Changelogs » Pillow




- Changed Image size property to be read-only by default 3203

- Add warnings if image file identification fails due to lack of WebP support 3169
[radarhere, hugovk]

- Hide the Ghostscript progress dialog popup on Windows 3378

- Adding support to reading tiled and YcbCr jpeg tiffs through libtiff 3227

- Fixed None as TIFF compression argument 3310

- Changed GIF seek to remove previous info items 3324

- Improved PDF document info 3274

- Add line width parameter to rectangle and ellipse-based shapes 3094
[hugovk, radarhere]

- Fixed decompression bomb check in _crop 3313
[dinkolubina, hugovk]

- Added support to ImageDraw.floodfill for non-RGB colors 3377

- Tests: Avoid catching unexpected exceptions in tests 2203

- Use TextIOWrapper.detach() instead of NoCloseStream 2214

- Added transparency to matrix conversion 3205

- Added ImageOps pad method 3364

- Give correct extrema for I;16 format images 3359

- Added PySide2 3279

- Corrected TIFF tags 3369

- CI: Install CFFI and pycparser without any PYTHONOPTIMIZE 3374

- Read/Save RGB webp as RGB (instead of RGBX) 3298

- ImageDraw: Add line joints 3250

- Improved performance of ImageDraw floodfill method 3294

- Fix builds with --parallel 3272

- Add more raw Tiff modes (RGBaX, RGBaXX, RGBAX, RGBAXX) 3335

- Close existing WebP fp before setting new fp 3341

- Add orientation, compression and id_section as TGA save keyword arguments 3327

- Convert int values of RATIONAL TIFF tags to floats 3338
[radarhere, wiredfool]

- Fix code for PYTHONOPTIMIZE 3233

- Changed ImageFilter.Kernel to subclass ImageFilter.BuiltinFilter, instead of the other way around 3273

- Remove unused draw.draw_line, draw.draw_point and font.getabc methods 3232

- Tests: Added ImageFilter tests 3295

- Tests: Added ImageChops tests 3230
[hugovk, radarhere]

- AppVeyor: Download lib if not present in pillow-depends 3316

- Travis CI: Add Python 3.7 and Xenial 3234

- Docs: Added documentation for NumPy conversion 3301

- Depends: Update libimagequant to 2.12.1 3281

- Add three-color support to ImageOps.colorize 3242

- Tests: Add LA to TGA test modes 3222

- Skip outline if the draw operation fills with the same colour 2922

- Flake8 fixes 3173, 3380

- Avoid deprecated 'U' mode when opening files 2187



- Fixed saving a multiframe image as a single frame PDF 3137

- If a Qt version is already imported, attempt to use it first 3143

- Fix transform fill color for alpha images 3147

- TGA: Add support for writing RLE data 3186

- TGA: Read and write LA data 3178

- QuantOctree.c: Remove erroneous attempt to average over an empty range 3196

- Changed ICNS format tests to pass on OS X 10.11 3202

- Fixed bug in ImageDraw.multiline_textsize() 3114

- Added getsize_multiline support for PIL.ImageFont 3113

- Added ImageFile get_format_mimetype method 3190

- Changed mmap file pointer to use context manager 3216

- Changed ellipse point calculations to be more evenly distributed 3142

- Only extract first Exif segment 2946

- Tests: Test ImageDraw2, WalImageFile 3135, 2989

- Remove unnecessary 'if 0' code 3075

- Tests: Added GD tests 1817

- Fix collections ABCs DeprecationWarning in Python 3.7 3123

- unpack_from is faster than unpack of slice 3201

- Docs: Add coordinate system links and file handling links in documentation 3204, 3214

- Tests: TestFilePng: Fix test_save_l_transparency() 3182

- Docs: Correct argument name 3171

- Docs: Update CMake download URL 3166

- Docs: Improve Image.transform documentation 3164

- Fix transform fillcolor argument when image mode is RGBA or LA 3163

- Tests: More specific Exception testing 3158

- Add getrgb HSB/HSV color strings 3148

- Allow float values in getrgb HSL color string 3146

- AppVeyor: Upgrade to Python 2.7.15 and 3.4.4 3140

- AppVeyor: Upgrade to PyPy 6.0.0 3133

- Deprecate PILLOW_VERSION and VERSION 3090

- Support Python 3.7 3076

- Depends: Update freetype to 2.9.1, libjpeg to 9c, libwebp to 1.0.0 3121, 3136, 3108

- Build macOS wheels with Xcode 6.4, supporting older macOS versions 3068

- Fix _i2f compilation on some GCC versions 3067

- Changed encoderinfo to have priority over info when saving GIF images 3086

- Rename PIL.version to PIL._version and remove it from module 3083

- Enable background colour parameter on rotate 3057

- Remove unnecessary `if 1` directive 3072

- Remove unused Python class, Path 3070

- Fix dereferencing type-punned pointer will break strict-aliasing 3069



- Close fp before return in ImagingSavePPM 3061

- Added documentation for ICNS append_images 3051

- Docs: Move intro text below its header 3021

- CI: Rename appveyor.yml as .appveyor.yml 2978

- Fix TypeError for JPEG2000 parser feed 3042

- Certain corrupted jpegs can result in no data read 3023

- Add support for BLP file format 3007

- Simplify version checks 2998

- Fix "invalid escape sequence" warning on Python 3.6+ 2996

- Allow append_images to set .icns scaled images 3005

- Support appending to existing PDFs 2965

- Fix and improve efficient saving of ICNS on macOS 3004

- Build: Enable pip cache in AppVeyor build 3009

- Trim trailing whitespace 2985

- Docs: Correct reference to method 3000

- Rearrange ImageFilter classes into alphabetical order 2990

- Test: Remove duplicate line 2983

- Build: Update AppVeyor PyPy version 3003

- Tiff: Open 8 bit Tiffs with 5 or 6 channels, discarding extra channels 2938

- Readme: Added Twitter badge 2930

- Removed __main__ code from ImageCms 2942

- Test: Changed assert statements to unittest calls 2961

- Depends: Update libimagequant to 2.11.10, raqm to 0.5.0, freetype to 2.9 3036, 3017, 2957

- Remove _imaging.crc32 in favor of builtin Python crc32 implementation 2935

- Move Tk directory to src directory 2928

- Enable pip cache in Travis CI 2933

- Remove unused and duplicate imports 2927

- Docs: Changed documentation references to 2.x to 2.7 2921

- Fix memory leak when opening webp files 2974

- Setup: Fix "TypeError: 'NoneType' object is not iterable" for PPC and CRUX 2951

- Setup: Add libdirs for ppc64le and armv7l 2968



- Docs: Added docstrings from documentation 2914

- Test: Switch from nose to pytest 2815

- Rework Source directory layout, preventing accidental import of PIL. 2911

- Dynamically link libraqm 2753

- Removed scripts directory 2901

- TIFF: Run all compressed tiffs through libtiff decoder 2899

- GIF: Add disposal option when saving GIFs 2902
[linnil1, wiredfool]

- EPS: Allow for an empty line in EPS header data 2903

- PNG: Add support for sRGB and cHRM chunks, permit sRGB when no iCCP chunk present 2898

- Dependencies: Update Tk Tcl to 8.6.8 2905

- Decompression bomb error now raised for images 2x larger than a decompression bomb warning 2583

- Test: avoid random failure in test_effect_noise 2894

- Increased epsilon for due to Arch update. 2896

- Removed check parameter from _save in BmpImagePlugin, PngImagePlugin, ImImagePlugin, PalmImagePlugin, and PcxImagePlugin. 2873

- Make PngImagePlugin.add_text() zip argument type bool 2890

- Depends: Updated libwebp to 0.6.1 2880

- Remove unnecessary bool() calls in Image.registered_extensions and skipKnownBadTests 2891

- Fix count of BITSPERSAMPLE items in broken TIFF files 2883

- Fillcolor parameter for Image.Transform 2852

- Test: Display differences for test failures 2862

- Added executable flag to file with shebang line 2884

- Setup: Specify compatible Python versions for pip 2877

- Dependencies: Updated libimagequant to 2.11.4 2878

- Setup: Warn if trying to install for Py3.7 on Windows 2855

- Doc: Fonts can be loaded from a file-like object, not just filename 2861

- Add eog support for Ubuntu Image Viewer 2864

- Test: Test on 3.7-dev on Travis CI 2870

- Dependencies: Update libtiff to 4.0.9 2871

- Setup: Replace deprecated platform.dist with file existence check 2869

- Build: Fix on Debian 2853

- Docs: Correct error in ImageDraw documentation 2858

- Test: Drop Ubuntu Precise, Fedora 24, Fedora 25, add Fedora 27, Centos 7, Amazon v2 CI Support 2854, 2843, 2895, 2897

- Dependencies: Updated libimagequant to 2.11.3 2849

- Test: Fix to use tempfile 2841

- Replace PIL.OleFileIO deprecation warning with descriptive ImportError 2833

- WebP: Add support for animated WebP files 2761

- PDF: Set encoderinfo for images when saving multi-page PDF. Fixes 2804. 2805

- Allow the olefile dependency to be optional 2789

- GIF: Permit LZW code lengths up to 12 bits in GIF decode 2813

- Fix unterminated string and unchecked exception in _font_text_asBytes. 2825

- PPM: Use fixed list of whitespace, rather relying on locale, fixes 272. 2831

- Added support for generators when using append_images 2829, 2835

- Doc: Correct PixelAccess.rst 2824

- Depends: Update raqm to 0.3.0 2822

- Docs: Link to maintained version of aggdraw 2809

- Include license file in the generated wheel packages 2801

- Depends: Update openjpeg to 2.3.0 2791

- Add option to Makefile to build and install with C coverage 2781

- Add context manager support to ImageFile.Parser and PngImagePlugin.ChunkStream 2793

- ImageDraw.textsize: fix zero length error 2788
[wiredfool, hugovk]



- Fix warning on pointer cast in isblock 2775, 2778

- Doc: Added macOS High Sierra tested Pillow version 2777

- Use correct Windows handle type on 64 bit in imagingcms  2774

- 64 Bit Windows fix for block storage 2773

- Fix "expression result unused" warning 2764

- Add 16bit Read/Write and RLE read support to SgiImageFile 2769
[jbltx, wiredfool]

- Block & array hybrid storage 2738

- Common seek frame position check 1849

- Doc: Add note about aspect ratio to Image thumbnail script 2281

- Fix ValueError: invalid version number '1.0.0rc1' in scipy release candidate 2771

