(FROM PREVIOUS PAGE) In my opinion, they’re the most readable fonts I’ve ever seen on a screen.
I like them better than Verdana and Georgia, because they have a subtlety which just wasn’t possible to achieve when we had to work with whole pixels – a constraint from which ClearType freed us...
I use Calibri as my default font for reading and writing mail in Outlook.
Cambria is a terrific serif typeface for reading long passages of text. It also has a very full set of around 4000 Math characters, which we created in collaboration with the American Mathematical Society.
Candara is different enough that it gives a great look to text in both normal body text sizes and in larger sizes used for headings, like “Bill Hill’s Blog”, the masthead for all of these pages.
All of these fonts have their embedding bits set to Editable, which means that if you’re using Windows Vista, or you bought a copy of Microsoft Office 2007, or Mac Office 2008, you can freely embed them in your Web pages using EOT, and they can also be used for text input by your readers.
Why No New Webfont Pack?
People have asked me: “Why doesn’t Microsoft create another free font pack for the Web?” Well, I think the answer’s now pretty obvious.
There’s no need for another Webfont Pack with EOT. Embedding fonts in pages is much more reliable. With a Webfont pack, you still have to rely on propagation, which always takes time.
The Web needs a solution that doesn’t rely on either some limited set of fonts being available, or free fonts.
I know how much time and money goes into making a font that works well for reading on screen. And I’ve seen thousands of terrible TrueType fonts.
Microsoft has been willing to make large investments over the years, because more than half a billion people will end up using fonts we ship with Windows and Office. And we have focused on screen readability more than any other area.
I hope EOT really does become a standard. We’ve done everything we can think of to help make it one, and with any luck it will be a routine process for the W3C Working Group, and other Web browsers will also implement it.
We’re providing source code to make that as easy as possible.
_______________________________
In The Details... August 19 2008
I t’s very much harder to set text well than to set it badly. A huge amount of effort has to be put into getting the tiniest details correct.
Why do people go to all that trouble? Because badly-set text is just not acceptable; no-one will read it. It’s painful.
The complexities involved in setting type properly took hundreds of years to develop.
All the letters have to fit together exactly to create the familiar shape of a word. Some letters, like “T” and “o”, for instance, don’t fit naturally, and need special treatment (kerning).
Words need to be spaced at precise intervals on a line; lines need to be a certain length. Pages need to be a certain size.
“People go to all the trouble of setting text well because badly-set text is not acceptable. No-one will read it. It’s painful”
All of these parameters (and more) are entirely driven by human physiology.
Our eyes and brain together form a high-precision, rapid-scanning machine. When we read, it scans four targets per second, each 5-7 characters wide, and takes only 25ms to move between each target. It works at a resolution of 600 dots per inch.
It’s a machine with tolerances as tight as an internal combustion engine. If we provide it with text that’s not properly balanced, it can’t settle into the smooth rhythm we need to make the scanning part of reading occur below our conscious cognitive processes.
That part of our cognition has to remain free of any scanning load, so it can focus on the meaning of the text.
It’s an almost-magical process. We take it for granted, because we learned how to do it when we were five years old, and we’ve forgotten the magic moment when “dirty marks on shredded trees” first took us into another world.