OnScreen Hinted Typeface - help needed for design guidelines

yennok's picture

Hi all,

I'm about to design a Hebrew on-screen typeface, which later on gonna be hinted.
lately I've been researching the technical parts, and think i quite got it well,

but my main questions are design perspective:
1. Before designing the typeface, are there any limits I need to take in account? (White spaces? open glyphs?)
2. What's the difference between the process of designing a typeface for print and for screen?
3. how free-form can the glyph be? or should it be very mathematical?

Thanks,
Yennok.

hrant's picture

{To Follow}

Té Rowan's picture

Can I assume without making ASS of U and ME that you have already trawled through the relevant threads in the Special Interest Group on Hebrew Typography and Type Design??

yennok's picture

The purpose I wrote it here is because I Want to know if there are any design guidelines when considering a screen font. something like written in this discussion: http://www.typophile.com/node/18685.

The "Hebrew" part is less important. Actually I was searching on that forum but nobody did Hebrew hinted as far as I know.

It would be great if someone has any conclusions about the issue...

John Hudson's picture

When designing for screen, it helps a lot if you have a target pixel-per-em (ppem) size or range of sizes in mind. This is especially true today, given the prevalence of asymmetric rendering, i.e. systems such as ClearType that apply different different grid resolutions in the x and y directions. If you can target a particular ppem size, you can design your glyphs so that their proportions naturally give the kind of onscreen shapes that you want at that size (generally providing positive results for adjacent sizes too). You will inevitably face the frustration of asymmetric rendering as the same outlines look less good as you move away from that target size before looking good again as they approach the next 'magic' size at which the proportions just work. [The only way to avoid this is to produce more than one design, targeting different ppem size ranges.]

hrant's picture

Or superhint.

hhp

Frode Bo Helland's picture

Nono, Hrant.

hrant's picture

Because it's too much work, or because it's no longer technically robust?

hhp

John Hudson's picture

Much depends on what you mean by 'superhint'. If you mean something like heavily delta'd instructions controlling precisely the pixel appearance of the type at individual sizes, then I think it is obvious that the diversity of rendering methods now in play make this technically impossible, especially since some of these rendering methods ignore some or all of the hint instruction set most of the time. Beat Stamm has suggested what I think is a different kind of superhinting: programming in the TT instruction language that allows for dependent hints targeting different environments. Even in that case, though, I think the reality is that design for screen has to begin at the outline level, and isn't something that can be applied post-facto in hinting as was the case for binary bitmap display. The days when Times New Roman could be turned into a credible screen reading font with hinting are long gone. The push to enhance individual typeface fidelity via antialiasing, to represent text type on screen recognisable as individual type designs rather than as highly readable but generic brick patterns, necessarily implies that the qualities that make for good screen readability must reside in the design. That was the genius of Verdana and Georgia and why they continue to work so well across so many different rendering technologies, and why that approach disitinguishes the best work being done in UI and webfont design.

Syndicate content Syndicate content