- Unfreeze requirements.txt 2766

- Test: ResourceWarning tests 2756

- Use n_frames to determine is_animated if possible 2315

- Doc: Corrected parameters in documentation 2768

- Avoid unnecessary Image operations 1891

- Added register_extensions method 1860

- Fix TIFF support for I;16S, I;16BS, and I;32BS rawmodes 2748

- Fixed doc syntax in ImageDraw 2752

- Fixed support for building on Windows/msys2. Added Appveyor CI coverage for python3 on msys2 2746

- Fix ValueError in Exif/Tiff IFD 2719

- Use pathlib2 for Path objects on Python < 3.4 2291

- Export only required properties in unsafe_ptrs 2740

- Alpha composite fixes 2709

- Faster Transpose operations, added 'Transverse' option 2730

- Deprecate ImageOps undocumented functions gaussian_blur, gblur, unsharp_mask, usm and box_blur in favor of ImageFilter implementations 2735

- Dependencies: Updated freetype to 2.8.1 2741

- Bug: Player skipped first image 2742

- Faster filter operations for Kernel, Gaussian, and Unsharp Mask filters 2679

- EPS: Add showpage to force rendering of some EPS images 2636

- DOC: Fix type of palette parameter in Image.quantize. 2703

- DOC: Fix Ico docs to match code 2712

- Added file pointer save to SpiderImagePlugin 2647

- Add targa version 2 footer 2713

- Removed redundant lines 2714

- Travis CI: Use default pypy/pypy3 2721

- Fix for SystemError when rendering an empty string, added in 4.2.0 2706

- Fix for memory leaks in font handling added in 4.2.0 2634

- Tests:  cleanup, more tests. Fixed WMF save handler 2689

- Removed debugging interface for Image.core.grabclipboard 2708

- Doc syntax fix 2710

- Faster packing and unpacking for RGB, LA, and related storage modes 2693

- Use RGBX rawmode for RGB JPEG images where possible 1989

- Remove palettes from non-palette modes in _new 2704

- Delete transparency info when convert'ing RGB/L to RGBA 2633

- Code tweaks to ease type annotations 2687

- Fixed incorrect use of 's' to byteslike object 2691

- Fix JPEG subsampling labels for subsampling=2  2698

- Region of interest (box) for resampling 2254

- Basic support for Termux (android) in 2684

- Bug: Fix Image.fromarray for numpy.bool type. 2683

- CI: Add Fedora 24 and 26 to Docker tests

- JPEG: Fix ZeroDivisionError when EXIF contains invalid DPI (0/0). 2667

- Depends: Updated openjpeg to 2.2.0 2669

- Depends: Updated Tk Tcl to 8.6.7 2668

- Depends: Updated libimagequant to 2.10.2 2660

- Test: Added test for ImImagePlugin tell() 2675

- Test: Additional tests for SGIImagePlugin 2659

- New Image.getchannel method 2661

- Remove unused im.copy2 and core.copy methods 2657

- Fast Image.merge() 2677

- Fast Image.split() 2676

- Fast image allocation 2655

- Storage cleanup 2654

- FLI: Use frame count from FLI header 2674

- Test: Test animated FLI file 2650

- Bug: Fixed uninitialized memory in bc5 decoding 2648

- Moved SgiImagePlugin save error to before the start of write operations 2646

- Move so isn't installed globally 2645

- Bug: Fix unexpected keyword argument 'align' 2641

- Add newlines to error message for clarity 2640

- Docs: Updated redirected URL 2637

- Bug: Fix JPEG DPI when EXIF is invalid 2632

- Bug: Fix for font getsize on empty string 2624

- Docs: Improved ImageDraw documentation 2626

- Docs: Corrected alpha_composite args documentation 2627

- Docs: added the description of the filename attribute to images.rst 2621

- Dependencies: Updated libimagequant to 2.10.1 2616

- PDF: Renamed parameter to not shadow built-in dict 2612



- CI: Fix version specification and test on CI for PyPy/Windows 2608



- Doc: Clarified documentation 2604

- CI: Amazon Linux and Centos6 docker images added to Travis CI 2585

- Image.alpha_composite added 2595

- Complex Text Support 2576
[ShamsaHamed, Fahad-Alsaidi, wiredfool]

- Added threshold parameter to ImageDraw.floodfill 2599

- Added dBATCH parameter to ghostscript command 2588

- JPEG: Adjust buffer size when icc_profile > MAXBLOCK 2596

- Specify Pillow Version in one place 2517

- CI: Change the owner of the TRAVIS_BUILD_DIR, fixing broken docker runs 2587

- Fix truncated PNG loading for some images, Fix memory leak on truncated PNG images. 2541, 2598

- Add decompression bomb check to Image.crop 2410

- ImageFile: Ensure that the `err_code` variable is initialized in case of exception. 2363

- Tiff: Support append_images for saving multipage TIFFs 2406

- Doc: Clarify that draft is only implemented for JPEG and PCD 2409

- Test: MicImagePlugin 2447

- Use round() instead of floor() to eliminate zero coefficients in resample 2558

- Remove deprecated code 2549

- Added append_images to PDF saving 2526

- Remove unused function core image function new_array 2548

- Remove unnecessary calls to dict.keys() 2551

- Add more tests and remove unused Draw.c code 2533

- Test: More tests for ImageMorph 2554

- Test: McIDAS area file 2552

- Update Feature Detection 2520

- CI: Update pypy on Travis CI 2573

- ImageMorph: Fix wrong expected size of MRLs read from disk 2561

- Docs: Update install docs for FreeBSD 2546

- Build: Ignore OpenJpeg 1.5 on FreeBSD 2544

- Remove 'not yet implemented' methods from PIL 1.1.4 2538

- Dependencies: Update FreeType to 2.8, LibTIFF to 4.0.8 and libimagequant to 2.9.1 2535 2537 2540

- Raise TypeError and not also UnboundLocalError in ImageFile.Parser() 2525

- Test: Use Codecov for coverage 2528

- Use PNG for 2527
[HinTak, wiredfool]

- Remove WITH_DEBUG compilation flag 2522

- Fix return value on parameter parse error in _webp.c 2521

- Set executable flag on scripts with shebang line 2295

- Flake8 2460

- Doc: Release Process Changes 2516

- CI: Added region for s3 deployment on appveyor 2515

- Doc: Updated references to point to existing files 2507

- Return copy on Image crop if crop dimensions match the image 2471

- Test: Optimize CI speed 2464, 2466



- Undef PySlice_GetIndicesEx, see 2493

- Fix for file with DPI in EXIF but not metadata, and XResolution is an int rather than tuple 2484

- Docs: Removed broken download counter badge 2487

- Docs: Fixed rst syntax error 2477



- Close files after loading if possible 2330
[homm, wiredfool]

- Fix Image Access to be reloadable when embedding the Python interpreter 2296
[wiredfool, cgohlke]

- Fetch DPI from EXIF if not specified in JPEG header 2449, 2472

- Removed winbuild checksum verification 2468

- Git: Set ContainerIO test file as binary 2469

- Remove superfluous import of FixTk 2455

- Fix import of tkinter/Tkinter 2456

- Pure Python Decoders, including Python decoder to fix for MSP images 1938
[wiredfool, hugovk]

- Reorganized GifImagePlugin, fixes 2314.  2374
[radarhere, wiredfool]

- Doc: Reordered operating systems in Compatibility Matrix 2436

- Test: Additional tests for BufrStub, Eps, Container, GribStub, IPTC, Wmf, XVThumb, ImageDraw, ImageMorph, ImageShow 2425

- Health fixes 2437

- Test: Correctness tests ContainerIO, XVThumbImagePlugin, BufrStubImagePlugin, GribStubImagePlugin, FitsStubImagePlugin, Hdf5StubImagePlugin, PixarImageFile, PsdImageFile 2443, 2442, 2441, 2440, 2431, 2430, 2428, 2427

- Remove unused imports 1822

- Replaced KeyError catch with dictionary get method 2424

- Test: Removed unrunnable code in test_image_toqimage 2415

- Removed use of spaces in TIFF kwargs names, deprecated in 2.7 1390

- Removed deprecated ImageDraw setink, setfill, setfont methods 2220

- Send unwanted subprocess output to /dev/null 2253

- Fix division by zero when creating 0x0 image from numpy array 2419

- Test: Added matrix convert tests 2381

- Replaced broken URL to 2413

- Removed unused private functions in and 2414

- Test: Fixed Qt tests for QT5 and saving 1 bit PNG 2394

- Test: docker builds for Arch and Debian Stretch 2394

- Updated libwebp to 0.6.0 on appveyor 2395

- More explicit error message when saving to a file with invalid extension 2399

- Docs: Update some http urls to https 2403

- Preserve aux/alpha channels when performing Imagecms transforms 2355

- Test linear and radial gradient effects 2382

- Test ImageDraw.Outline and and ImageDraw.Shape 2389

- Added PySide to ImageQt documentation 2392

- BUG: Empty image mode no longer causes a crash 2380

- Exclude .travis and contents from manifest 2386

- Remove 'MIT-like' from license 2145

- Tests: Add tests for several Image operations 2379

- PNG: Moved iCCP chunk before PLTE chunk when saving as PNG, restricted chunks known value/ordering 2347

- Default to inch-interpretation for missing ResolutionUnit in TiffImagePlugin 2365

- Bug: Fixed segfault when using ImagingTk on pypy Issue 2376, 2359.

