>From: Selva Nair <[EMAIL PROTECTED]>
>
>> This is all revealed in the table of frequency response graphs on the
>> web page. (I need to add a table of timing benchmarks, too.)
>
>Those links (http://www.mir.com/DMG/Software/freq-graph-table.html and
>http://www.mir.com/DMG/Software/step-
>> (With the two bug fixes, I can now wholeheartedly recommend the sinc4lan
>> and sinc8lan kernels if you have cycles to burn.)
>
> Hmmm, what makes one kernel preferable to another? If, for example,
> I'm mainly cropping/scaling from 720x480 to 352x480 (for either CVD
> o
Hi!
> From: Matto Marjanovic <[EMAIL PROTECTED]>
> Version 0.5.0 of y4mscaler is now available at
Great!
> (With the two bug fixes, I can now wholeheartedly recommend the sinc4lan
> and sinc8lan kernels if you have cycles to burn.)
Hmmm, what makes one kernel preferable to anot
Hi, all,
Version 0.5.0 of y4mscaler is now available at
http://www.mir.com/DMG/Software/
Please try it out, and send any bug reports/comments/etc to dmg @ mir.com
New features:
o Geometry strings for active and matte regions are now allowed to specify
just the offset, with the size in
