emerth
6ed94cfa3a
Update README.md
...
Windows build tries to link in functions from libdeflate and zstd library. These are neither included in the build depends scheme nor mentioned in the instructions. So, I propose they at least be mentioned.
2022-02-08 20:00:10 -07:00
Alexander Karpinsky
6fdb1faf39
Merge pull request #110 from radarhere/patch-3
...
Fixed typos
2022-01-17 13:12:52 +03:00
Andrew Murray
c4f10c8b10
Fixed typos
2022-01-17 15:29:07 +11:00
Alexander Karpinsky
fb9ed499a2
SIMD. version
2022-01-04 19:10:00 +03:00
Alexander Karpinsky
abd6708434
Merge pull request #109 from uploadcare/simd/fix-builds
...
Simd/fix builds
2022-01-04 19:05:54 +03:00
Alexander Karpinsky
b5f03c9e74
Two items, not two bytes
2022-01-04 18:02:25 +03:00
Alexander Karpinsky
e8fa86d630
Exclude CIFuzz tests
2022-01-04 17:32:18 +03:00
Alexander Karpinsky
afac531208
Exclude other arch tests
2022-01-04 17:01:05 +03:00
Alexander Karpinsky
4e4dc1dc15
Protect double inclusion of ImagingSIMD.h
2022-01-04 16:55:13 +03:00
Alexander Karpinsky
da01a5f472
SIMD. version
2022-01-03 12:20:10 +03:00
Alexander Karpinsky
d40b9f0ab4
SIMD. resolve imports
2022-01-03 12:20:10 +03:00
Alexander Karpinsky
fc915a4ed5
SIMD Filter. Correct offset for 3x3f_u8
2022-01-03 12:20:10 +03:00
Roman Sedykh
18783eeefa
New Uploadcare logo and date correction
2022-01-03 12:20:10 +03:00
lsb
8db8caa119
Markdown typo
...
Remove extra parenthesis. Now the link can be parsed as a link
2022-01-03 12:20:10 +03:00
Monirzadeh
c0d7cd9602
#84 add install prerequisites for avoid confutation
2022-01-03 12:20:10 +03:00
Alexander
d37299775b
SIMD. version
2022-01-03 12:20:10 +03:00
Alexander
4f92606af0
SIMD. update changes link
2022-01-03 12:20:09 +03:00
Alexander
000dafeb6d
SIMD. MSVC compiler
2022-01-03 12:20:09 +03:00
Alexander
1e70975dc1
SIMD ColorLUT. 16bit arithmetic. Access violation fixed
2022-01-03 12:20:09 +03:00
Alexander
085b72935b
SIMD. version
2022-01-03 12:20:09 +03:00
Alexander
6e22415933
SIMD ColorLUT. finally fix alpha copy
2022-01-03 12:20:09 +03:00
Alexander
c31c78819f
SIMD ColorLUT. Fast AVX2 implementation with very wired slowdown
2022-01-03 12:20:09 +03:00
Alexander
8a52520e66
SIMD ColorLUT. AVX2 implementation (near the same speed)
2022-01-03 12:20:09 +03:00
Alexander
b4a77d406c
SIMD ColorLUT. remove left_mask and right_mask
2022-01-03 12:20:09 +03:00
Alexander
0520521560
SIMD ColorLUT. remove unused utility
2022-01-03 12:20:09 +03:00
Alexander
91dcda6b43
SIMD ColorLUT. table_channels==4 case, minor optimizations
2022-01-03 12:20:09 +03:00
Alexander
c7e3b0ead9
SIMD ColorLUT. improve performance by preliminary index calculation
2022-01-03 12:20:09 +03:00
Alexander
f50dba4aad
SIMD ColorLUT. remove SHIFT_ROUNDING
2022-01-03 12:20:09 +03:00
Alexander
16f69404f1
SIMD ColorLUT. First try
2022-01-03 12:20:09 +03:00
Alexander
23451b8423
SIMD. version
2022-01-03 12:20:09 +03:00
Alexander
1b8030989f
SIMD. sse only when needed
2022-01-03 12:20:09 +03:00
Alexander
be70e9eded
SIMD Bands. access violation in getband
2022-01-03 12:20:09 +03:00
Alexander
d74ab21ac9
SIMD Reduce. sse4 implementation
2022-01-03 12:20:09 +03:00
Alexander
33b11280b0
SIMD. Fix memory read error on some compilers
2022-01-03 12:20:09 +03:00
Alexander
f9c5589cac
SIMD. Gather all SSE and AVX imports in ImPlatform.h
2022-01-03 12:20:09 +03:00
Alexander
e4ca18a9c8
SIMD Bands
2022-01-03 12:20:09 +03:00
Alexander
8c3d6a32d1
SIMD Convert. Fix rounding error on RGB to L conversion
2022-01-03 12:20:09 +03:00
Alexander
52cfbf339a
SIMD Convert. Speedup other 2L convertions
2022-01-03 12:20:09 +03:00
Alexander
4fbc9d1f37
SIMD Convert. RGB → L 2.2 times faster
2022-01-03 12:20:09 +03:00
Alexander
7398f9e1c9
SIMD Convert. fix rounding and speedup a bit
2022-01-03 12:20:09 +03:00
Alexander
3322b77c9a
SIMD Convert. use 16bit arithmetics
2022-01-03 12:20:09 +03:00
Alexander
03c150d0a6
SIMD Convert. sse4 version (still 1.4x faster than previous avx2 implementation)
2022-01-03 12:20:09 +03:00
Alexander
d82fe3dd74
SIMD Convert. use float div instead of gather
2022-01-03 12:20:09 +03:00
homm
95a68340c9
SIMD Convert. fix RGBa → RGBA conversion on AVX2
2022-01-03 12:20:09 +03:00
homm
f888dbe1cc
SIMD Convert. RGBa → RGBA convert using gather
2022-01-03 12:20:09 +03:00
homm
55b71f2edb
SIMD Convert. avx2 implementation
2022-01-03 12:20:09 +03:00
homm
6e91b2b41f
SIMD Convert. faster implementation
2022-01-03 12:20:09 +03:00
homm
a7de792025
SIMD Convert. sse4 implementation
2022-01-03 12:20:09 +03:00
Alexander
a71b748bde
SIMD Resample. Correct max coefficient calculation
...
Minimal test case:
Image.new('RGB', (341, 60), 'pink').resize((170, 60), Image.LANCZOS).save('_out.png')
2022-01-03 12:20:09 +03:00
Alexander
b3c28911c0
SIMD Resample. fix wrong usage of xmax for division compensation
2022-01-03 12:20:09 +03:00