[reactos.git] / reactos / dll / 3rdparty / libjpeg / change.log
1 CHANGE LOG for Independent JPEG Group's JPEG software
4 Version 9b 17-Jan-2016
5 -----------------------
7 Improvements and optimizations in DCT and color calculations.
8 Normalize range limit array composition and access pattern.
9 Thank to Sia Furler and Maddie Ziegler for inspiration.
11 Use merged upsample with scaled DCT sizes larger than 8.
12 Thank to Taylor Hatala for inspiration.
14 Check for excessive comment lengths in argument parsing in wrjpgcom.c.
15 Thank to Julian Cohen for hint.
17 Add makefile.b32 for use with Borland C++ 32-bit (bcc32).
18 Thank to Joe Slater for contribution.
20 Document 'f' specifier for jpegtran -crop specification.
21 Thank to Michele Martone for suggestion.
23 Use defined value from header instead of hardwired number in rdswitch.c.
24 Thank to Robert Sprowson for hint.
27 Version 9a 19-Jan-2014
28 -----------------------
30 Add support for wide gamut color spaces (JFIF version 2).
31 Improve clarity and accuracy in color conversion modules.
32 Note: Requires rebuild of test images.
34 Extend the bit depth support to all values from 8 to 12
35 (BITS_IN_JSAMPLE configuration option in jmorecfg.h).
36 jpegtran now supports N bits sample data precision with all N from 8 to 12
37 in a single instance. Thank to Roland Fassauer for inspiration.
39 Try to resolve issues with new boolean type definition.
40 Thank also to v4hn for suggestion.
42 Enable option to use default Huffman tables for lossless compression
43 (for hardware solution), and in this case improve lossless RGB compression
44 with reversible color transform. Thank to Benny Alexandar for hint.
46 Extend the entropy decoding structure, so that extraneous bytes between
47 compressed scan data and following marker can be reported correctly.
48 Thank to Nigel Tao for hint.
50 Add jpegtran -wipe option and extension for -crop.
51 Thank to Andrew Senior, David Clunie, and Josef Schmid for suggestion.
54 Version 9 13-Jan-2013
55 ----------------------
57 Add cjpeg -rgb1 option to create an RGB JPEG file, and insert
58 a simple reversible color transform into the processing which
59 significantly improves the compression.
60 The recommended command for lossless coding of RGB images is now
61 cjpeg -rgb1 -block 1 -arithmetic.
62 As said, this option improves the compression significantly, but
63 the files are not compatible with JPEG decoders prior to IJG v9
64 due to the included color transform.
65 The used color transform and marker signaling is compatible with
66 other JPEG standards (e.g., JPEG-LS part 2).
68 Remove the automatic de-ANSI-fication support (Automake 1.12).
69 Thank also to Nitin A Kamble for suggestion.
71 Add remark for jpeg_mem_dest() in jdatadst.c.
72 Thank to Elie-Gregoire Khoury for the hint.
74 Support files with invalid component identifiers (created
75 by Adobe PDF). Thank to Robin Watts for the suggestion.
77 Adapt full buffer case in jcmainct.c for use with scaled DCT.
78 Thank to Sergii Biloshytskyi for the suggestion.
80 Add type identifier for declaration of noreturn functions.
81 Thank to Brett L. Moore for the suggestion.
83 Correct argument type in format string, avoid compiler warnings.
84 Thank to Vincent Torri for hint.
86 Add missing #include directives in configuration checks, avoid
87 configuration errors. Thank to John Spencer for the hint.
90 Version 8d 15-Jan-2012
91 -----------------------
93 Add cjpeg -rgb option to create RGB JPEG files.
94 Using this switch suppresses the conversion from RGB
95 colorspace input to the default YCbCr JPEG colorspace.
96 This feature allows true lossless JPEG coding of RGB color images.
97 The recommended command for this purpose is currently
98 cjpeg -rgb -block 1 -arithmetic.
99 SmartScale capable decoder (introduced with IJG JPEG 8) required.
100 Thank to Michael Koch for the initial suggestion.
102 Add option to disable the region adjustment in the transupp crop code.
103 Thank to Jeffrey Friedl for the suggestion.
105 Thank to Richard Jones and Edd Dawson for various minor corrections.
107 Thank to Akim Demaille for configure.ac cleanup.
110 Version 8c 16-Jan-2011
111 -----------------------
113 Add option to compression library and cjpeg (-block N) to use
114 different DCT block size.