- Bug: Fixed Integer overflow using ImagingTk on 32 bit platforms 2359
[wiredfool, QuLogic]

- Tests: Added docker images for testing alternate platforms. See also 2368

- Removed PIL 1.0 era TK readme that concerns Windows 95/NT 2360

- Prevent `nose -v` printing docstrings 2369

- Replaced absolute PIL imports with relative imports 2349

- Added context managers for file handling 2307

- Expose registered file extensions in Image 2343
[iggomez, radarhere]

- Make mode descriptor cache initialization thread-safe. 2351

- Updated Windows test dependencies: Freetype 2.7.1, zlib 1.2.11 2331, 2332, 2357

- Followed upstream pngquant packaging reorg to libimagquant 2354

- Fix invalid string escapes 2352

- Add test for crop operation with no argument 2333



- Refactor out postprocessing hack to load_end in PcdImageFile

- Add center and translate option to Image.rotate. 2328

- Test: Relax WMF test condition, fixes 2323.  2327

- Allow 0 size images, Fixes 2259, Reverts to pre-3.4 behavior.  2262

- SGI: Save uncompressed SGI/BW/RGB/RGBA files 2325

- Depends: Updated pngquant to 2.8.2 2319

- Test: Added correctness tests for opening SGI images 2324

- Allow passing a list or tuple of individual frame durations when saving a GIF 2298

- Unified different GIF optimize conditions 2196

- Build: Refactor dependency installation 2305

- Test: Add python 3.6 to travis, tox 2304

- Test: Fix coveralls coverage for Python+C 2300

- Remove executable bit and shebang from 2308
[jwilk, radarhere]

- PyPy: Buffer interface workaround 2294

- Test: Switch to Ubuntu Trusty 14.04 on Travis CI 2294

- Remove vendored version of olefile Python package in favor of upstream 2199

- Updated comments to use print as a function 2234

- Set executable flag on, and added shebang line 2282, 2277
[radarhere, homm]

- Test: Increase epsilon for FreeType 2.7 as rendering is slightly different. 2286

- Test: Faster assert_image_similar 2279

- Removed depreciated internal "stretch" method 2276

- Removed the handles_eof flag in decode.c 2223

- Tiff: Fix for writing Tiff to BytesIO using libtiff 2263

- Doc: Design docs 2269

- Test: Move tests requiring libtiff to test_file_libtiff 2273

- Update Maxblock heuristic 2275

- Fix for 2-bit palette corruption 2274
[pdknsk, wiredfool]

- Tiff: Update info.icc_profile when using libtiff reader. 2193

- Test: Fix bug in test_ifd_rational_save when libtiff is not available 2270

- ICO: Only save relevant sizes 2267

- ICO: Allow saving .ico files of 256x256 instead of 255x255 2265

- Fix TIFFImagePlugin ICC color profile saving. 2087

- Doc: Improved description of ImageOps.deform resample parameter 2256

- EMF: support negative bounding box coordinates 2249

- Close file if opened in WalImageFile 2216

- Use Image._new() instead of _makeself() 2248

- SunImagePlugin fixes 2241

- Use minimal scale for jpeg drafts 2240

- Updated dependency scripts to use FreeType 2.7, OpenJpeg 2.1.2, WebP 0.5.2 and Tcl/Tk 8.6.6 2235, 2236, 2237, 2290, 2302

- Fix "invalid escape sequence" bytestring warnings in Python 3.6 2186

- Removed support for Python 2.6 and Python 3.2 2192

- Setup: Raise custom exceptions when required/requested dependencies are not found 2213

- Use a context manager in to ensure file is always closed 2226

- Fixed bug in saving to fp-objects in Python >= 3.4 2227

- Use a context manager in ImageFont._load_pilfont() to ensure file is always closed 2232

- Use generator expressions instead of list comprehension 2225

- Close file after reading in ImagePalette.load() 2215

- Changed behaviour of default box argument for paste method to match docs 2211

- Add support for another BMP bitfield 2221

- Added missing top-level test __main__ 2222

- Replaced range(len()) 2197

- Fix for ImageQt Segfault, fixes 1370 2182

- Setup: Close file in after finished reading 2208

- Setup: optionally use pkg-config (when present) to detect dependencies 2074

- Search for tkinter first in builtins 2210

- Tests: Replace try/except/fail pattern with TestCase.assertRaises() 2200

- Tests: Remove unused, open files at top level of tests 2188

- Replace type() equality checks with isinstance 2184

- Doc: Move ICO out of the list of read-only file formats 2180

- Doc: Fix formatting, too-short title underlines and malformed table 2175

- Fix BytesWarnings 2172

- Use Integer division to eliminate deprecation warning. 2168

- Doc: Update compatibility matrix
[daavve, wiredfool]



- Fix Resample coefficient calculation 2162



- Allow lists as arguments for 2149

- Fix fix for map.c overflow 2151  (also in 3.3.3)



- Removed Image.core.open_ppm, added negative image size checks in 2146

- Windows build: fetch dependencies from pillow-depends 2095

- Add TIFF save_all writer. 2140
[lambdafu, vashek]

- Move libtiff fd duplication to _load_libtiff 2141

- Speed up GIF save optimization step, fixes 2093. 2133

- Fix for ImageCms Segfault, Issue 2037. 2131

- Make Image.crop an immediate operation, not lazy. 2138

- Skip empty values in ImageFileDirectory 2024

- Force reloading palette when using mmap in ImageFile. 2139

- Fix "invalid escape sequence" warning in Python 3.6 2136

- Update documentation about drafts 2137

- Converted documentation parameter format, comments to docstrings 2021

- Fixed typos 2128 2142

- Renamed references to OS X to macOS 2125 2130

- Use truth value when checking for progressive and optimize option on save 2115, 2129

- Convert DPI to ints when saving as JPEG 2102

- Added append_images parameter to GIF saving 2103

- Speedup paste with masks up to 80% 2015

- Rewrite DDS decoders in C, add DXT3 and BC7 decoders 2068

- Fix PyArg_ParseTuple format in getink() 2070

- Fix saving originally missing TIFF tags. 2111

- Allow pathlib.Path in on Python 2.7 2110

- Use modern base64 interface over deprecated 2121

- ImageColor.getrgb hexadecimal RGBA 2114

- Test fix for bigendian machines 2092

- Resampling lookups, trailing empty coefficients, precision 2008

- Add (un)packing between RGBA and BGRa 2057

- Added return for J2k (and fpx) Load to return a pixel access object 2061

- Skip failing numpy tests on Pypy <= 5.3.1 2090

- Show warning when trying to save RGBA image as JPEG 2010

- Respect pixel centers during transform 2022

- TOC for supported file formats 2056

- Fix conversion of bit images to numpy arrays Fixes 350, 2058

- Add ImageOps.scale to expand or contract a PIL image by a factor 2011

- Flake8 fixes 2050

- Updated freetype to 2.6.5 on Appveyor builds 2035

- PCX encoder fixes 2023, pr 2041

- Docs: Windows console prompts are > 2031

- Expose Pillow package version as PIL.__version__ 2027

- Add Box and Hamming filters for resampling 1959

- Retain a reference to core image object in PyAccess 2009



- Fix fix for map.c overflow 2151



- Fix negative image sizes in Storage.c 2146

- Fix integer overflow in map.c 2146

3.3.1 insecure


- Fix C90 compilation error for Tcl / Tk rewrite 2033

- Fix image loading when rotating by 0 deg 2052

3.3.0 insecure


- Fixed enums for Resolution Unit and Predictor in 1998

- Fix issue converting P mode to LA 1986

- Moved test_j2k_overflow to check_j2k_overflow, prevent DOS of our 32bit testing machines 1995

- Skip CRC checks in PNG files when LOAD_TRUNCATED_IMAGES is enabled 1991

- Added CMYK mode for opening EPS files 1826

- Docs: OSX build instruction clarification 1994

- Docs: Filter comparison table 1993

- Removal of pthread based Incremental.c, new interface for file decoders/encoders to access the python file. Fixes assorted J2k Hangs. 1934

- Skip unnecessary passes when resizing 1954

- Removed duplicate code in ImagePalette 1832

- test_imagecms: Reduce precision of extended info due to 32 bit machine precision 1990

- Binary Tiff Metadata/ICC profile. 1988

- Ignore large text blocks in PNG if LOAD_TRUNCATED_IMAGES is enabled 1970

- Replace index = index+1 in docs with +=1

- Skip extra 0xff00 in jpeg 1977

- Use bytearray for palette mutable storage 1985
[radarhere, wiredfool]

- Added additional uint modes for Image.fromarray, more extensive tests of fromarray 1984
[mairsbw, wiredfool]

- Fix for program importing PyQt4 when PyQt5 also installed 1942

- Changed depends/install_*.sh urls to point to github pillow-depends repo 1983

- Allow ICC profile from `encoderinfo` while saving PNGs 1909

- Fix integer overflow on ILP32 systems (32-bit Linux). 1975

- Change function declaration to match Tcl_CmdProc type 1966

- Integer overflow checks on all calls to *alloc 1781

- Change equals method on Image so it short circuits 1967

- Runtime loading of TCL/TK libraries, eliminating build time dependency. 1932

- Cleanup of transform methods 1941

- Fix "Fatal Python error: UNREF invalid object" in debug builds 1936

- Setup fixes for Alpine linux 1937

- Split resample into horizontal + vertical passes 1933

- Box blur with premultiplied alpha 1914

- Add libimagequant support in quantize() 1889

- Added internal Premultiplied luminosity (La) mode 1912

- Fixed point integer resample 1881

- Removed docs/BUILDME script 1924

- Moved comments to docstrings  1926

- Include Python.h before wchar.h so _GNU_SOURCE is set consistently 1906

- Updated example decoder in documentation 1899

- Added support for GIF comment extension 1896

- Removed support for pre- 1.5.2 list form of Image info in 1897

- Fix typos in TIFF tags 1918

- Skip tests that require libtiff if it is not installed 1893 (fixes 1866)

- Skip test when icc profile is not available, fixes 1887.  1892

- Make deprecated functions raise NotImplementedError instead of Exception. 1862, 1890
[daniel-leicht, radarhere]

- Replaced os.system with in 1879

- Corrected Image show documentation 1886

- Added check for executable permissions to ImageShow 1880

- Fixed tutorial code and added explanation 1877

- Added OS X support for ImageGrab grabclipboard 1837

- Combined duplicate code in ImageTk 1856

- Added --disable-platform-guessing option to build extension 1861

- Fixed loading Transparent PNGs with a transparent black color 1840

- Add support for LA mode in Image.fromarray 1865

- Make ImageFile load images in read-only mode 1864

- Added _accept hook for XVThumbImagePlugin 1853

- Test TIFF with LZW compression 1855, TGA RLE file 1854

- Improved SpiderImagePlugin help text 1863

- Updated Sphinx project description 1870

- Remove support for Python 3.0 from _imaging.c 1851

- Jpeg qtables are unsigned chars 1814, 1921

- Added additional EXIF tags 1841, TIFF Tags 1821

- Changed documentation to refer to ImageSequence Iterator 1833

- Fix Fedora prerequisites in installation docs, depends script 1842

- Added _accept hook for PixarImagePlugin 1843

- Removed outdated scanner classifier 1823

- Combined identical error messages in _imaging 1825

- Added debug option for to trace header and library finding 1790

- Fix doc building on travis 1820, 1844

- Fix for DIB/BMP images 1813, 1847

- Add PixarImagePlugin file extension 1809

- Catch struct.errors when verifying png files 1805

- SpiderImagePlugin: raise an error when seeking in a non-stack file 1794
[radarhere, jmichalon]

