David Lawley wrote

>I notice that with some typefaces,
>for example Bliss, the 'Cap height' and
>'Ascender height' are different, is there
>a reason for doing this, or is it just
>a look thing?


It's rarely just a look thing. As Hrant says, readability is usually paramount.

But ascenders can be exactly the same height as cap H, successfully. As long as the stem weights of the caps (like H, with its vertical straights) are properly balanced with the (slightly thinner) weight of the lowercase letters with vertical stems, it can work fine.

The same applies to italics and obliques.

For a very interesting, decades old and quite delightful experiment with ascender/descender variance, look at the typeface series called 'Parsons'.

Try to find an original cut, or an original type specimen or promotional booklet from the founder, showing all the designed alternates.

It's a lovely thing, really.

Joe

> ascenders can be exactly the same height as cap H, successfully. As long as the stem weights ....

I don't see it as a weight issue - instead it's a matter of proportions: caps generally fill most of the rectangle defined by their width and height, while lc characters fill the x-height or at most the x-height and a stick to the ascender height (except the wonderful "f"). So the only way you can safely match the cap and ascender heights (in a text face) is if: the lc x-height is very large and/or the caps are very narrow. But the former is damaging to readability while the latter is usually aesthetically fatal.

> ascender/descender variance

Such variance is sady under-rated, in this Modernist era of regularity. ATF Garamond (AKA Garamond #3) has some amazing ascender variance, even though aesthetically I wouldn't touch it with a 10-foot composing stick.

hhp

If you make the caps as high as the ascenders, in real text (as opposed to in something like a poster where lining things up might be great) they look too big, reducing readability.

hhp