115 All N from 1 to 16 are possible. Default is 8 (baseline format).
116 Larger values produce higher compression,
117 smaller values produce higher quality.
118 SmartScale capable decoder (introduced with IJG JPEG 8) required.
121 Version 8b 16-May-2010
122 -----------------------
124 Repair problem in new memory source manager with corrupt JPEG data.
125 Thank to Ted Campbell and Samuel Chun for the report.
127 Repair problem in Makefile.am test target.
128 Thank to anonymous user for the report.
130 Support MinGW installation with automatic configure.
131 Thank to Volker Grabsch for the suggestion.
134 Version 8a 28-Feb-2010
135 -----------------------
137 Writing tables-only datastreams via jpeg_write_tables works again.
139 Support 32-bit BMPs (RGB image with Alpha channel) for read in cjpeg.
140 Thank to Brett Blackham for the suggestion.
142 Improve accuracy in floating point IDCT calculation.
143 Thank to Robert Hooke for the hint.
146 Version 8 10-Jan-2010
147 ----------------------
149 jpegtran now supports the same -scale option as djpeg for "lossless" resize.
150 An implementation of the JPEG SmartScale extension is required for this
151 feature. A (draft) specification of the JPEG SmartScale extension is
152 available as a contributed document at ITU and ISO. Revision 2 or later
153 of the document is required (latest document version is Revision 3).
154 The SmartScale extension will enable more features beside lossless resize
155 in future implementations, as described in the document (new compression
156 options).
158 Add sanity check in BMP reader module to avoid cjpeg crash for empty input
159 image (thank to Isaev Ildar of ISP RAS, Moscow, RU for reporting this error).
161 Add data source and destination managers for read from and write to
162 memory buffers. New API functions jpeg_mem_src and jpeg_mem_dest.
163 Thank to Roberto Boni from Italy for the suggestion.
166 Version 7 27-Jun-2009
167 ----------------------
169 New scaled DCTs implemented.
170 djpeg now supports scalings N/8 with all N from 1 to 16.
171 cjpeg now supports scalings 8/N with all N from 1 to 16.
172 Scaled DCTs with size larger than 8 are now also used for resolving the
173 common 2x2 chroma subsampling case without additional spatial resampling.
174 Separate spatial resampling for those kind of files is now only necessary
175 for N>8 scaling cases.
176 Furthermore, separate scaled DCT functions are provided for direct resolving
177 of the common asymmetric subsampling cases (2x1 and 1x2) without additional
178 spatial resampling.
180 cjpeg -quality option has been extended for support of separate quality
181 settings for luminance and chrominance (or in general, for every provided
182 quantization table slot).
183 New API function jpeg_default_qtables() and q_scale_factor array in library.
185 Added -nosmooth option to cjpeg, complementary to djpeg.
186 New variable "do_fancy_downsampling" in library, complement to fancy
187 upsampling. Fancy upsampling now uses direct DCT scaling with sizes
188 larger than 8. The old method is not reversible and has been removed.
190 Support arithmetic entropy encoding and decoding.
191 Added files jaricom.c, jcarith.c, jdarith.c.
193 Straighten the file structure:
194 Removed files jidctred.c, jcphuff.c, jchuff.h, jdphuff.c, jdhuff.h.
196 jpegtran has a new "lossless" cropping feature.
198 Implement -perfect option in jpegtran, new API function
199 jtransform_perfect_transform() in transupp. (DP 204_perfect.dpatch)
201 Better error messages for jpegtran fopen failure.
202 (DP 203_jpegtran_errmsg.dpatch)
204 Fix byte order issue with 16bit PPM/PGM files in rdppm.c/wrppm.c:
205 according to Netpbm, the de facto standard implementation of the PNM formats,
206 the most significant byte is first. (DP 203_rdppm.dpatch)
208 Add -raw option to rdjpgcom not to mangle the output.
209 (DP 205_rdjpgcom_raw.dpatch)
211 Make rdjpgcom locale aware. (DP 201_rdjpgcom_locale.dpatch)
213 Add extern "C" to jpeglib.h.
214 This avoids the need to put extern "C" { ... } around #include "jpeglib.h"
215 in your C++ application. Defining the symbol DONT_USE_EXTERN_C in the
216 configuration prevents this. (DP 202_jpeglib.h_c++.dpatch)
219 Version 6b 27-Mar-1998
220 -----------------------
222 jpegtran has new features for lossless image transformations (rotation
223 and flipping) as well as "lossless" reduction to grayscale.
