[cairo-bugs] [Bug 84396] cairo-tor-scan-converter is off by half of a subrow.

[bugzilla-daemon at freedesktop.org]

https://bugs.freedesktop.org/show_bug.cgi?id=84396

Massimo <sixtysix at inwind.it> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
 Attachment #107095|0                           |1
        is obsolete|                            |

--- Comment #25 from Massimo <sixtysix at inwind.it> ---
Created attachment 107344
  --> https://bugs.freedesktop.org/attachment.cgi?id=107344&action=edit
increase coverage area precision

There is a loss of precision in the last step of the analytical
coverage computation, that is, the uncovered area does not depend
on the number of subrows and can be computed to a greater accuracy,
improving the overall symmetrical behaviour:

the output images of record{,flip}-paint-alpha-clip-mask become
the 180°-rotated version of each other

and the output of a8-clear is within +-1 from its 180°-rotated
version.

If something like this is considered worth it, perhaps after playing
with GRID_Y or GRID_Y_BITS and EXTRA_Y, a specialized version
of GRID_AREA_TO_ALPHA could be added for the new value of GRID_XY.

-- 
You are receiving this mail because:
You are the QA Contact for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.cairographics.org/archives/cairo-bugs/attachments/20141005/b86e21eb/attachment.html>