DRC
|
552dd27bde
|
tjTransform() was not working properly if r.w=0 or r.h=0
|
2011-03-15 20:09:47 +00:00 |
|
DRC
|
074a6f24bb
|
Handle 4:4:0 (transposed 4:2:2 subsampling)
|
2011-03-04 12:52:45 +00:00 |
|
DRC
|
5e9da424a4
|
Tile generation did not work with TJXFORM_HFLIP, because the underlying transform code was using an in-place algorithm, which modified the source coefficients after the first tile was generated. Thus, create a new option which allows TurboJPEG to turn off the in-place horizontal flip if there are multiple transforms being performed from the same set of coefficients.
|
2011-03-04 03:20:34 +00:00 |
|
DRC
|
66b6d80ed6
|
Since lossless transforms have such a high fixed performance cost, implement a mechanism for performing more than one at a time on the same source image.
|
2011-03-02 09:22:41 +00:00 |
|
DRC
|
0c7b322efe
|
Fix Windows build; Add grayscale tests to Windows build; Add proper implementation of snprintf on Windows and use snprintf() in TurboJPEG library and tests instead of sprintf()
|
2011-03-02 02:17:30 +00:00 |
|
DRC
|
b02715c1b1
|
Fix compiler warnings in Visual C++
|
2011-03-02 01:27:26 +00:00 |
|
DRC
|
be085e08b3
|
tjGetScaledSize() would never be able to accommodate scaling factors > 1, so replace it with a function that returns a list of fractional scaling factors that TurboJPEG supports.
|
2011-03-01 09:53:07 +00:00 |
|
DRC
|
da6a1f6c76
|
Don't call jpeg_finish_compress() with YUV output, because it tries to insert an EOI marker right in the middle of our image data.
|
2011-03-01 08:23:53 +00:00 |
|
DRC
|
3f2ef301c4
|
Expose libjpeg lossless transform feature in TurboJPEG/OSS
|
2011-02-26 22:02:37 +00:00 |
|
DRC
|
28db388687
|
Fix compiler warnings
|
2011-02-26 21:20:46 +00:00 |
|
DRC
|
02156d7c53
|
"tjScaledSize"="tjGetScaledSize"
|
2011-02-26 19:53:58 +00:00 |
|
DRC
|
3405865cd2
|
Replace the TJ_YUV flag with two new API functions
|
2011-02-25 02:08:23 +00:00 |
|
DRC
|
38218222cd
|
Make the scaling API a bit more friendly
|
2011-02-22 06:41:29 +00:00 |
|
DRC
|
9eecea2a8e
|
Add TJBUFSIZEYUV() convenience function
|
2011-02-22 00:16:14 +00:00 |
|
DRC
|
aa8b75afa8
|
Added scaling API to TurboJPEG/OSS
|
2011-02-15 08:31:34 +00:00 |
|
DRC
|
50103b80ca
|
Ensure that YUV temp buffers get properly freed if a failure occurs
|
2011-02-15 05:26:28 +00:00 |
|
DRC
|
21b5c4bc4d
|
The fast integer DCT degrades for qualities > 95, so use the slow integer DCT instead to ensure that perceptually lossless performance is maintained.
|
2011-01-25 06:52:31 +00:00 |
|
DRC
|
2e1b59c09d
|
Eliminate all of the __memalign() mess and just handle the alignment ourselves
|
2011-01-05 23:35:53 +00:00 |
|
DRC
|
3c7c57b3ad
|
Ensure temp. buffers for YUV output are 16-byte aligned on 32-bit Windows
|
2011-01-05 22:33:24 +00:00 |
|
DRC
|
f0d2577482
|
Added extended tjDecompressHeader() function which can determine the type of subsampling used in the JPEG image
|
2010-12-14 01:21:29 +00:00 |
|
DRC
|
55a9cbffe7
|
Fix various memory corruption issues in the new JPEG->YUV feature
|
2010-12-11 06:01:11 +00:00 |
|
DRC
|
f54cc16225
|
If the output width/height are not an even number of MCU's, then use an intermediate buffer to ensure that the output format is XVideo-compatible
|
2010-12-10 10:58:49 +00:00 |
|
DRC
|
f80fb63847
|
Provide TJ_YUV option for tjDecompress() as well
|
2010-12-10 04:59:13 +00:00 |
|
DRC
|
979a590374
|
Merge the TurboJPEG planar YUV encoding feature from VirtualGL 2.2
|
2010-11-24 04:02:37 +00:00 |
|
DRC
|
366f303099
|
Grayscale bitmap support in TurboJPEG/OSS
|
2010-11-04 22:39:59 +00:00 |
|
DRC
|
a478af780a
|
Space = Tab
|
2010-10-13 19:22:50 +00:00 |
|
DRC
|
0298a12bd8
|
Bleepin' Windows uses LLP64, not LP64
|
2010-02-26 23:01:19 +00:00 |
|
DRC
|
f72fb65759
|
Make TJ_FORCE* options work correctly
|
2010-02-22 08:34:44 +00:00 |
|
DRC
|
cf75c6f1be
|
Test fast upsampling
"4:1:1 = 4:2:0"
|
2009-04-05 21:53:20 +00:00 |
|
DRC
|
65b7cbecf3
|
Include low-level unit tests borrowed from VirtualGL
|
2009-04-03 12:04:24 +00:00 |
|