225 jpegtran now copies comments by default; it has a -copy switch to enable
226 copying all APPn blocks as well, or to suppress comments. (Formerly it
227 always suppressed comments and APPn blocks.) jpegtran now also preserves
228 JFIF version and resolution information.
230 New decompressor library feature: COM and APPn markers found in the input
231 file can be saved in memory for later use by the application. (Before,
232 you had to code this up yourself with a custom marker processor.)
234 There is an unused field "void * client_data" now in compress and decompress
235 parameter structs; this may be useful in some applications.
237 JFIF version number information is now saved by the decoder and accepted by
238 the encoder. jpegtran uses this to copy the source file's version number,
239 to ensure "jpegtran -copy all" won't create bogus files that contain JFXX
240 extensions but claim to be version 1.01. Applications that generate their
241 own JFXX extension markers also (finally) have a supported way to cause the
242 encoder to emit JFIF version number 1.02.
244 djpeg's trace mode reports JFIF 1.02 thumbnail images as such, rather
245 than as unknown APP0 markers.
247 In -verbose mode, djpeg and rdjpgcom will try to print the contents of
248 APP12 markers as text. Some digital cameras store useful text information
249 in APP12 markers.
251 Handling of truncated data streams is more robust: blocks beyond the one in
252 which the error occurs will be output as uniform gray, or left unchanged
253 if decoding a progressive JPEG. The appearance no longer depends on the
254 Huffman tables being used.
256 Huffman tables are checked for validity much more carefully than before.
258 To avoid the Unisys LZW patent, djpeg's GIF output capability has been
259 changed to produce "uncompressed GIFs", and cjpeg's GIF input capability
260 has been removed altogether. We're not happy about it either, but there
261 seems to be no good alternative.
263 The configure script now supports building libjpeg as a shared library
264 on many flavors of Unix (all the ones that GNU libtool knows how to
265 build shared libraries for). Use "./configure --enable-shared" to
266 try this out.
268 New jconfig file and makefiles for Microsoft Visual C++ and Developer Studio.
269 Also, a jconfig file and a build script for Metrowerks CodeWarrior
270 on Apple Macintosh. makefile.dj has been updated for DJGPP v2, and there
271 are miscellaneous other minor improvements in the makefiles.
273 jmemmac.c now knows how to create temporary files following Mac System 7
274 conventions.
276 djpeg's -map switch is now able to read raw-format PPM files reliably.
278 cjpeg -progressive -restart no longer generates any unnecessary DRI markers.
280 Multiple calls to jpeg_simple_progression for a single JPEG object
281 no longer leak memory.
284 Version 6a 7-Feb-96
285 --------------------
287 Library initialization sequence modified to detect version mismatches
288 and struct field packing mismatches between library and calling application.
289 This change requires applications to be recompiled, but does not require
290 any application source code change.
292 All routine declarations changed to the style "GLOBAL(type) name ...",
293 that is, GLOBAL, LOCAL, METHODDEF, EXTERN are now macros taking the
294 routine's return type as an argument. This makes it possible to add
295 Microsoft-style linkage keywords to all the routines by changing just
296 these macros. Note that any application code that was using these macros
297 will have to be changed.
299 DCT coefficient quantization tables are now stored in normal array order
300 rather than zigzag order. Application code that calls jpeg_add_quant_table,
301 or otherwise manipulates quantization tables directly, will need to be
302 changed. If you need to make such code work with either older or newer
303 versions of the library, a test like "#if JPEG_LIB_VERSION >= 61" is
304 recommended.
306 djpeg's trace capability now dumps DQT tables in natural order, not zigzag
307 order. This allows the trace output to be made into a "-qtables" file
308 more easily.
310 New system-dependent memory manager module for use on Apple Macintosh.
312 Fix bug in cjpeg's -smooth option: last one or two scanlines would be
313 duplicates of the prior line unless the image height mod 16 was 1 or 2.