- Added support for 2/4 bpp Tiff grayscale images 1789

- Removed unused variable from selftest 1788

- Added warning for as_dict method (deprecated in 3.0.0) 1799

- Removed powf support for older Python versions 1784

- Health fixes 1625 1903

3.2.0 insecure


- Added install docs for Fedora 23 and FreeBSD 1729, 1739, 1792
[koobs, zandermartin, wiredfool]

- Fixed TIFF multiframe load when the frames have different compression types 1782
[radarhere, geka000]

- Added __copy__ method to Image 1772

- Updated dates in PIL license in OleFileIO README 1787

- Corrected Tiff tag names 1786

- Fixed documented name of JPEG property 1783

- Fixed UnboundLocalError when loading a corrupt jpeg2k file 1780

- Fixed integer overflow in path.c 1773
[wiredfool, nedwill]

- Added debug to command line help text for pilprint 1766

- Expose many more fields in ICC Profiles 1756

- Documentation changes, URL update, transpose, release checklist

- Fixed saving to nonexistant files specified by pathlib.Path objects 1748 (fixes 1747)

- Round Image.crop arguments to the nearest integer 1745 (fixes 1744)

- Fix uninitialized variable warning in _imaging.c:getink 1663 (fixes 486)

- Disable multiprocessing install on cygwin 1700 (fixes 1690)

- Fix the error reported when libz is not found 1764

- More general error check to avoid Symbol not found: _PyUnicodeUCS2_AsLatin1String on OS X 1761

- Added py35 to tox envlist 1724

- Fix EXIF tag name typos 1736
[zarlant, radarhere]

- Updated freetype to 2.6.3, Tk/Tcl to 8.6.5 and 8.5.19 1725, 1752

- Add a loader for the FTEX format from Independence War 2: Edge of Chaos 1688

- Improved alpha_composite documentation 1698

- Extend ImageDraw.text method to pass on multiline_text method specific arguments 1647

- Allow ImageSequence to seek to zero 1686

- ImageSequence Iterator is now an iterator 1649

- Updated windows test builds to jpeg9b 1673

- Fixed support for .gbr version 1 images, added support for version 2 in GbrImagePlugin 1653

- Clarified which YCbCr format is used 1677

- Added TiffTags documentation, Moved windows build documentation to winbuild/ 1667

- Add tests for OLE file based formats 1678

- Add TIFF IFD test 1671

- Add a basic DDS image plugin with more tests 1654
[jleclanche, hugovk, wiredfool]

- Fix incorrect conditional in encode.c 1638

3.1.2 insecure


- Fixed an integer overflow in Jpeg2KEncode.c causing a buffer overflow. CVE-2016-3076

3.1.1 insecure


- Fixed an integer overflow in Resample.c causing writes in the Python heap.

- Fixed a buffer overflow in PcdDecode.c causing a segfault when opening PhotoCD files. CVE-2016-2533

- Fixed a buffer overflow in FliDecode.c causing a segfault when opening FLI files. CVE-2016-0775

- Fixed a buffer overflow in TiffDecode.c causing an arbitrary amount of memory to be overwritten when opening a specially crafted invalid TIFF file. CVE-2016-0740

3.1.0 insecure


- Fixing test failures on Python 2.6/Windows 1633

- Limit metadata tags when writing using libtiff 1620

- Rolling back exif support to pre-3.0 format 1627

- Fix Divide by zero in Exif, add IFDRational class 1531

- Catch the IFD error near the source 1622

- Added release notes for 3.1.0 1623

- Updated spacing to be consistent between multiline methods 1624

- Let EditorConfig take care of some basic formatting 1489

- Restore gpsexif data to the v1 form 1619

- Add /usr/local include and library directories for freebsd 1613

- Updated installation docs for new versions of dependencies 1611

- Removed unrunnable test file 1610

- Changed register calls to use format property 1608

- Added field type constants to TiffTags 1596

- Allow saving RowsPerStrip with libtiff 1594

- Enabled conversion to numpy array for HSV images 1578

- Changed some urls in the docs to use https 1580

- Removed logger.exception from 1590

- Removed warnings module check 1587

- Changed arcs, chords and pie slices to use floats 1577

- Update unit test asserts 1584, 1598

- Fix command to invoke ghostscript for eps files 1478
[baumatron, radarhere]

- Consistent multiline text spacing 1574
[wiredfool, hugovk]

- Removed unused lines in BDFFontFile 1530

- Changed ImageQt import of Image 1560
[radarhere, ericfrederich]

- Throw TypeError if no cursors were found in .cur file 1556

- Fix crash in ImageTk.PhotoImage on win-amd64 1553

- ExtraSamples tag should be a SHORT, not a BYTE 1555

- Docs and code health fixes 1565 1566 1581 1586 1591 1621

- Updated freetype to 2.6.2 1564

- Updated WebP to 0.5.0 for Travis 1515 1609

- Fix missing 'version' key value in __array_interface__ 1519

- Replaced os.popen with subprocess.Popen to pilprint script 1523

- Catch OverflowError in SpiderImagePlugin 1545
[radarhere, MrShark]

- Fix the definition of icc_profile in TiffTags 1539

- Remove old _imagingtiff.c and pilplus stuff 1499

- Fix Exception when requiring jpeg 1501

- Dependency scripts for Debian and Ubuntu 1486

- Added Usage message to painter script 1482

- Add tag info for iccprofile, fixes 1462. 1465

- Added some requirements for make release-test 1451

- Flatten tiff metadata value SAMPLEFORMAT to initial value 1467 (fixes 1466)

- Fix handling of pathlib in 1464 (fixes 1460)

- Make tests more robust 1469

- Use correctly sized pointers for windows handle types 1458

3.0.0 insecure


- Check flush method existence for file-like object 1398
[mrTable, radarhere]

- Added PDF multipage saving 1445

- Removed deprecated code, Image.tostring, Image.fromstring, Image.offset, ImageDraw.setink, ImageDraw.setfill, ImageFileIO, ImageFont.FreeTypeFont and ImageFont.truetype `file` kwarg, ImagePalette private _make functions, ImageWin.fromstring and ImageWin.tostring 1343

- Load more broken images 1428

- Require zlib and libjpeg 1439

- Preserve alpha when converting from a QImage to a Pillow Image by using png instead of ppm 1429

- Qt needs 32 bit aligned image data 1430

- Tiff ImageFileDirectory rewrite 1419
[anntzer, wiredfool, homm]

- Removed spammy debug logging 1423

- Save as GiF89a with support for animation parameters 1384

- Correct convert matrix docs 1426

- Catch TypeError in _getexif 1414
[radarhere, wiredfool]

- Fix for UnicodeDecodeError in TiffImagePlugin 1416
[bogdan199, wiredfool]

- Dedup code in 1415

- Skip any number extraneous chars at the end of JPEG chunks 1337

- Single threaded build for pypy3, refactor 1413

- Fix loading of truncated images with LOAD_TRUNCATED_IMAGES enabled 1366

- Documentation update for concepts: bands 1406

- Add Solaris/SmartOS include and library directories 1356

- Improved handling of getink color 1387

- Disable compiler optimizations for topalette and tobilevel functions for all msvc versions 1402 (fixes 1357)

- Skip ImageFont_bitmap test if _imagingft C module is not installed 1409

- Add param documentation to ImagePalette 1381

- Corrected scripts path 1407

- Updated libtiff to 4.0.6 1405, 1421

- Updated Platform Support for Yosemite 1403

- Fixed infinite loop on truncated file 1401

- Check that images are L mode in ImageMorph methods 1400

- In tutorial of pasting images, add to mask text 1389

- Style/health fixes 1391, 1397, 1417, 1418

- Test on Python 3.5 dev and 3.6 nightly 1361

- Fix fast rotate operations 1373

- Added support for pathlib Path objects to open and save 1372

- Changed register calls to use format property 1333

- Added support for ImageGrab.grab to OS X 1367, 1443
[radarhere, hugovk]

- Fixed PSDraw stdout Python 3 compatibility 1365

- Added Python 3.3 to AppVeyor 1363

- Treat MPO with unknown header as base JPEG file 1350
[hugovk, radarhere]

- Added various tests 1330, 1344

- More ImageFont tests 1327

- Use logging instead of print 1207

2.9.0 insecure


- Added test for GimpPaletteFile 1324

- Merged gifmaker script to allow saving of multi-frame GIF images 1320

- Added is_animated property to multi-frame formats 1319

- Fixed ValueError in Python 2.6 1315 1316
[cgohlke, radarhere]

- Fixed tox test script path 1308

- Added width and height properties 1304

- Update tiff and tk tcl 8.5 versions 1303
[radarhere, wiredfool]

- Add functions to convert: Image <-> QImage; Image <-> QPixmap 1217
[radarhere, rominf]

- Remove duplicate code in gifmaker script 1294

- Multiline text in ImageDraw 1177
[allo-, radarhere]

- Automated Windows CI/build support 1278

- Removed support for Tk versions earlier than 8.4 1288

- Fixed polygon edge drawing 1255 (fixes 1252)

- Check prefix length in _accept methods 1267

- Register MIME type for BMP 1277

- Adjusted ImageQt use of unicode() for 2/3 compatibility 1218

- Identify XBM file created with filename including underscore 1230 (fixes 1229)

- Copy image when saving in GifImagePlugin 1231 (fixes 718)

- Removed support for FreeType 2.0 1247

- Added background saving to GifImagePlugin 1273

- Provide n_frames attribute to multi-frame formats 1261
[anntzer, radarhere]

- Add duration and loop set to GifImagePlugin 1172, 1269

- Ico files are little endian 1232

- Upgrade olefile from 0.30 to 0.42b 1226
[radarhere, decalage2]

- Setting transparency value to 0 when the tRNS contains only null byte(s) 1239

- Separated out feature checking from selftest 1233

- Style/health fixes

- Update WebP from 0.4.1 to 0.4.3 1235

- Release GIL during image load (decode) 1224

- Added icns save 1185

- Fix putdata memory leak 1196

- Keep user-specified ordering of icon sizes 1193

- Tiff: allow writing floating point tag values 1113

2.8.2 insecure


- Bug fix: Fixed Tiff handling of bad EXIF data

2.8.1 insecure


- Bug fix: Catch struct.error on invalid JPEG, fixes 1163.  1165
[wiredfool, hugovk]

