Glyphs going above ascent height/below descent.

Get help with FontCreator here. Please do not post feature requests or bug reports here.
Post Reply
David Webber
Posts: 44
Joined: Mon Aug 19, 2013 10:33 pm

Glyphs going above ascent height/below descent.

Post by David Webber »

If I include glyphs in a font which either descend well below the descent height, or ascend well above the ascent height, am I likely to run into trouble?

[I'll probably find out the hard way, but I ask because it would be nice, if necessary, to plan an alternative to what I'm doing before I invest too much time in it!]

The context is my Unicode font with latin text characters in the usual place, and music symbols in the private use area. A sample symbol is the treble clef:
treble clef
treble clef
TrebleClef.jpg (1.99 KiB) Viewed 5048 times
It is very convenient to have its base line at the position shown by the red line: where it intersects the note G on the staff. But that will mean that it descends well below the descenders of the lower case Latin characters. Can I get away with it?

David Webber
Bhikkhu Pesala
Top Typographer
Top Typographer
Posts: 9876
Joined: Tue Oct 29, 2002 5:28 am
Location: Seven Kings, London UK
Contact:

Re: Glyphs going above ascent height/below descent.

Post by Bhikkhu Pesala »

You may run into problems with the glyphs being clipped below WinDescent or above WinAscent. I suggest simply recalculating the metrics after designing the font as you want. The other glyphs will be smaller and the default line-spacing for single spaced text will be excessive for text, but I don't know of any other reliable method.

The font may work in some applications, but not others.

Take a look at my Talapanna Font to see the similar problems that I faced when designing the Initial Forms.
My FontsReviews: MainTypeFont CreatorHelpFC15 + MT12.0 @ Win 10 64-bit build 19045.2486
William
Top Typographer
Top Typographer
Posts: 2038
Joined: Tue Sep 14, 2004 6:41 pm
Location: Worcestershire, England
Contact:

Re: Glyphs going above ascent height/below descent.

Post by William »

Something that I sometimes do is to add a character at U+E700 that has an advance width of 2048 font units and a glyph that is a rectangle that goes from 0 to 256 font units horizontally and from 2048 font units vertically down to some whole number multiple of -256 vertically and then calculate the metrics.

Then keep every other glyph within those vertical measurements.

I have a great liking for multiples of 256, or 128 or 64 or sometimes 32 for positioning points when designing glyphs.

I feel that it helps the on-screen rendering at 12 point, 24 point, 36 point and so on. However 14 point on the screen can then look terrible!

Those are just some things that I do myself. Other people may have different views. However, it seems to work for me.

William Overington

29 August 2013
David Webber
Posts: 44
Joined: Mon Aug 19, 2013 10:33 pm

Re: Glyphs going above ascent height/below descent.

Post by David Webber »

Bhikkhu Pesala wrote:You may run into problems with the glyphs being clipped below WinDescent or above WinAscent. I suggest simply recalculating the metrics after designing the font as you want. The other glyphs will be smaller and the default line-spacing for single spaced text will be excessive for text, but I don't know of any other reliable method. ...
Thanks for this. But I'm not sure about just recalculating the metrics.

At the moment I'm combining two of my fonts into one by:
- making a copy of my Unicode text font
- importing music symbols from a Symbol font into the private use area U+E000 onwards

For historical reasons the two original fonts had very different WinAscent, WinDescent, x-Height etc, (The symbol one was much larger.)

So, having imported all the musical symbols, I think I am going to have to use the Glyph transformer to scale all the musical symbols down by the same factor, to get them within the [WinDescent,WinAscent] range. And check that the baseline and left-bearing of each is correct (as these are crucial when drawing the music).

If I leave the metrics alone (as inherited from my earlier text font), then

a) other programs (eg Microsoft Word) will be able to open the font and use the Latin text with the expected size and line spacing. The music symbols (if used) will be at some arbitrary scale relative to them, but that isn't too important - there are no rules!

b) my music software will have to find a way of selecting the font so that a specific symbol (U+E000 as it happens) is printed at a specified pixel height. (The text will scale in a fairly arbitrary way, but I can live with that.)

David Webber

[PS love your Talapanna font in the pdf link! I see you have made it available - I will have a look inside and see how you've done it!]
David Webber
Posts: 44
Joined: Mon Aug 19, 2013 10:33 pm

Re: Glyphs going above ascent height/below descent.

Post by David Webber »

Having combined the two fonts by making a copy of the Unicode text font, and then importing the symbols, this is what I get in Microsoft Word:
firstversion.jpg
firstversion.jpg (62 KiB) Viewed 5025 times
I haven't touched the metrics and as Bhikkhu predicted, symbols which go outside the WinDescent-WinAscent range are cropped.

But the comments here have helped me focus on the problem: if the music symbols are sized relative to the text as I thought they should be (as in the picture here) then the ascent/descent of some symbols will be so large that they'll force a very large line spacing in the text, which I don't want.

My immediate feeling now is that this is unavoidable, and that I should scale down the music symbols by something like 50%. I could allow for that when loading the font to draw the music. But my second thought is that I don't understand how Bhikkhu's enormous initial characters work in the exciting Talapanna font. Pause for more research I think!

David Webber
Post Reply