1 NOTE: This file was modified by The libjpeg-turbo Project to include only
2 information relevant to libjpeg-turbo.
4 CHANGE LOG for Independent JPEG Group's JPEG software
8 -----------------------
10 Add cjpeg -rgb option to create RGB JPEG files.
11 Using this switch suppresses the conversion from RGB
12 colorspace input to the default YCbCr JPEG colorspace.
13 Thank to Michael Koch for the initial suggestion.
15 Add option to disable the region adjustment in the transupp crop code.
16 Thank to Jeffrey Friedl for the suggestion.
19 Version 8b 16-May-2010
20 -----------------------
22 Repair problem in new memory source manager with corrupt JPEG data.
23 Thank to Ted Campbell and Samuel Chun for the report.
26 Version 8a 28-Feb-2010
27 -----------------------
29 Writing tables-only datastreams via jpeg_write_tables works again.
31 Support 32-bit BMPs (RGB image with Alpha channel) for read in cjpeg.
32 Thank to Brett Blackham for the suggestion.
36 ----------------------
38 Add sanity check in BMP reader module to avoid cjpeg crash for empty input
39 image (thank to Isaev Ildar of ISP RAS, Moscow, RU for reporting this error).
41 Add data source and destination managers for read from and write to
42 memory buffers. New API functions jpeg_mem_src and jpeg_mem_dest.
43 Thank to Roberto Boni from Italy for the suggestion.
47 ----------------------
49 New scaled DCTs implemented.
50 djpeg now supports scalings N/8 with all N from 1 to 16.
52 cjpeg -quality option has been extended for support of separate quality
53 settings for luminance and chrominance (or in general, for every provided
54 quantization table slot).
55 New API function jpeg_default_qtables() and q_scale_factor array in library.
57 Support arithmetic entropy encoding and decoding.
58 Added files jaricom.c, jcarith.c, jdarith.c.
60 jpegtran has a new "lossless" cropping feature.
62 Implement -perfect option in jpegtran, new API function
63 jtransform_perfect_transform() in transupp. (DP 204_perfect.dpatch)
65 Better error messages for jpegtran fopen failure.
66 (DP 203_jpegtran_errmsg.dpatch)
68 Fix byte order issue with 16bit PPM/PGM files in rdppm.c/wrppm.c:
69 according to Netpbm, the de facto standard implementation of the PNM formats,
70 the most significant byte is first. (DP 203_rdppm.dpatch)
72 Add -raw option to rdjpgcom not to mangle the output.
73 (DP 205_rdjpgcom_raw.dpatch)
75 Make rdjpgcom locale aware. (DP 201_rdjpgcom_locale.dpatch)
77 Add extern "C" to jpeglib.h.
78 This avoids the need to put extern "C" { ... } around #include "jpeglib.h"
79 in your C++ application. Defining the symbol DONT_USE_EXTERN_C in the
80 configuration prevents this. (DP 202_jpeglib.h_c++.dpatch)
83 Version 6b 27-Mar-1998
84 -----------------------
86 jpegtran has new features for lossless image transformations (rotation
87 and flipping) as well as "lossless" reduction to grayscale.
89 jpegtran now copies comments by default; it has a -copy switch to enable
90 copying all APPn blocks as well, or to suppress comments. (Formerly it
91 always suppressed comments and APPn blocks.) jpegtran now also preserves
92 JFIF version and resolution information.
94 New decompressor library feature: COM and APPn markers found in the input
95 file can be saved in memory for later use by the application. (Before,
96 you had to code this up yourself with a custom marker processor.)
98 There is an unused field "void * client_data" now in compress and decompress
99 parameter structs; this may be useful in some applications.
101 JFIF version number information is now saved by the decoder and accepted by
102 the encoder. jpegtran uses this to copy the source file's version number,
103 to ensure "jpegtran -copy all" won't create bogus files that contain JFXX
104 extensions but claim to be version 1.01. Applications that generate their
105 own JFXX extension markers also (finally) have a supported way to cause the
106 encoder to emit JFIF version number 1.02.
108 djpeg's trace mode reports JFIF 1.02 thumbnail images as such, rather
109 than as unknown APP0 markers.
111 In -verbose mode, djpeg and rdjpgcom will try to print the contents of
112 APP12 markers as text. Some digital cameras store useful text information
115 Handling of truncated data streams is more robust: blocks beyond the one in
116 which the error occurs will be output as uniform gray, or left unchanged
117 if decoding a progressive JPEG. The appearance no longer depends on the
118 Huffman tables being used.
120 Huffman tables are checked for validity much more carefully than before.
122 To avoid the Unisys LZW patent, djpeg's GIF output capability has been
123 changed to produce "uncompressed GIFs", and cjpeg's GIF input capability
124 has been removed altogether. We're not happy about it either, but there
125 seems to be no good alternative.
127 The configure script now supports building libjpeg as a shared library
128 on many flavors of Unix (all the ones that GNU libtool knows how to
129 build shared libraries for). Use "./configure --enable-shared" to
132 New jconfig file and makefiles for Microsoft Visual C++ and Developer Studio.
133 Also, a jconfig file and a build script for Metrowerks CodeWarrior
134 on Apple Macintosh. makefile.dj has been updated for DJGPP v2, and there
135 are miscellaneous other minor improvements in the makefiles.
137 jmemmac.c now knows how to create temporary files following Mac System 7
140 djpeg's -map switch is now able to read raw-format PPM files reliably.
142 cjpeg -progressive -restart no longer generates any unnecessary DRI markers.
