[Devel] Rendering (by FreeType) and hinting of Latin small 'a'

[Vadim Plessky]

I'd like to discuss (again) rendering of Serif fonts, and, in particular (to 
narrow the scope) - Latin small 'a'.
I have prepared a web page with examples demonstrating different rendering of 
this glyph
http://freetype.newmail.ru/misc/hinting-a.html

You may want also to download experimental 3-glyph font which I prepared for 
testing 
http://freetype.newmail.ru/misc/hinting-a-test.zip
Zip archive contains PfaEdit's .sfd, PostScript Type1 .pfb  and OpenType .otf 
font files.
PfaEdit can be downloaded from http://pfaedit.sourceforge.net

Now about problem I see at a moment.
1) topology of 'a' is destroyed at all examples, except non-AA rendering.
I refer here to central, near-horizontal curve, in the middle of 'a'
enabling sub-pixel rgb decimation doesn't help here.
 
2) in ftview rendering, even bottom part of 'a' is disconnected from the 
vertical stem.
It's very visible at screenshot in natural size, and on enlarged s/s as well.

3) My testing of 5 different Times clones (variants), in addition to the 
version I develop,  demonstrates, that all variants of Times, with exception
 of Times New Roman, have similiar problem described above.
See screenshots at th ebottom of th epage (Times New Roman is 2nd in series, 
each three rendered glyphs are 12pt, 11pt, 10pt Latin 'a')

Here are my questions:
a) can this *topology* (visually broken  curve) issue be fixed?
I demonstrated problem on example of 'a', but many other glyphs are affected 
(in AA or RGN-AA modes), too.

b) how rendering (by FreeType, not too much interested in Adobe Type Manager 
or Windpows' rasterizer) of middle part of 'a' can be enhanced?
Should I artificially deformate curve to make constrains snapping to 
horizontal position? (like is done in Times New Roman)
This is not very desirable, as 'a' in Times New Roman has too small bottom 
part, IMO, and this is not easy for screen reading.
Diagonal hinting also doesn't help here, as, frankly speaking, this part 
does't conform to diagonal hinting definition.

c) Is it possible, in some way, to increase *tension* for FT rasterizer when 
it follows some part of the contour/curve?
It's like of driving a car with high speed, and follow smooth turn to, say, 
left at 80km/hour.
It can be done, if you increase your speed instantly /with acceleration/ :-)
So, I think we should find a way to do the same with such curves.
Is it possible to do?  What are your opinions?

-- 

Vadim Plessky
http://kde2.newmail.ru  (English)
33 Window Decorations and 6 Widget Styles for KDE
http://kde2.newmail.ru/kde_themes.html
KDE mini-Themes
http://kde2.newmail.ru/themes/