2.8.0 insecure


- Fix 32-bit BMP loading (RGBA or RGBX) 1125

- Fix UnboundLocalError in ImageFile 1131

- Re-enable test image caching 982
[hugovk, homm]

- Fix: Cannot identify EPS images 1152 (fixes 1104)

- Configure setuptools to run nosetests, fixes 729

- Style/health fixes
[radarhere, hugovk]

- Add support for HTTP response objects to 1151

- Improve reference docs for PIL.ImageDraw.Draw.pieslice() 1145

- Added copy method font_variant() and accessible properties to truetype() 1123

- Fix ImagingEffectNoise 1128

- Remove unreachable code 1126

- Let Python do the endian stuff + tests 1121
[amoibos, radarhere]

- Fix webp decode memory leak 1114

- Fast path for opaque pixels in RGBa unpacker 1088

- Enable basic support for 'RGBa' raw encoding/decoding 1096

- Fix pickling L mode images with no palette, 1095

- iPython display hook 1091

- Adjust buffer size when quality=keep 1079 (fixes 148 again)

- Fix for corrupted bitmaps embedded in truetype fonts 1072
[jackyyf, wiredfool]

2.7.0 insecure


- Split Sane into a separate repo:

- Look for OS X and Linux fonts in common places 1054

- Fix CVE-2014-9601, potential PNG decompression DOS 1060

- Use underscores, not spaces, in TIFF tag kwargs 1044, 1058
[anntzer, hugovk]

- Update PSDraw for Python3, add tests 1055

- Use Bicubic filtering by default for thumbnails. Don't use Jpeg Draft mode for thumbnails 1029

- Fix MSVC compiler error: Use Py_ssize_t instead of ssize_t 1051

- Fix compiler error: MSVC needs variables defined at the start of the block 1048

- The GIF Palette optimization algorithm is only applicable to mode='P' or 'L' 993

- Use PySide as an alternative to PyQt4/5 1024

- Replace affine-based im.resize implementation with convolution-based im.stretch 997

- Replace Gaussian Blur implementation with iterated fast box blur. 961  Note: Radius parameter is interpreted differently than before.

- Better docs explaining import _imaging failure 1016, build 1017, mode 1018, PyAccess, PixelAccess objects 1019 Image.quantize 1020 and 1021

- Fix for saving TIFF image into an io.BytesIO buffer 1011

- Fix antialias compilation on debug versions of Python 1010

- Fix for Image.putdata segfault 1009

- Ico save, additional tests 1007

- Use PyQt4 if it has already been imported, otherwise prefer PyQt5 1003

- Speedup resample implementation up to 2.5 times 977

- Speed up rotation by using cache aware loops, added transpose to rotations 994

- Fix Bicubic interpolation 970

- Support for 4-bit greyscale TIFF images 980

- Updated manifest 957

- Fix PyPy 2.4 regression 958

- Webp Metadata Skip Test comments 954

- Fixes for things rpmlint complains about 942

2.6.2 insecure


- Fix CVE-2014-9601, potential PNG decompression DOS 1060

- Fix Regression in PyPy 2.4 in streamio  958

2.6.1 insecure


- Fix SciPy regression in Image.resize 945

- Fix manifest to include all test files.

2.6.0 insecure


- Relax precision of ImageDraw tests for x86, GimpGradient for PPC 930



- Use redistributable image for testing 884

- Use redistributable ICC profiles for testing, skip if not available 923

- Additional documentation for JPEG info and save options 922

- Fix JPEG Encoding memory leak when exif or qtables were specified 921

- Image.tobytes() and Image.tostring() documentation update 916 917

- On Windows, do not execute convert.exe without specifying path 912

- Fix msvc build error 911

- Fix for handling P + transparency -> RGBA conversions 904

- Retain alpha in ImageEnhance operations 909

- Jpeg2k Decode/encode memory leak fix 898
[joshware, wiredfool]

- EpsFilePlugin Speed improvements 886
[wiredfool, karstenw]

- Don't resize if already the right size 892

- Fix for reading multipage TIFFs 885
[kostrom, wiredfool]

- Correctly handle saving gray and CMYK JPEGs with quality=keep 857

- Correct duplicate Tiff Metadata and Exif tag values

- Windows fixes 871

- Fix TGA files with image ID field 856

- Fixed wrong P-mode of small, unoptimized L-mode GIF 843

- Fixed CVE-2014-3598, a DOS in the Jpeg2KImagePlugin
[Andrew Drake]

- Fixed CVE-2014-3589, a DOS in the IcnsImagePlugin
[Andrew Drake]

- Close open file handle before deleting 844

- Return Profile with Transformed Images 837

- Changed docstring to refer to the correct function 836

- Adding coverage support for C code tests 833

- PyPy performance improvements 821

- Added support for reading MPO files 822

- Added support for encoding and decoding iTXt chunks 818

- HSV Support 816

- Removed unusable

- Fix Scrambled XPM 808

- Doc cleanup

- Fix `ImageStat` docs 796

- Added docs for ExifTags 794

- More tests for CurImagePlugin, DcxImagePlugin, Effects.c, GimpGradientFile, ImageFont, ImageMath, ImagePalette, IptcImagePlugin, SpiderImagePlugin, SgiImagePlugin, XpmImagePlugin and _util

- Fix return value of FreeTypeFont.textsize() does not include font offsets 784

- Fix dispose calculations for animated GIFs 765

- Added class checking to Image __eq__ function 775
[radarhere, hugovk]

- Test PalmImagePlugin and method to skip known bad tests 776
[hugovk, wiredfool]

2.5.3 insecure


- Fixed CVE-2014-3598, a DOS in the Jpeg2KImagePlugin (backport)
[Andrew Drake]

2.5.2 insecure


- Fixed CVE-2014-3589, a DOS in the IcnsImagePlugin (backport)
[Andrew Drake]

2.5.1 insecure


- Fixed install issue if Multiprocessing.Pool is not available

- 32bit mult overflow fix 782

2.5.0 insecure


- Imagedraw rewrite 737
[terseus, wiredfool]

- Add support for multithreaded test execution 755

- Prevent shell injection 748
[mbrown1413, wiredfool]

- Support for Resolution in BMP files 734

- Fix error in for Python 3 744

- Pyroma fix and add Python 3.4 to setup metadata 742

- Top level flake8 fixes 741

- Remove obsolete Animated Raster Graphics (ARG) support 736

- Fix test_imagedraw failures 727

- Fix AttributeError: class Image has no attribute 'DEBUG' 726

- Fix msvc warning: 'inline' : macro redefinition 725

- Cleanup 654
[dvska, hugovk, wiredfool]

- 16-bit monochrome support for JPEG2000 730

- Fixed

- Support JPEG qtables 677

- Add binary morphology addon
[dov, wiredfool]

- Decompression bomb protection 674

- Put images in a single directory 708

- Support OpenJpeg 2.1 681
[al45tair, wiredfool]

- Remove unistd.h include for all platforms 704

- Use unittest for tests

- ImageCms fixes

- Added more ImageDraw tests

- Added tests for Spider files

- Use libtiff to write any compressed tiff files 669

- Support for pickling Image objects

- Fixed resolution handling for EPS thumbnails 619

- Fixed rendering of some binary EPS files (Issue 302)

- Rename variables not to use built-in function names 670

- Ignore junk JPEG markers

- Change default interpolation for Image.thumbnail to Image.ANTIALIAS

- Add tests and fixes for saving PDFs

- Remove transparency resource after P->RGBA conversion

- Clean up preprocessor cruft for Windows 652

- Adjust Homebrew freetype detection logic 656

- Added Image.close, context manager support

- Added support for 16 bit PGM files

- Updated OleFileIO to version 0.30 from upstream 618

- Added support for additional TIFF floating point format

- Have the tempfile use a suffix with a dot

- Fix variable name used for transparency manipulations 604

2.4.0 insecure


- Indexed Transparency handled for conversions between L, RGB, and P modes 574 (fixes 510)

- Conversions enabled from RGBA->P 574 (fixes 544)

- Improved icns support 565

- Fix libtiff leaking open files 580 (fixes 526)

- Fixes for Jpeg encoding in Python 3 578 (fixes 577)

- Added support for JPEG 2000 547

- Add more detailed error messages to 566

- Avoid conflicting _expand functions in PIL & MINGW, fixes 538

- Merge from Philippe Lagadec’s OleFileIO_PL fork 512

- Fix ImageColor.getcolor 534

- Make ICO files work with the ImageFile.Parser interface 525 (fixes 522)

- Handle 32bit compiled python on 64bit architecture 521

- Fix support for characters >128 using .pcf or .pil fonts in Py3k 517 (fixes 505)

- Skip CFFI test earlier if it's not installed 516

- Fixed opening and saving odd sized .pcx files 535 (fixes 523)

- Fixed palette handling when converting from mode P->RGB->P

- Fixed saving mode P image as a PNG with transparency = palette color 0

- Improve heuristic used when saving progressive and optimized JPEGs with high quality values 504

- Fixed DOS with invalid palette size or invalid image size in BMP file

- Added support for BMP version 4 and 5
[eddwardo, wiredfool]

- Fix segfault in getfont when passed a memory resident font

- Fix crash on Saving a PNG when icc-profile is None 496

- Cffi+Python implementation of the PixelAccess object

- PixelAccess returns unsigned ints for I16 mode

- Minor patch on booleans + Travis 474

- Look in multiarch paths in GNU platforms 511

- Add arch support for pcc64, s390, s390x, armv7l, aarch64 475

- Add arch support for ppc

- Correctly quote file names for WindowsViewer command

- Prefer homebrew freetype over X11 freetype (but still allow both) 466

2.3.2 insecure


- Fixed CVE-2014-3589, a DOS in the IcnsImagePlugin (backport)
[Andrew Drake]

2.3.1 insecure


- Fix insecure use of tempfile.mktemp (CVE-2014-1932 CVE-2014-1933)

