1CHANGE LOG for Independent JPEG Group's JPEG software 2 3 4Version 7 27-Jun-2009 5---------------------- 6 7New scaled DCTs implemented. 8djpeg now supports scalings N/8 with all N from 1 to 16. 9cjpeg now supports scalings 8/N with all N from 1 to 16. 10Scaled DCTs with size larger than 8 are now also used for resolving the 11common 2x2 chroma subsampling case without additional spatial resampling. 12Separate spatial resampling for those kind of files is now only necessary 13for N>8 scaling cases. 14Furthermore, separate scaled DCT functions are provided for direct resolving 15of the common asymmetric subsampling cases (2x1 and 1x2) without additional 16spatial resampling. 17 18cjpeg -quality option has been extended for support of separate quality 19settings for luminance and chrominance (or in general, for every provided 20quantization table slot). 21New API function jpeg_default_qtables() and q_scale_factor array in library. 22 23Added -nosmooth option to cjpeg, complementary to djpeg. 24New variable "do_fancy_downsampling" in library, complement to fancy 25upsampling. Fancy upsampling now uses direct DCT scaling with sizes 26larger than 8. The old method is not reversible and has been removed. 27 28Support arithmetic entropy encoding and decoding. 29Added files jaricom.c, jcarith.c, jdarith.c. 30 31Straighten the file structure: 32Removed files jidctred.c, jcphuff.c, jchuff.h, jdphuff.c, jdhuff.h. 33 34jpegtran has a new "lossless" cropping feature. 35 36Implement -perfect option in jpegtran, new API function 37jtransform_perfect_transform() in transupp. (DP 204_perfect.dpatch) 38 39Better error messages for jpegtran fopen failure. 40(DP 203_jpegtran_errmsg.dpatch) 41 42Fix byte order issue with 16bit PPM/PGM files in rdppm.c/wrppm.c: 43according to Netpbm, the de facto standard implementation of the PNM formats, 44the most significant byte is first. (DP 203_rdppm.dpatch) 45 46Add -raw option to rdjpgcom not to mangle the output. 47(DP 205_rdjpgcom_raw.dpatch) 48 49Make rdjpgcom locale aware. (DP 201_rdjpgcom_locale.dpatch) 50 51Add extern "C" to jpeglib.h. 52This avoids the need to put extern "C" { ... } around #include "jpeglib.h" 53in your C++ application. Defining the symbol DONT_USE_EXTERN_C in the 54configuration prevents this. (DP 202_jpeglib.h_c++.dpatch) 55 56 57Version 6b 27-Mar-1998 58----------------------- 59 60jpegtran has new features for lossless image transformations (rotation 61and flipping) as well as "lossless" reduction to grayscale. 62 63jpegtran now copies comments by default; it has a -copy switch to enable 64copying all APPn blocks as well, or to suppress comments. (Formerly it 65always suppressed comments and APPn blocks.) jpegtran now also preserves 66JFIF version and resolution information. 67 68New decompressor library feature: COM and APPn markers found in the input 69file can be saved in memory for later use by the application. (Before, 70you had to code this up yourself with a custom marker processor.) 71 72There is an unused field "void * client_data" now in compress and decompress 73parameter structs; this may be useful in some applications. 74 75JFIF version number information is now saved by the decoder and accepted by 76the encoder. jpegtran uses this to copy the source file's version number, 77to ensure "jpegtran -copy all" won't create bogus files that contain JFXX 78extensions but claim to be version 1.01. Applications that generate their 79own JFXX extension markers also (finally) have a supported way to cause the 80encoder to emit JFIF version number 1.02. 81 82djpeg's trace mode reports JFIF 1.02 thumbnail images as such, rather 83than as unknown APP0 markers. 84 85In -verbose mode, djpeg and rdjpgcom will try to print the contents of 86APP12 markers as text. Some digital cameras store useful text information 87in APP12 markers. 88 89Handling of truncated data streams is more robust: blocks beyond the one in 90which the error occurs will be output as uniform gray, or left unchanged 91if decoding a progressive JPEG. The appearance no longer depends on the 92Huffman tables being used. 93 94Huffman tables are checked for validity much more carefully than before. 95 96To avoid the Unisys LZW patent, djpeg's GIF output capability has been 97changed to produce "uncompressed GIFs", and cjpeg's GIF input capability 98has been removed altogether. We're not happy about it either, but there 99seems to be no good alternative. 100 101The configure script now supports building libjpeg as a shared library 102on many flavors of Unix (all the ones that GNU libtool knows how to 103build shared libraries for). Use "./configure --enable-shared" to 104try this out. 105 106New jconfig file and makefiles for Microsoft Visual C++ and Developer Studio. 107Also, a jconfig file and a build script for Metrowerks CodeWarrior 108on Apple Macintosh. makefile.dj has been updated for DJGPP v2, and there 109are miscellaneous other minor improvements in the makefiles. 110 111jmemmac.c now knows how to create temporary files following Mac System 7 112conventions. 113 114djpeg's -map switch is now able to read raw-format PPM files reliably. 115 116cjpeg -progressive -restart no longer generates any unnecessary DRI markers. 117 118Multiple calls to jpeg_simple_progression for a single JPEG object 119no longer leak memory. 120 121 122Version 6a 7-Feb-96 123-------------------- 124 125Library initialization sequence modified to detect version mismatches 126and struct field packing mismatches between library and calling application. 127This change requires applications to be recompiled, but does not require 128any application source code change. 129 130All routine declarations changed to the style "GLOBAL(type) name ...", 131that is, GLOBAL, LOCAL, METHODDEF, EXTERN are now macros taking the 132routine's return type as an argument. This makes it possible to add 133Microsoft-style linkage keywords to all the routines by changing just 134these macros. Note that any application code that was using these macros 135will have to be changed. 