[XeTeX] Devanagari rendering with XeTeX - a comparision

[Yves Codet]

Hello Deepak.

I am afraid I cannot answer your technical questions, well, only  
partly. OpenType rendering is not so bad in itself, I suppose it  
mainly depends on the font. AAT rendering is not always good, I  
suppose it mainly depends on the font :) Please have a look at the  
attached sample (compiled on a Mac with TL 2009). You will notice that  
placement of accents is much better with Sanskrit2003 than with  
Devanagari MT. But both can display the right shape for rr̥.

Best wishes,

Yves

-------------- next part --------------
A non-text attachment was scrubbed...
Name: test.tex
Type: application/octet-stream
Size: 624 bytes
Desc: not available
URL: <http://tug.org/pipermail/xetex/attachments/20100106/ba80ee58/attachment-0001.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: test.pdf
Type: application/pdf
Size: 24062 bytes
Desc: not available
URL: <http://tug.org/pipermail/xetex/attachments/20100106/ba80ee58/attachment-0001.pdf>
-------------- next part --------------


Le 6 janv. 10 à 19:10, Deepak Jois a écrit :

>
> Hi All
>
> I have been using XeTeX to test the rendering of Devanagari using
> different fonts and layout engines. Here is a sample containing
> multiple renderings of a word from the Rig Veda.
>
> Sample Rendering :
> http://farm3.static.flickr.com/2755/4250942627_469615ffbc_o.png
>
> TeX source : http://gist.github.com/270468
>
> The first four are using OpenType fonts (Nakula, Sahadeva, Akshar and
> Gargi respectively). The last one uses Devanagari MT. In my (not so
> expert) opinion, the Devangari MT/AAT rendering while not perfect, is
> superior to the OpenType rendering by a fair stretch. I am mailing
> this list to get some insights about the rendering process, and see if
> anything can be done to improve it for OpenType fonts. Specifically :
>
> 1. Why is the OpenType/ICU rendering bad, compared to AAT. Is it the
> font, or is it the layout engine, or both? My limited understanding is
> that the Devangari MT font contains glyph layout information within
> itself, as opposed to the OpenType fonts which rely more on software
> to layout the glyphs. Is there an inherent advantage in the font
> containing the layout information, or can software be made to
> replicate a similar quality?
>
> 2. What can be done to improve this? Are there better fonts? Does a
> later version of ICU (i believe XeTeX uses ICU 4.0) contain any
> enhancements?
>
> 3. Can anyone tell if Uniscribe on Windows Vista does a better job. I
> would appreciate a sample, since I dont have a Windows system here to
> test it out myself.
>
> Thanks
> Deepak
>
>
> --------------------------------------------------
> Subscriptions, Archive, and List information, etc.:
>  http://tug.org/mailman/listinfo/xetex
>