2.3.0 insecure


- Stop leaking filename parameter passed to getfont 459

- Report availability of LIBTIFF during setup and selftest

- Fix msvc build error C1189: "No Target Architecture" 460

- Fix memory leak in font_getsize

- Correctly prioritize include and library paths 442

- Image.point fixes for numpy.array and docs 441

- Save the transparency header by default for PNGs 424

- Support for PNG tRNS header when converting from RGB->RGBA 423

- PyQT5 Support 418

- Updates for saving color tiffs w/compression using libtiff 417

- 2gigapix image fixes and redux

- Save arbitrary tags in Tiff image files 369

- Quote filenames and title before using on command line 398

- Fixed to return properly 399

- Documentation fixes

- Fixed memory leak saving images as webp when webpmux is available 429

- Fix compiling with FreeType 2.5.1 427

- Adds directories for NetBSD 411

- Support RGBA TIFF with missing ExtraSamples tag 393

- Lossless WEBP Support 390

- Take compression as an option in the save call for tiffs 389

- Add support for saving lossless WebP. Just pass 'lossless=True' to save() 386

- LCMS support upgraded from version 1 to version 2 380 (fixes 343)

- Added more raw decoder 16 bit pixel formats 379

- Document remaining Image* modules listed in PIL handbook

- Document ImageEnhance, ImageFile, ImageFilter, ImageFont, ImageGrab, ImageMath, and ImageOps

- Port and update docs for Image, ImageChops, ImageColor, and ImageDraw

- Move or copy content from README.rst to docs/

- Respect CFLAGS/LDFLAGS when searching for headers/libs

- Port PIL Handbook tutorial and appendices

- Alpha Premultiplication support for transform and resize 364

- Fixes to make Pypy 2.1.0 work on Ubuntu 12.04/64 359

2.2.2 insecure


- Fix compiling with FreeType 2.5.1 427

2.2.1 insecure


- Error installing Pillow 2.2.0 on Mac OS X (due to hard dep on brew) 357 (fixes 356)

2.2.0 insecure


- Bug in image transformations resulting from uninitialized memory 348 (fixes 254)

- Fix for encoding of b_whitespace 346 (similar to closed issue 272)

- Add numpy array interface support for 16 and 32 bit integer modes 347 (fixes 273)

- Partial fix for 290: Add preliminary support for TIFF tags.

- Fix 251 and 326: circumvent classification of pngtest_bad.png as malware

- Add typedef uint64_t for MSVC 339

- better support for C_INCLUDE_PATH, LD_RUN_PATH, etc. 336 (fixes 329)

- _imagingcms.c: include windef.h to fix build issue on MSVC 335 (fixes 328)

- Automatically discover homebrew include/ and lib/ paths on OS X 330

- Fix bytes which should be bytearray 325

- Add respective paths for C_INCLUDE_PATH, LD_RUN_PATH (rpath) to build
if specified as environment variables 324

- Fix 312 + gif optimize improvement

- Be more tolerant of tag read failures 320

- Catch truncated zTXt errors 321 (fixes 318)

- Fix IOError when saving progressive JPEGs 313

- Add RGBA support to ImageColor 309

- Test for `str`, not `"utf-8"` 306 (fixes 304)

- Fix missing import os in 303

- Added missing exif tags 300

- Fail on all import errors 298, 299 (fixes 297)
[macfreek, wiredfool]

- Fixed Windows fallback (wasn't using correct file in Windows fonts) 295

- Moved ImageFile and ImageFileIO comments to docstrings 293

- Restore compatibility with ISO C 289

- Use correct format character for C int type 288

- Allocate enough memory to hold pointers in encode.c 287

- Fillorder double shuffling bug when FillOrder ==2 and decoding using libtiff 284 (fixes 279)

- Moved Image module comments to docstrings.

- Add 16-bit TIFF support 277 (fixes 274)

- Ignore high ascii characters in string.whitespace 276 (fixes 272)

- Added clean/build to tox to make it behave like Travis 275

- Adding support for metadata in webp images 271

2.1.0 insecure


- Add /usr/bin/env python shebangs to all scripts in /Scripts 197

- Add several TIFF decoders and encoders 268

- Added support for alpha transparent webp images.

- Adding Python 3 support for StringIO.

- Adding Python3 basestring compatibility without changing basestring.

- Fix webp encode errors on win-amd64 259

- Better fix for ZeroDivisionError in for image.size height is 1 267

- Better support for ICO images.

- Changed PY_VERSION_HEX 190 (fixes 166)

- Changes to put everything under the PIL namespace 191

- Changing StringIO to BytesIO.

- Cleanup whitespace.

- Don't skip 'import site' on initialization when running tests for inplace builds.

- Enable warnings for test suite 227

- Fix for ZeroDivisionError in for image.size == (1,1) 255

- Fix for if isinstance(filter, collections.Callable) crash. Python bug 7624 on <2.6.6

- Remove double typedef declaration 194 (fixes 193)

- Fix msvc compile errors (230).

- Fix rendered characters have been chipped for some TrueType fonts

- Fix usage of script 184

- Fresh start for docs, generated by sphinx-apidoc.

- Introduce --enable-x and fail if it is given and x is not available.

- Partial work to add a wrapper for WebPGetFeatures to correctly support 220 (fixes 204)

- Significant performance improvement of `alpha_composite` function 156

- Support explicitly disabling features via --disable-* options 240

- Support --installed, fixes 263

- Transparent WebP Support 220 (fixes 204)
[euangoddard, wiredfool]

- Use PyCapsule for py3.1 238 (fixes 237)

- Workaround for: in 3.2.x < 3.2.4 and 3.3.0.

2.0.0 insecure


.. Note:: Special thanks to Christoph Gohlke and Eric Soroos for assisting with a pre-PyCon 2013 release!

- Many other bug fixes and enhancements by many other people.

- Add Python 3 support. (Pillow >= 2.0.0 supports Python 2.6, 2.7, 3.2, 3.3. Pillow < 2.0.0 supports Python 2.4, 2.5, 2.6, 2.7.)

- Add PyPy support (experimental, please see 67)

- Add WebP support 96

- Add Tiff G3/G4 support (experimental)

- Backport PIL's PNG/Zip improvements 95, 97

- Various 64-bit and Windows fixes.

- Add testing suite.
[cgohlke, fluggo]

- Added support for PNG images with transparency palette.

1.7.8 insecure


- Removed that made tests of other packages fail.

- Fix opening psd files with RGBA layers when A mode is not of type 65535 but 3.
Fixes 3

1.7.7 insecure


- UNDEF more types before including windows headers

1.7.6 insecure


- Bug fix: freetype not found on Mac OS X with case-sensitive filesystem

- Bug fix: Backport fix to split() after open() (regression introduced in PIL 1.1.7).

1.7.5 insecure


- Fix for sys.platform = "linux3"

- Package cleanup and additional documentation

1.7.4 insecure


- Fix brown bag release

1.7.3 insecure


- Fix : resize need int values, append int conversion in thumbnail method

1.7.2 insecure


- Bug fix: Python 2.4 compat

1.7.1 insecure


- More multi-arch support
[SteveM, regebro, barry, aclark4life]

1.7.0 insecure


- Add support for multi-arch library directory /usr/lib/x86_64-linux-gnu

1.6 insecure


- Bug fix: /usr/x11/include should be added to include_dirs not library_dirs

- Doc fixes

1.5 insecure


- Module and package fixes

1.4 insecure


- Doc fixes

1.3 insecure


- Add support for /lib64 and /usr/lib64 library directories on Linux

- Doc fixes

1.2 insecure


- On OS X also check for freetype2 in the X11 path

- Doc fixes


This section may not be fully complete.  For changes since this file
was last updated, see the repository revision history:

(1.1.7 final)

+ Set GIF loop info property to the number of iterations if a NETSCAPE
loop extension is present, instead of always setting it to 1 (from
Valentino Volonghi).


+ Improved PNG compression (from Alexey Borzenkov).

+ Read interlaced PNG files (from Conrado Porto Lopes Gouvêa)

+ Added various TGA improvements from Alexey Borzenkov, including
support for specifying image orientation.

+ Bumped block threshold to 16 megabytes, made size estimation a bit
more accurate.  This speeds up allocation of large images.

+ Fixed rounding error in ImagingDrawWideLine.

"gormish" writes: ImagingDrawWideLine() in Draw.c has a bug in every
version I've seen, which leads to different width lines depending on
the order of the points in the line. This is especially bad at some
angles where a 'width=2' line can completely disappear.

+ Added support for RGBA mode to the SGI module (based on code by
Karsten Hiddemann).

+ Handle repeated IPTC tags (adapted from a patch by Eric Bruning).

Eric writes: According to the specification, some IPTC tags can be
repeated, e.g., tag 2:25 (keywords). PIL 1.1.6 only retained the last
instance of that tag. Below is a patch to store all tags. If there are
multiple tag instances, they are stored in a (python) list. Single tag
instances remain as strings.

+ Fixed potential crash in ImageFilter for small target images
(reported by Zac Burns and Daniel Fetchinson).

+ Use BMP instead of JPEG as temporary show format on Mac OS X.

+ Fixed putpixel/new for I;16 with colors > 255.

+ Added integer power support to ImagingMath.

+ Added limited support for I;16L mode (explicit little endian).

+ Moved WMF support into Image.core; enable WMF rendering by default
if renderer is available.

+ Mark the ARG plugin as obsolete.

+ Added version query mechanism to ImageCms and ImageFont, for

+ Added (experimental) ImageCms function for fetching the ICC profile
for the current display (currently Windows only).

Added HWND/HDC support to ImageCms.get_display_profile().

+ Added WMF renderer (Windows only).

+ Added ImagePointHandler and ImageTransformHandler mixins; made
ImageCmsTransform work with im.point.

+ Fixed potential endless loop in the XVThumbnail reader (from Nikolai

+ Added Kevin Cazabon's pyCMS package.

The C code has been moved to _imagingcms.c, the Python interface
module is installed as PIL.ImageCMS.

Added support for in-memory ICC profiles.

Unified buildTransform and buildTransformFromOpenProfiles.

The profile can now be either a filename, a profile object, or a
file-like object containing an in-memory profile.

Additional fixes from Florian Böch:

Very nice - it just needs LCMS flags support so we can use black
point compensation and softproofing :) See attached patches.  They
also fix a naming issue which could cause confusion - display
profile (ImageCms wording) actually means proof profile (lcms
wording), so I changed variable names and docstrings where
applicable. Patches are tested under Python 2.6.

+ Improved support for layer names in PSD files (from Sylvain Baubeau)

Sylvain writes: I needed to be able to retrieve the names of the
layers in a PSD files. But didn't do the job so I
wrote this very small patch.

+ Improved RGBA support for ImageTk for 8.4 and newer (from Con

This replaces the slow run-length based encoding model with true
compositing at the Tk level.

