freetype-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [ft-devel] How the new lcd filter coefficients in 2.6.2 were choosen


From: Markus Trippelsdorf
Subject: Re: [ft-devel] How the new lcd filter coefficients in 2.6.2 were choosen?
Date: Wed, 3 May 2017 15:36:46 +0200

On 2017.05.03 at 09:20 -0400, Alexei Podtelezhnikov wrote:
> >> So, how did you get the values used currently as the default? Are there 
> >> other
> >> properties besides the two mentioned, that helped you to "calculate" such a
> >> filter? Or did you just try different variants and choose the one that 
> >> looked
> >> better for you?
> 
> Without filtering, an outline covers unequal amounts of color
> channels, which is visible as color fringes especially with thin
> stems. So filtering should enforce equality between color channels.
> The box fiilter (aka light) "00 55 56 55 00" is the absolute minimum.
> The light filter is actually my favorite. The default filter
> (suggested by Dave Arnold from Adobe) "08 4d 56 4d 08" is only
> slightly different from the light filter. It is slightly wider and
> blurrier than the light filter and, as such, more tolerant to
> incorrect gamma setting.

Apropos gamma setting. CFF fonts will get linear blended (with stem
darkening) by default in upcoming Qt 5.9:
http://code.qt.io/cgit/qt/qtbase.git/commit/?h=5.9&id=21f27c9a4

-- 
Markus



reply via email to

[Prev in Thread] Current Thread [Next in Thread]