitalized words and a limit on the number of possible consecutive hyphens. These parameters can then be saved and reused as named H & J settings. If just one parameter is changed it can completely change the composition of the entire document.


Hyphenation is inextricably linked with justification and good text composition.

The automatic composition control offered by programs such as XPress and PageMaker is comprehensive, but it's not foolproof and you'll often need to go in manually to sort out problems. You'll want to make sure that names aren't hyphenated, for example, which you can usually do by inserting a discretionary hyphen before the word's first letter. You'll also want to prevent lines breaking between titles and surnames so that "Mr/Mrs/Ms" doesn't appear separated from its name. This can usually be done either with a hard, non-breaking space or by marking sections of text as non-breaking. More importantly you'll often find that you can improve the consistency and typographic colour of a paragraph with some judicious use of discretionary hyphens and hard spaces - though you'll have to be careful to check that this doesn't interfere with future editing.

All told the computerized text composition system is by no means perfect but has proved workable and the loss of absolute typographic quality compared to manual setting by a professional has been more than offset by the gains in design flexibility. More to the point it's all we've been offered, so of course we've accepted it. With the release of InDesign, however, Adobe has taken the opportunity of a fresh start to completely rewrite the rulebook and to set new typographic standards in the process.

At first sight nothing seems that different apart from the presence of new alignment options of four justified paragraphs in which the last line is aligned left, right, centred or fully. It might not sound that revolutionary but it does open up some nice design effects that were previously awkward to produce. The same is true of the introduction of a new "flush space". In all other paragraph alignments this character acts just like a normal space, but in fully justified paragraphs it soaks up all the unused space. If used before a special end of story marker this will automatically neatly align the marker flush with the right edge of the line.

Such changes are welcome but hardly push back the typographic boundaries. Dig a little deeper, however, and you begin to discover some serious advances. Call up InDesign's Justification dialog, for example, and you can set not just desired word and letter spacing but glyph scaling. This means that rather than just relying on varying spacing you can actually change the width of your characters to ensure that they fill the line! Clearly this is a potentially dangerous option with the possibility of radically different letter shapes from one line to the next. If you know what you are doing, however, you can quickly produce some striking special effects. Alternatively, by limiting the minimum and maximum scaling to 98% and 102%, you can cut down on undesirable letter spacing variations without the changes in letter shape becoming distinguishable to the untrained eye.

InDesign's glyph scaling

InDesign has introduced new typographic features such as glyph scaling and optical margin alignment.

In fact this move towards optically-based standards is typical of InDesign. Another of its innovations is Optical Margin Adjustment. This is set for the story as a whole rather than on an individual paragraph or style basis, and subtly alters the position of the last and first character on each line. This is most obvious with punctuation marks which are left hanging outside the main text block, but all other characters are also slightly affected. The end result is a text block that is actually slightly ragged left and right but to the eye seems much more flush.

InDesign's Optical Kerning feature also chooses visual appearance over in-built metrics. Essentially it overrides pair kerning and looks at the actual character shapes in letter combinations to determine ideal kerning on the fly. This might seem pretty pointless when most fonts now include extensive kerning information but it does have its advantages. To begin with it works with all fonts including those without any kerning pairs built in. More importantly it can work with combinations of different fonts and point-sizes which would throw off metric-based kerning. Finally it also takes account of the fact that text in larger point sizes, such as headings, usually looks better if it is more tightly spaced than body copy.

These are all important typographic advances but by far the most significant is a radical reworking of the core composition engine. As we have seen, with features like proportional font metrics, variable spacing, kerning and hyphenation, the typical DTP engine has come a long way from the crude, ragged monospaced composition that a typewriter is capable of. In one important aspect though nothing has changed. Both typewriter and DTP engine work on a line-by-line basis. In other words once the spacing of one line has been determined it's back to square one for the next.

With InDesign this single-line method is still available, but there's also a new multi-line composer on offer. This engine considers a whole network of possible break points over a range of lines and assigns them a weighted penalty according to how much they deviate from optimal spacing. It then chooses the option with the lowest total over the paragraph as a whole, or rather over the number of lines that you specify. The process can be slightly disconcerting as, if you edit an existing line, the lines above can suddenly rearrange themselves. In terms of end results, however, there's no comparison - the multi-line composer produces much more evenly spaced and more evenly coloured text.

Multiline composing

InDesign's multi-line composer acts like a human operator judging the typographic quality of the paragraph as a whole.

The results are excellent and, if you switch on InDesign's multiple shade highlighting of compositional problems, you'll see how the multi-line composer produces far fewer and far less serious violations than the single line composer - especially after the first few lines of a paragraph when the engine has more options to play with. The fact that any violations still exist, however, shows that the system isn't - and can never be - perfect. It can always be better, however, and InDesign's final typographic strength is its extensibility that means that you can plug-in new compositional engines as they are made available. In other words, in future we'll no longer just have to accept the typographical quality that the software developers have thought fit to provide, but will be able to demand the best available.

With features such as its new multi-line composer, it's not surprising that InDesign claims to set new typographic standards. In fact of course it is only setting new standards compared to previous mainstream DTP applications. After all the principle of multiple-line composition was already established back in the early 1980's by Donald Knuth in his ground-breaking public-domain typesetting application, TEX. More to the point, the vital breakthrough of trying out different options to see which one is more visually attractive has a very long pedigree - Gutenberg himself would approve.

Tom Arah

November 1999

Hopefully you've found the information you were looking for. For further information please click here.

For free trials and special offers please click the following recommended links:

For further information on the following design applications and subjects please click on the links below:

[3D], [3ds max], [Adobe], [Acrobat], [Cinema 4D], [Corel], [CorelDRAW], [Creative Suite], [Digital Image], [Dreamweaver], [Director], [Fireworks], [Flash], [FreeHand], [FrameMaker], [FrontPage], [GoLive], [Graphic Design], [HTML/CSS], [Illustrator], [InDesign], [Macromedia], [Macromedia Studio], [Microsoft], [NetObjects Fusion], [PageMaker], [Paint Shop Pro], [Painter], [Photo Editing], [PhotoImpact], [Photoshop], [Photoshop Elements], [Publisher], [QuarkXPress], [Web Design]

To continue your search on the designer-info.com site and beyond please use the Google and Amazon search boxes below:

Web designer-info.com

designer-info.com: independent, informed, intelligent, incisive, in-depth...

All the work on the site (over 250 reviews, over 100 articles and tutorials) has been written by me, Tom Arah It's also me who maintains the site, answers your emails etc. The site is very popular and from your feedback I know it's a useful resource - but it takes a lot to keep it up.

You can help keep the site running, independent and free by Bookmarking the site (if you don't you might never find it again), telling others about it and by coming back (new content is added every month). Even better you can make a donation eg $5 the typical cost of just one issue of a print magazine or buy anything via Amazon.com or Amazon.co.uk (now or next time you feel like shopping) using these links or the designer-info.com shop - it's a great way of quickly finding the best buys, it costs you nothing and I gain a small but much-appreciated commission.

Thanks very much, Tom Arah

[DTP/Publishing] [Vector Drawing] [Bitmap/Photo] [Web] [3D]
[Reviews/Archive] [Shop]  [Home/What's New]

Copyright 1995-2005, Tom Arah, Designer-Info.com. Please get in contact to let me know what you think about a particular piece or the site in general.