136 137DCT coefficient quantization tables are now stored in normal array order 138rather than zigzag order. Application code that calls jpeg_add_quant_table, 139or otherwise manipulates quantization tables directly, will need to be 140changed. If you need to make such code work with either older or newer 141versions of the library, a test like "#if JPEG_LIB_VERSION >= 61" is 142recommended. 143 144djpeg's trace capability now dumps DQT tables in natural order, not zigzag 145order. This allows the trace output to be made into a "-qtables" file 146more easily. 147 148New system-dependent memory manager module for use on Apple Macintosh. 149 150Fix bug in cjpeg's -smooth option: last one or two scanlines would be 151duplicates of the prior line unless the image height mod 16 was 1 or 2. 152 153Repair minor problems in VMS, BCC, MC6 makefiles. 154 155New configure script based on latest GNU Autoconf. 156 157Correct the list of include files needed by MetroWerks C for ccommand(). 158 159Numerous small documentation updates. 160 161 162Version 6 2-Aug-95 163------------------- 164 165Progressive JPEG support: library can read and write full progressive JPEG 166files. A "buffered image" mode supports incremental decoding for on-the-fly 167display of progressive images. Simply recompiling an existing IJG-v5-based 168decoder with v6 should allow it to read progressive files, though of course 169without any special progressive display. 170 171New "jpegtran" application performs lossless transcoding between different 172JPEG formats; primarily, it can be used to convert baseline to progressive 173JPEG and vice versa. In support of jpegtran, the library now allows lossless 174reading and writing of JPEG files as DCT coefficient arrays. This ability 175may be of use in other applications. 176 177Notes for programmers: 178* We changed jpeg_start_decompress() to be able to suspend; this makes all 179decoding modes available to suspending-input applications. However, 180existing applications that use suspending input will need to be changed 181to check the return value from jpeg_start_decompress(). You don't need to 182do anything if you don't use a suspending data source. 183* We changed the interface to the virtual array routines: access_virt_array 184routines now take a count of the number of rows to access this time. The 185last parameter to request_virt_array routines is now interpreted as the 186maximum number of rows that may be accessed at once, but not necessarily 187the height of every access. 188 189 190Version 5b 15-Mar-95 191--------------------- 192 193Correct bugs with grayscale images having v_samp_factor > 1. 194 195jpeg_write_raw_data() now supports output suspension. 196 197Correct bugs in "configure" script for case of compiling in 198a directory other than the one containing the source files. 199 200Repair bug in jquant1.c: sometimes didn't use as many colors as it could. 201 202Borland C makefile and jconfig file work under either MS-DOS or OS/2. 203 204Miscellaneous improvements to documentation. 205 206 207Version 5a 7-Dec-94 208-------------------- 209 210Changed color conversion roundoff behavior so that grayscale values are 211represented exactly. (This causes test image files to change.) 212 213Make ordered dither use 16x16 instead of 4x4 pattern for a small quality 214improvement. 215 216New configure script based on latest GNU Autoconf. 217Fix configure script to handle CFLAGS correctly. 218Rename *.auto files to *.cfg, so that configure script still works if 219file names have been truncated for DOS. 220 221Fix bug in rdbmp.c: didn't allow for extra data between header and image. 222 223Modify rdppm.c/wrppm.c to handle 2-byte raw PPM/PGM formats for 12-bit data. 224 225Fix several bugs in rdrle.c. 226 227NEED_SHORT_EXTERNAL_NAMES option was broken. 228 229Revise jerror.h/jerror.c for more flexibility in message table. 230 231Repair oversight in jmemname.c NO_MKTEMP case: file could be there 232but unreadable. 233 234 235Version 5 24-Sep-94 236-------------------- 237 238Version 5 represents a nearly complete redesign and rewrite of the IJG 239software. Major user-visible changes include: 240 * Automatic configuration simplifies installation for most Unix systems. 241 * A range of speed vs. image quality tradeoffs are supported. 242 This includes resizing of an image during decompression: scaling down 243 by a factor of 1/2, 1/4, or 1/8 is handled very efficiently. 244 * New programs rdjpgcom and wrjpgcom allow insertion and extraction 245 of text comments in a JPEG file. 246 247The application programmer's interface to the library has changed completely. 248Notable improvements include: 249 * We have eliminated the use of callback routines for handling the 250 uncompressed image data. The application now sees the library as a 251 set of routines that it calls to read or write image data on a 252 scanline-by-scanline basis. 253 * The application image data is represented in a conventional interleaved- 254 pixel format, rather than as a separate array for each color channel. 255 This can save a copying step in many programs. 256 * The handling of compressed data has been cleaned up: the application can 257 supply routines to source or sink the compressed data. It is possible to 258 suspend processing on source/sink buffer overrun, although this is not 259 supported in all operating modes. 260 * All static state has been eliminated from the library, so that multiple 261 instances of compression or decompression can be active concurrently. 262 * JPEG abbreviated datastream formats are supported, ie, quantization and 263 Huffman tables can be stored separately from the image data. 264 * And not only that, but the documentation of the library has improved 265 considerably! 266 267 268The last widely used release before the version 5 rewrite was version 4A of 26918-Feb-93. Change logs before that point have been discarded, since they 270are not of much interest after the rewrite. 271