144 Multiple calls to jpeg_simple_progression for a single JPEG object
145 no longer leak memory.
151 Library initialization sequence modified to detect version mismatches
152 and struct field packing mismatches between library and calling application.
153 This change requires applications to be recompiled, but does not require
154 any application source code change.
156 All routine declarations changed to the style "GLOBAL(type) name ...",
157 that is, GLOBAL, LOCAL, METHODDEF, EXTERN are now macros taking the
158 routine's return type as an argument. This makes it possible to add
159 Microsoft-style linkage keywords to all the routines by changing just
160 these macros. Note that any application code that was using these macros
161 will have to be changed.
163 DCT coefficient quantization tables are now stored in normal array order
164 rather than zigzag order. Application code that calls jpeg_add_quant_table,
165 or otherwise manipulates quantization tables directly, will need to be
166 changed. If you need to make such code work with either older or newer
167 versions of the library, a test like "#if JPEG_LIB_VERSION >= 61" is
170 djpeg's trace capability now dumps DQT tables in natural order, not zigzag
171 order. This allows the trace output to be made into a "-qtables" file
174 New system-dependent memory manager module for use on Apple Macintosh.
176 Fix bug in cjpeg's -smooth option: last one or two scanlines would be
177 duplicates of the prior line unless the image height mod 16 was 1 or 2.
179 Repair minor problems in VMS, BCC, MC6 makefiles.
181 New configure script based on latest GNU Autoconf.
183 Correct the list of include files needed by MetroWerks C for ccommand().
185 Numerous small documentation updates.
191 Progressive JPEG support: library can read and write full progressive JPEG
192 files. A "buffered image" mode supports incremental decoding for on-the-fly
193 display of progressive images. Simply recompiling an existing IJG-v5-based
194 decoder with v6 should allow it to read progressive files, though of course
195 without any special progressive display.
197 New "jpegtran" application performs lossless transcoding between different
198 JPEG formats; primarily, it can be used to convert baseline to progressive
199 JPEG and vice versa. In support of jpegtran, the library now allows lossless
200 reading and writing of JPEG files as DCT coefficient arrays. This ability
201 may be of use in other applications.
203 Notes for programmers:
204 * We changed jpeg_start_decompress() to be able to suspend; this makes all
205 decoding modes available to suspending-input applications. However,
206 existing applications that use suspending input will need to be changed
207 to check the return value from jpeg_start_decompress(). You don't need to
208 do anything if you don't use a suspending data source.
209 * We changed the interface to the virtual array routines: access_virt_array
210 routines now take a count of the number of rows to access this time. The
211 last parameter to request_virt_array routines is now interpreted as the
212 maximum number of rows that may be accessed at once, but not necessarily
213 the height of every access.
217 ---------------------
219 Correct bugs with grayscale images having v_samp_factor > 1.
221 jpeg_write_raw_data() now supports output suspension.
223 Correct bugs in "configure" script for case of compiling in
224 a directory other than the one containing the source files.
226 Repair bug in jquant1.c: sometimes didn't use as many colors as it could.
228 Borland C makefile and jconfig file work under either MS-DOS or OS/2.
230 Miscellaneous improvements to documentation.
236 Changed color conversion roundoff behavior so that grayscale values are
237 represented exactly. (This causes test image files to change.)
239 Make ordered dither use 16x16 instead of 4x4 pattern for a small quality
242 New configure script based on latest GNU Autoconf.
243 Fix configure script to handle CFLAGS correctly.
244 Rename *.auto files to *.cfg, so that configure script still works if
245 file names have been truncated for DOS.
247 Fix bug in rdbmp.c: didn't allow for extra data between header and image.
249 Modify rdppm.c/wrppm.c to handle 2-byte raw PPM/PGM formats for 12-bit data.
251 Fix several bugs in rdrle.c.
253 NEED_SHORT_EXTERNAL_NAMES option was broken.
255 Revise jerror.h/jerror.c for more flexibility in message table.
257 Repair oversight in jmemname.c NO_MKTEMP case: file could be there
264 Version 5 represents a nearly complete redesign and rewrite of the IJG
265 software. Major user-visible changes include:
266 * Automatic configuration simplifies installation for most Unix systems.
267 * A range of speed vs. image quality tradeoffs are supported.
268 This includes resizing of an image during decompression: scaling down
269 by a factor of 1/2, 1/4, or 1/8 is handled very efficiently.
270 * New programs rdjpgcom and wrjpgcom allow insertion and extraction
271 of text comments in a JPEG file.
273 The application programmer's interface to the library has changed completely.
274 Notable improvements include:
275 * We have eliminated the use of callback routines for handling the
276 uncompressed image data. The application now sees the library as a
277 set of routines that it calls to read or write image data on a
278 scanline-by-scanline basis.
279 * The application image data is represented in a conventional interleaved-
280 pixel format, rather than as a separate array for each color channel.
281 This can save a copying step in many programs.
282 * The handling of compressed data has been cleaned up: the application can
283 supply routines to source or sink the compressed data. It is possible to
284 suspend processing on source/sink buffer overrun, although this is not
285 supported in all operating modes.
286 * All static state has been eliminated from the library, so that multiple
287 instances of compression or decompression can be active concurrently.
288 * JPEG abbreviated datastream formats are supported, ie, quantization and
289 Huffman tables can be stored separately from the image data.
290 * And not only that, but the documentation of the library has improved
294 The last widely used release before the version 5 rewrite was version 4A of
295 18-Feb-93. Change logs before that point have been discarded, since they
296 are not of much interest after the rewrite.