+ Added support for 16- and 32-bit images to McIdas loader.

Based on file samples and stand-alone reader code provided by Craig

+ Added ImagePalette support to putpalette.

+ Fixed problem with incremental parsing of PNG files.

+ Make report non-zero status on failure (from Mark

+ Add big endian save support and multipage infrastructure to the TIFF
writer (from Sebastian Haase).

+ Handle files with GPS IFD but no basic EXIF IFD (reported by Kurt

+ Added zTXT support (from Andrew Kuchling via Lowell Alleman).

+ Fixed potential infinite loop bug in ImageFont (from Guilherme Polo).

+ Added sample ICC profiles (from Kevin Cazabon)

+ Fixed array interface for I, F, and RGBA/RGBX images.

+ Added Chroma subsampling support for JPEG (from Justin Huff).

Justin writes: Attached is a patch (against PIL 1.1.6) to provide
control over the chroma subsampling done by the JPEG encoder.  This
is often useful for reducing compression artifacts around edges of
clipart and text.

+ Added USM/Gaussian Blur code from Kevin Cazabon.

+ Fixed bug w. uninitialized image data when cropping outside the
source image.

+ Use ImageShow to implement the method.

Most notably, this picks the 'display' utility when available.  It
also allows application code to register new display utilities via
the ImageShow registry.

+ Release the GIL in the PNG compressor (from Michael van Tellingen).

+ Revised JPEG CMYK handling.

Always assume Adobe behaviour, both when reading and writing (based on
a patch by Kevin Cazabon, and test data by Tim V. and Charlie Clark, and
additional debugging by Michael van Tellingen).

+ Support for preserving ICC profiles (by Florian Böch via Tim Hatch).

Florian writes:

It's a beta, so still needs some testing, but should allow you to:
- retain embedded ICC profiles when saving from/to JPEG, PNG, TIFF.
Existing code doesn't need to be changed.
- access embedded profiles in JPEG, PNG, PSD, TIFF.

It also includes patches for TIFF to retain IPTC, Photoshop and XMP
metadata when saving as TIFF again, read/write TIFF resolution
information correctly, and to correct inverted CMYK JPEG files.

+ Fixed potential memory leak in median cut quantizer (from Evgeny Salmin).

+ Fixed OverflowError when reading upside-down BMP images.

+ Added resolution save option for PDF files.

Andreas Kostyrka writes: I've included a patched
based on 1.1.6 as included in Ubuntu, that supports a "resolution"
save option. Not great, but it makes the PDF saving more useful by
allowing PDFs that are not exactly 72dpi.

+ Look for Tcl/Tk include files in version-specific include directory
(from Encolpe Degoute).

+ Fixed grayscale rounding error in ImageColor.getcolor (from Tim

+ Fixed calculation of mean value in ImageEnhance.Contrast (reported
by "roop" and Scott David Daniels).

+ Fixed truetype positioning when first character has a negative left
bearing (from Ned Batchelder):

Ned writes: In PIL 1.1.6, ImageDraw.text will position the string
incorrectly if the first character has a negative left bearing.  To
see the problem, show a string like "///" in an italic font.  The
first slash will be clipped at the left, and the string will be

+ Fixed resolution unit bug in tiff reader/writer (based on code by
Florian Höch, Gary Bloom, and others).

+ Added simple transparency support for RGB images (reported by
Sebastian Spaeth).

+ Added support for Unicode filenames in ImageFont.truetype (from Donn

+ Fixed potential crash in ImageFont.getname method (from Donn Ingle).

+ Fixed encoding issue in PIL/WalImageFile (from Santiago M. Mola).


+ Fixed some 64-bit compatibility warnings for Python 2.5.

+ Added threading support for the Sane driver (from Abel Deuring).


+ Added experimental "floodfill" function to the ImageDraw module
(based on code by Eric Raymond).

+ The default arguments for "frombuffer" doesn't match "fromstring"
and the documentation; this is a bug, and will most likely be fixed
in a future version.  In this release, PIL prints a warning message
instead.  To silence the warning, change any calls of the form
"frombuffer(mode, size, data)" to

frombuffer(mode, size, data, "raw", mode, 0, 1)

+ Added "fromarray" function, which takes an object implementing the
NumPy array interface and creates a PIL Image from it. (from Travis

+ Added NumPy array interface support (__array_interface__) to the
Image class (based on code by Travis Oliphant).

This allows you to easily convert between PIL image memories and
NumPy arrays:

import numpy, Image

im ='hopper.jpg')

a = numpy.asarray(im)  a is readonly

im = Image.fromarray(a)

+ Fixed CMYK polarity for JPEG images, by treating all images as
"Adobe CMYK" images. (thanks to Cesare Leonardi and Kevin Cazabon
for samples, debugging, and patches).


+ Added 'expand' option to the Image 'rotate' method.  If true, the
output image is made large enough to hold the entire rotated image.

+ Changed the ImageDraw 'line' method to always draw the last pixel in
a polyline, independent of line angle.

+ Fixed bearing calculation and clipping in the ImageFont truetype
renderer; this could lead to clipped text, or crashes in the low-
level _imagingft module.  (based on input from Adam Twardoch and

+ Added ImageQt wrapper module, for converting PIL Image objects to
QImage objects in an efficient way.

+ Fixed 'getmodebands' to return the number of bands also for "PA"
and "LA" modes.  Added 'getmodebandnames' helper that return the
band names.


+ Added float/double support to the TIFF loader (from Russell

+ Fixed broken use of realloc() in path.c (from Jan Matejek)

+ Added save support for Spider images (from William Baxter).

+ Fixed broken 'paste' and 'resize' operations in pildriver
(from Bill Janssen).

+ Added support for duplex scanning to the Sane interface (Abel


+ Fixed a memory leak in "convert(mode)", when converting from
L to P.

+ Added pixel access object.  The "load" method now returns a
access object that can be used to directly get and set pixel
values, using ordinary [x, y] notation:

pixel = im.load()
v = pixel[x, y]
pixel[x, y] = v

If you're accessing more than a few pixels, this is a lot
faster than using getpixel/putpixel.

+ Fixed building on Cygwin (from Miki Tebeka).

+ Fixed "point(callable)" on unloaded images (reported by Håkan

+ Fixed size bug in ImageWin.ImageWindow constructor (from Victor

+ Fixed ImageMath float() and int() operations for Python 2.4
(reported by Don Rozenberg).

+ Fixed "RuntimeError: encoder error -8 in tostring" problem for
wide "RGB", "I", and "F" images.

+ Fixed line width calculation.


+ Fixed byte order issue in Image.paste(ink) (from Ka-Ping Yee).

+ Fixed off-by-0.5 errors in the ANTIALIAS code (based on input
from Douglas Bagnall).

+ Added buffer interface support to the Path constructor.  If
a buffer is provided, it is assumed to contain a flat array
of float coordinates (e.g. array.array('f', seq)).

+ Added new ImageMath module.

+ Fixed ImageOps.equalize when used with a small number of distinct
values (reported by David Kirtley).

+ Fixed potential integer division in PSDraw.image (from Eric Etheridge).




+ Added experimental PERSPECTIVE transform method (from Jeff Breiden-


+ Make sure "thumbnail" never generates zero-wide or zero-high images
(reported by Gene Skonicki)

+ Fixed a "getcolors" bug that could result in a zero count for some
colors (reported by Richard Oudkerk).

+ Changed default "convert" palette to avoid "rounding errors" when
round-tripping white source pixels (reported by Henryk Gerlach and
Jeff Epler).


+ Don't crash in "quantize" method if the number of colors requested
is larger than 256.  This release raises a ValueError exception;
future versions may return a mode "RGB" image instead (reported
by Richard Oudkerk).

+ Added WBMP read/write support (based on code by Duncan Booth).


+ Added DPI read/write support to the PNG codec.  The decoder sets
the info["dpi"] attribute for PNG files with appropriate resolution
settings.  The encoder uses the "dpi" option (based on code by Niki

+ Added limited support for "point" mappings from mode "I" to mode "L".
Only 16-bit values are supported (other values are clipped), the lookup
table must contain exactly 65536 entries, and the mode argument must be
set to "L".

+ Added support for Mac OS X icns files (based on code by Bob Ippolito).

+ Added "ModeFilter" support to the ImageFilter module.

+ Added support for Spider images (from William Baxter).  See the
comments in PIL/ for more information on this


+ Added new Sane release (from Ralph Heinkel).  See the Sane/README
and Sane/CHANGES files for more information.

+ Added experimental PngInfo chunk container to the PngImageFile
module.  This can be used to add arbitrary chunks to a PNG file.
Create a PngInfo instance, use "add" or "add_text" to add chunks,
and pass the instance as the "pnginfo" option when saving the

+ Added "getpalette" method.  This returns the palette as a list,
or None if the image has no palette.  To modify the palette, use
"getpalette" to fetch the current palette, modify the list, and
put it back using "putpalette".

+ Added optional flattening to the ImagePath "tolist" method.
tolist() or tolist(0) returns a list of 2-tuples, as before.
tolist(1) returns a flattened list instead.


+ Fixed BILINEAR/BICUBIC/ANTIALIAS filtering for mode "LA".

+ Added "getcolors()" method.  This is similar to the existing histo-
gram method, but looks at color values instead of individual layers,
and returns an unsorted list of (count, color) tuples.

By default, the method returns None if finds more than 256 colors.
If you need to look for more colors, you can pass in a limit (this
is used to allocate internal tables, so you probably don't want to
pass in too large values).

+ Build improvements: Fixed building under AIX, improved detection of
FreeType2 and Mac OS X framework libraries, and more.  Many thanks
to everyone who helped test the new "" script!


+ The "save" method now looks for a file format driver before
creating the file.

+ Don't use antialiased truetype fonts when drawing in mode "P", "I",
and "F" images.

+ Rewrote the "" file.  The new version scans for available
support libraries, and configures both the libImaging core library
and the bindings in one step.

To use specific versions of the libraries, edit the ROOT variables
in the file.

+ Removed threaded "show" viewer; use the old "show" implementation
instead (Windows).

+ Added deprecation warnings to Image.offset, ImageDraw.setink, and

+ Added width option to ImageDraw.line().  The current implementation
works best for straight lines; it does not support line joins, so
polylines won't look good.

+ ImageDraw.Draw is now a factory function instead of a class.  If
you need to create custom draw classes, inherit from the ImageDraw
class.    All other code should use the factory function.

