[Pixman] [PATCH 0/2] 7-bit bilinear interpolation precision

[Søren Sandmann]

Søren Sandmann <sandmann at cs.au.dk> writes:

> For bilinear filtering, the question of rounding down is mostly
> irrelevant because we need both the pixel below and above. In this case
> step 3 and 4 become simply a truncation.

This is not true actually. To get the lower pixel, 0.5 is subtracted,
and then the position is truncated. But the question of rounding down or
up is irrelevant since when the location is exactly on top of a pixel,
the two cases both end up with that pixel having a weight of 1 and the
other (whether below or above) a weight of 0.


Soren