315 Repair minor problems in VMS, BCC, MC6 makefiles.
317 New configure script based on latest GNU Autoconf.
319 Correct the list of include files needed by MetroWerks C for ccommand().
321 Numerous small documentation updates.
324 Version 6 2-Aug-95
325 -------------------
327 Progressive JPEG support: library can read and write full progressive JPEG
328 files. A "buffered image" mode supports incremental decoding for on-the-fly
329 display of progressive images. Simply recompiling an existing IJG-v5-based
330 decoder with v6 should allow it to read progressive files, though of course
331 without any special progressive display.
333 New "jpegtran" application performs lossless transcoding between different
334 JPEG formats; primarily, it can be used to convert baseline to progressive
335 JPEG and vice versa. In support of jpegtran, the library now allows lossless
336 reading and writing of JPEG files as DCT coefficient arrays. This ability
337 may be of use in other applications.
339 Notes for programmers:
340 * We changed jpeg_start_decompress() to be able to suspend; this makes all
341 decoding modes available to suspending-input applications. However,
342 existing applications that use suspending input will need to be changed
343 to check the return value from jpeg_start_decompress(). You don't need to
344 do anything if you don't use a suspending data source.
345 * We changed the interface to the virtual array routines: access_virt_array
346 routines now take a count of the number of rows to access this time. The
347 last parameter to request_virt_array routines is now interpreted as the
348 maximum number of rows that may be accessed at once, but not necessarily
349 the height of every access.
352 Version 5b 15-Mar-95
353 ---------------------
355 Correct bugs with grayscale images having v_samp_factor > 1.
357 jpeg_write_raw_data() now supports output suspension.
359 Correct bugs in "configure" script for case of compiling in
360 a directory other than the one containing the source files.
362 Repair bug in jquant1.c: sometimes didn't use as many colors as it could.
364 Borland C makefile and jconfig file work under either MS-DOS or OS/2.
366 Miscellaneous improvements to documentation.
369 Version 5a 7-Dec-94
370 --------------------
372 Changed color conversion roundoff behavior so that grayscale values are
373 represented exactly. (This causes test image files to change.)
375 Make ordered dither use 16x16 instead of 4x4 pattern for a small quality
376 improvement.
378 New configure script based on latest GNU Autoconf.
379 Fix configure script to handle CFLAGS correctly.
380 Rename *.auto files to *.cfg, so that configure script still works if
381 file names have been truncated for DOS.
383 Fix bug in rdbmp.c: didn't allow for extra data between header and image.
385 Modify rdppm.c/wrppm.c to handle 2-byte raw PPM/PGM formats for 12-bit data.
387 Fix several bugs in rdrle.c.
389 NEED_SHORT_EXTERNAL_NAMES option was broken.
391 Revise jerror.h/jerror.c for more flexibility in message table.
393 Repair oversight in jmemname.c NO_MKTEMP case: file could be there
394 but unreadable.
397 Version 5 24-Sep-94
398 --------------------
400 Version 5 represents a nearly complete redesign and rewrite of the IJG
401 software. Major user-visible changes include:
402 * Automatic configuration simplifies installation for most Unix systems.
403 * A range of speed vs. image quality tradeoffs are supported.
404 This includes resizing of an image during decompression: scaling down
405 by a factor of 1/2, 1/4, or 1/8 is handled very efficiently.
406 * New programs rdjpgcom and wrjpgcom allow insertion and extraction
407 of text comments in a JPEG file.
409 The application programmer's interface to the library has changed completely.
410 Notable improvements include:
411 * We have eliminated the use of callback routines for handling the
412 uncompressed image data. The application now sees the library as a
413 set of routines that it calls to read or write image data on a
414 scanline-by-scanline basis.
415 * The application image data is represented in a conventional interleaved-
416 pixel format, rather than as a separate array for each color channel.
417 This can save a copying step in many programs.
418 * The handling of compressed data has been cleaned up: the application can
419 supply routines to source or sink the compressed data. It is possible to
420 suspend processing on source/sink buffer overrun, although this is not
421 supported in all operating modes.
422 * All static state has been eliminated from the library, so that multiple
423 instances of compression or decompression can be active concurrently.
424 * JPEG abbreviated datastream formats are supported, ie, quantization and
425 Huffman tables can be stored separately from the image data.
426 * And not only that, but the documentation of the library has improved
427 considerably!
430 The last widely used release before the version 5 rewrite was version 4A of
431 18-Feb-93. Change logs before that point have been discarded, since they
432 are not of much interest after the rewrite.