On Saturday 21 September 2002 6:29 am, Allen Akin wrote:
| On Fri, Sep 20, 2002 at 08:49:09PM +0400, Vadim Plessky wrote:
| | I am, as many other office workers, primary user of different word
| | processor(s), spearsheet(s), etc., and I am mostly concerned about
| | quality of rendering for typical text.
|
| I'm not an expert in digital typography. I've been running some
| experiments with OpenGL-based text rendering, but I've got a ways to go
| before I reach any strongly-defensible conclusions.
|
Can you please post some screenshots?
I also think that anti-aliased OpenGL can be used for text rendering.
But never had time to experiment with it.
| What I can say today is that using hardware full-scene antialiasing
| techniques for rendering small text as geometry works surprisingly well
| on a CRT. (4X supersampling with Gaussian filtering seems best at the
| moment. I'd like to try the new FSAA techniques available on the Radeon
| 9700.)
Question: what kind of filtration XRENDER / text in AA uses?
Isn't it Gaussian?
What is typical size of filtration matrix/operator kernel?
Let's suppose we have:
Ax=y
where 'x' is source image (pixmap), and 'y' is resulting image after
transform.
And, to optimize speed, we use integers to calculate transformation.
A is operator matrix:
| a(1,1) a(1,2) .... a(1,n) |
A = | a(2,1) a(2,2) .... a(2,n) |
| ........ |
| a(n,1) a(n,2) .... a(n,n) |
What is the value of 'n' in transformations used to render text in XFree86?
10 year ago, when I just graduated from the university, typical values for n
were either 3 or 5.
But hat was time of i386dx/33, and 486 was just coming.
I guess youwe can use n=25 or n=99 nowdays.
Therefor, quality of Gaussian filtration would be much higher.
It would be (almost) real Gaussian filter, not poor aproximation.
|
| I'd like to look into the process of choosing a level-of-detail that's
| appropriate for a given drawing size. Obviously this touches on hinting
| issues as well as more straightforward problems of approximating complex
| geometry.
|
| I speculate that using geometry rather than pixmaps of antialiased
| characters is a better approach in the long run, because with geometry
| it's easier to apply small transformations, e.g. subpixel positioning or
| scaling to preserve sharp vertical elements.
|
| I also plan to try multipass rendering to implement high-quality
| antialiasing filters with support greater than one pixel in area. Given
| current hardware performance levels, it's quite practical to do this,
| and it has the advantage that it can take into account the spot
| characteristics of CRTs.
|
| Allen
| _______________________________________________
| Render mailing list
| [EMAIL PROTECTED]
| http://XFree86.Org/mailman/listinfo/render
--
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/
_______________________________________________
Render mailing list
[EMAIL PROTECTED]
http://XFree86.Org/mailman/listinfo/render
