python-pillow/Pillow
1
1
Fork 0
mirror of https://github.com/python-pillow/Pillow.git synced 2025-11-04 01:47:47 +03:00
Code Issues Packages Projects Releases Wiki Activity

Precision improvement. No performance affected.

Browse Source
This commit is contained in:
homm 2013-03-25 00:06:02 +04:00
parent 0663e14444
commit af96b5d4b2
1 changed files with 15 additions and 13 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

28
libImaging/AlphaComposite.c
View File

@ -69,21 +69,23 @@ ImagingAlphaComposite(Imaging imDst, Imaging imSrc)
// almost equivalent to: // almost equivalent to:
// tmp = a + (2 << (n-1)), ((tmp >> n) + tmp) >> n // tmp = a + (2 << (n-1)), ((tmp >> n) + tmp) >> n
// 0xff * 0xff = 16 meaningful bits.
UINT16 blend = dst->a * (255 - src->a); UINT16 blend = dst->a * (255 - src->a);
// Shift 4 bits up, to don't loose blend precision UINT16 outa255 = src->a * 255 + blend;
// on very transparent pixels. // There we use 7 bits for precision.
UINT16 outa = (src->a << 4) + (((blend << 4) + (blend >> 4) + 0x80) >> 8); // We could use more, but we go beyond 32 bits.
UINT16 coef1 = (((src->a << 8) - src->a) << 8) / outa; // 12 UINT16 coef1 = src->a * 255 * 255 * 128 / outa255;
UINT16 coef2 = (blend << 8) / outa; // 12 UINT16 coef2 = blend * 255 * 128 / outa255;
UINT32 tmpr = src->r * coef1 + dst->r * coef2 + 0x800; #define SHIFTFORDIV255(a)\
out->r = ((tmpr >> 8) + tmpr) >> 12; ((a >> 8) + a >> 8)
UINT32 tmpg = src->g * coef1 + dst->g * coef2 + 0x800;
out->g = ((tmpg >> 8) + tmpg) >> 12; UINT32 tmpr = src->r * coef1 + dst->r * coef2 + (0x80 << 7);
UINT32 tmpb = src->b * coef1 + dst->b * coef2 + 0x800; out->r = SHIFTFORDIV255(tmpr) >> 7;
out->b = ((tmpb >> 8) + tmpb) >> 12; UINT32 tmpg = src->g * coef1 + dst->g * coef2 + (0x80 << 7);
out->a = (outa + 0x7) >> 4; out->g = SHIFTFORDIV255(tmpg) >> 7;
UINT32 tmpb = src->b * coef1 + dst->b * coef2 + (0x80 << 7);
out->b = SHIFTFORDIV255(tmpb) >> 7;
out->a = SHIFTFORDIV255(outa255 + 0x80);
} }
dst++; src++; out++; dst++; src++; out++;