Re: [ft-devel] FT_LOAD_FORCE_AUTOHINT and FT_LOAD_FORCE_AUTOHINT

[David Turner]

On Fri, 15 Jun 2007 13:07:34 -0700, "Derek Clegg" <address@hidden> said:
> Just to follow up a bit, it appears that part of my problem was my  
> choice of flags to "FT_Load_Glyph" &#8212; since I wanted the glyph path in  
> font units, I passed in "FT_LOAD_NO_SCALE |  
> FT_LOAD_IGNORE_TRANSFORM".  It appears, however, that this ended up  
> by-passing the hinter, and so my glyph paths were incorrect.  I'm not  
> sure whether this is by design, however.
> 
> In any case, thanks again for your help.
> 
> Derek
>

yes, in these fonts, the original glyph coordinates in font units, as stored
in the file, are totally un-appropriate. only the bytecode interpreter can
correct them to give you something usable.

if you want something more or less "equivalent" to original font units, load
the glyphs at a very high pixel size, since this should minimize the hinting
distorsion, then scale down to the appropriate value.

hope this helps,

- David 

 
> On Jun 14, 2007, at 3:48 PM, Derek Clegg wrote:
> 
> > Thanks very much to you and Werner &#8212; this is excellent and very  
> > helpful information.  I appreciate it a lot.
> >
> > Derek
> >
> > On Jun 10, 2007, at 4:47 PM, David Turner wrote:
> >
> >> Hello Derek,
> >>
> >> FreeType can return glyph paths for these fonts, but you need to
> >> use either the patented or unpatented bytecode interpreter to get
> >> correct results (auto-hinting or even no hinting will return
> >> garbage). this means one of the following:
> >>
> >> - activate the patented bytecode interpreter (since you have an  
> >> @apple.com
> >>   e-mail address, I guess this shouldn't be a problem to you :o) )
> >>
> >> - the font face name is "recognized" by FreeType (which contains a  
> >> hard-coded
> >>   list in src/truetype/ttobjs.c). In this case, the font engine  
> >> automatically
> >>   uses the unpatented hinter when calling FT_Load_Glyph (unless  
> >> you disable
> >>   hinting, force auto-hinting, or use the "light" hinting mode)
> >>
> >> - you activate the unpatented hinter by using the appropriate  
> >> parameter
> >>   (FT_PARAM_TAG_UNPATENTED_HINTING) when calling FT_Open_Face. See  
> >> Werner's
> >>   last e-mail on the subject for the details.
> >>
> >> I've just added a new API called FT_Face_CheckTrueTypePatents that  
> >> scans
> >> *all* the glyph hints in a given TrueType fonts to see if they use  
> >> any of
> >> the patented opcodes. This should help you determine if you should  
> >> use the
> >> unpatented hinter or not.
> >>
> >> Note that this call is *slow* by design, since it parses the whole  
> >> font file,
> >> so caching its result is better.
> >>
> >> There is also now a program called "ftpatentcheck" in 'ft2demos'  
> >> that can
> >> be used to check wether a given font uses patented opcodes (still  
> >> very rough,
> >> but functional)
> >>
> >> I'll later try to add the ability to switch the unpatented hinter  
> >> *after*
> >> creating/opening a FT_Face; that should be more convenient to use.
> >>
> >> Hope this helps,
> >>
> >> - David Turner
> >> - The FreeType Project  (www.freetype.org)
> >>
> >>
> >> On Tue, 29 May 2007 10:04:26 -0700, "Derek Clegg"  
> >> <address@hidden> said:
> >>> On May 27, 2007, at 11:34 PM, Werner LEMBERG wrote:
> >>>
> >>>> Whatever font it is, the images do *not* show `PMingLiU' from
> >>>> mingliu.ttc!  This font can't be displayed with the autohinter  
> >>>> because
> >>>> the glyph components need the bytecode interpreter to be positioned
> >>>> correctly.  Please always use ftview, ftstring, or ftdiff to  
> >>>> inspect
> >>>> the font.  At the font size you are showing, mingliu.ttc contains
> >>>> embedded bitmaps to improve the rendering, BTW.
> >>>
> >>> Can FreeType return glyph paths from mingliu and other fonts which
> >>> need the bytecode interpreter for positioning?  In my experiments,
> >>> I've failed to get correct results, but that may be my error  
> >>> &#8212; I may
> >>> not have supplied the correct combination of build parameters and
> >>> flags.  In my case, I'm only interested in the path, not bitmaps.
> >>> Thanks for any info!
> >>>
> >>> Derek
> >>>
> >>>
> >>>
> >>> _______________________________________________
> >>> Freetype-devel mailing list
> >>> address@hidden
> >>> http://lists.nongnu.org/mailman/listinfo/freetype-devel
> >
> >
> >
> > _______________________________________________
> > Freetype-devel mailing list
> > address@hidden
> > http://lists.nongnu.org/mailman/listinfo/freetype-devel
> 
> 
> 
> _______________________________________________
> Freetype-devel mailing list
> address@hidden
> http://lists.nongnu.org/mailman/listinfo/freetype-devel