+ Fixed loading of certain PCX files (problem reported by Greg
Hamilton, who also provided samples).

+ Changed _imagingft.c to require FreeType 2.1 or newer.  The
module can still be built with earlier versions; see comments
in _imagingft.c for details.


+ Added 'getim' method, which returns a PyCObject wrapping an
Imaging pointer.  The description string is set to IMAGING_MAGIC.
See Imaging.h for pointer and string declarations.

+ Fixed reading of TIFF JPEG images (problem reported by Ulrik

+ Made ImageColor work under Python 1.5.2

+ Fixed division by zero "equalize" on very small images (from
Douglas Bagnall).


+ The "paste" method now supports the alternative "paste(im, mask)"
syntax (in this case, the box defaults to im's bounding box).

+ The "ImageFile.Parser" class now works also for PNG files with
more than one IDAT block.

+ Added DPI read/write to the TIFF codec, and fixed writing of
rational values.  The decoder sets the info["dpi"] attribute
for TIFF files with appropriate resolution settings.  The
encoder uses the "dpi" option.

+ Disable interlacing for small (or narrow) GIF images, to
work around what appears to be a hard-to-find bug in PIL's
GIF encoder.

+ Fixed writing of mode "P" PDF images.  Made mode "1" PDF
images smaller.

+ Made the XBM reader a bit more robust; the file may now start
with a few whitespace characters.

+ Added support for enhanced metafiles to the WMF driver.  The
separate PILWMF kit lets you render both placeable WMF files
and EMF files as raster images.  See


+ Replaced broken WMF driver with a WMF stub plugin (see below).

+ Fixed writing of mode "1", "L", and "CMYK" PDF images (based on
input from Nicholas Riley and others).

+ Fixed adaptive palette conversion for zero-width or zero-height
images (from Chris Cogdon)

+ Fixed reading of PNG images from QuickTime 6 (from Paul Pharr)

+ Added support for StubImageFile plugins, including stub plugins
for BUFR, FITS, GRIB, and HDF5 files.  A stub plugin can identify
a given file format, but relies on application code to open and
save files in that format.

+ Added optional "encoding" argument to the ImageFont.truetype
factory.  This argument can be used to specify non-Unicode character
maps for fonts that support that.  For example, to draw text using
the Microsoft Symbol font, use:

font = ImageFont.truetype("symbol.ttf", 16, encoding="symb")
draw.text((0, 0), unichr(0xF000 + 0xAA))

(note that the symbol font uses characters in the 0xF000-0xF0FF

Common encodings are "unic" (Unicode), "symb" (Microsoft Symbol),
"ADOB" (Adobe Standard), "ADBE" (Adobe Expert), and "armn" (Apple
Roman).  See the FreeType documentation for more information.

+ Made "putalpha" a bit more robust; you can now attach an alpha
layer to a plain "L" or "RGB" image, and you can also specify
constant alphas instead of alpha layers (using integers or colour

+ Added experimental "LA" mode support.

An "LA" image is an "L" image with an attached transparency layer.
Note that support for "LA" is not complete; some operations may
fail or produce unexpected results.

+ Added "RankFilter", "MinFilter", "MedianFilter", and "MaxFilter"
classes to the ImageFilter module.

+ Improved support for applications using multiple threads; PIL
now releases the global interpreter lock for many CPU-intensive
operations (based on work by Kevin Cazabon).

+ Ignore Unicode characters in the PCF loader (from Andres Polit)

+ Fixed typo in OleFileIO.loadfat, which could affect loading of
FlashPix and Image Composer images (Daniel Haertle)

+ Fixed building on platforms that have Freetype but don't have
Tcl/Tk (Jack Jansen, Luciano Nocera, Piet van Oostrum and others)

+ Added EXIF GPSInfo read support for JPEG files.  To extract
GPSInfo information, open the file, extract the exif dictionary,
and check for the key 0x8825 (GPSInfo).  If present, it contains
a dictionary mapping GPS keys to GPS values.  For a list of keys,
see the EXIF specification.

The "ExifTags" module contains a GPSTAGS dictionary mapping GPS
tags to tag names.

+ Added DPI read support to the PCX and DCX codecs (info["dpi"]).

+ The "show" methods now uses a built-in image viewer on Windows.
This viewer creates an instance of the ImageWindow class (see
below) and keeps it running in a separate thread.  NOTE: This
was disabled in 1.1.5a4.

+ Added experimental "Window" and "ImageWindow" classes to the
ImageWin module.  These classes allow you to create a WCK-style
toplevel window, and use it to display raster data.

+ Fixed some Python 1.5.2 issues (to build under 1.5.2, use the approach)

+ Added support for the TIFF FillOrder tag.  PIL can read mode "1",
"L", "P" and "RGB" images with non-standard FillOrder (based on
input from Jeff Breidenbach).


+ Fixed ImageTk build problem on Unix.


+ Improved building on Mac OS X (from Jack Jansen).

+ Improved building on Windows with MinGW (from Klamer Shutte).

+ If no font is specified, ImageDraw now uses the embedded default
font.  Use the "load" or "truetype" methods to load a real font.

+ Added embedded default font to the ImageFont module (currently
an 8-pixel Courier font, taken from the X window distribution).


+ Added experimental EXIF support for JPEG files.  To extract EXIF
information from a JPEG file, open the file as usual, and call the
"_getexif" method.  If successful, this method returns a dictionary
mapping EXIF TIFF tags to values.  If the file does not contain EXIF
data, the "_getexif" method returns None.

The "ExifTags" module contains a dictionary mapping tags to tag

This interface will most likely change in future versions.

+ Fixed a bug when using the "transparency" option with the GIF

+ Added limited support for "bitfield compression" in BMP files
and DIB buffers, for 15-bit, 16-bit, and 32-bit images.  This
also fixes a problem with ImageGrab module when copying screen-
dumps from the clipboard on 15/16/32-bit displays.

+ Added experimental WAL (Quake 2 textures) loader.  To use this
loader, import WalImageFile and call the "open" method in that


+ Added updated SANE driver (Andrew Kuchling, Abel Deuring)

+ Use Python's "mmap" module on non-Windows platforms to read some
uncompressed formats using memory mapping.  Also added a "frombuffer"
function that allows you to access the contents of an existing string
or buffer object as if it were an image object.

+ Fixed a memory leak that could appear when processing mode "P"
images (from Pier Paolo Glave)

+ Ignore Unicode characters in the BDF loader (from Graham Dumpleton)


+ Added experimental RGBA-on-RGB drawing support.  To use RGBA
colours on an RGB image, pass "RGBA" as the second string to
the ImageDraw.Draw constructor.

+ Added support for non-ASCII strings (Latin-1) and Unicode
to the truetype font renderer.

+ The ImageWin "Dib" object can now be constructed directly from
an image object.

+ The ImageWin module now allows you use window handles as well
as device contexts.  To use a window handle, wrap the handle in
an ImageWin.HWND object, and pass in this object instead of the
device context.


+ Improved support for 16-bit unsigned integer images (mode "I;16").
This includes TIFF reader support, and support for "getextrema"
and "point" (from Klamer Shutte).

+ Made the BdfFontFile reader a bit more robust (from Kevin Cazabon
and Dmitry Vasiliev)

+ Changed TIFF writer to always write Compression tag, even when
using the default compression (from Greg Couch).

+ Added "show" support for Mac OS X (from Dan Wolfe).

+ Added clipboard support to the "ImageGrab" module (Windows only).
The "grabclipboard" function returns an Image object, a list of
filenames (not in 1.1.4), or None if neither was found.


+ Improved support for drawing RGB data in palette images.  You can
now use RGB tuples or colour names (see below) when drawing in a
mode "P" image.  The drawing layer automatically assigns color
indexes, as long as you don't use more than 256 unique colours.

+ Moved self test from MiniTest/ to ./

+ Added support for CSS3-style color strings to most places that
accept colour codes/tuples.  This includes the "ImageDraw" module,
the Image "new" function, and the Image "paste" method.

Colour strings can use one of the following formats: "f00",
"ff0000", "rgb(255,0,0)", "rgb(100%,0%,0%)", "hsl(0, 100%, 50%)",
or "red" (most X11-style colour names are supported).  See the
documentation for the "ImageColor" module for more information.

+ Fixed DCX decoder (based on input from Larry Bates)

+ Added "IptcImagePlugin.getiptcinfo" helper to extract IPTC/NAA
newsphoto properties from JPEG, TIFF, or IPTC files.

+ Support for TrueType/OpenType fonts has been added to
the standard distribution.  You need the freetype 2.0

+ Made the PCX reader a bit more robust when reading 2-bit
and 4-bit PCX images with odd image sizes.

+ Added "Kernel" class to the ImageFilter module.  This class
allows you to filter images with user-defined 3x3 and 5x5
convolution kernels.

+ Added "putdata" support for mode "I", "F" and "RGB".

+ The GIF writer now supports the transparency option (from
Denis Benoit).

+ A HTML version of the module documentation is now shipped
with the source code distribution.  You'll find the files in
the Doc subdirectory.

+ Added support for Palm pixmaps (from Bill Janssen).  This
change was listed for 1.1.3, but the "PalmImagePlugin" driver
didn't make it into the distribution.

+ Improved decoder error messages.


+ Made look for old versions of zlib.  For some back-
ground, see:


+ Added file (tested on Unix and Windows).  You still
need to build libImaging/imaging.lib in the traditional way,
but the script takes care of the rest.

The old build method is still

+ Fixed segmentation violation in ANTIALIAS filter (an internal
buffer wasn't properly allocated).


+ Added ANTIALIAS downsampling filter for high-quality "resize"
and "thumbnail" operations.  Also added filter option to the
"thumbnail" operation; the default value is NEAREST, but this
will most likely change in future versions.

+ Fixed plugin loader to be more robust if the __file__
variable isn't set.

+ Added seek/tell support (for layers) to the PhotoShop
loader.  Layer 0 is the main image.

+ Added new (but experimental) "ImageOps" module, which provides
shortcuts for commonly used operations on entire images.

+ Don't mess up when loading PNG images if the decoder leaves
data in the output buffer.  This could cause internal errors
on some PNG images, with some versions of ZLIB. (Bug report
and patch provided by Bernhard Herzog.)

+ Don't mess up on Unicode filenames.

+ Don't mess up when drawing on big endian platforms.

+ Made the TIFF loader a bit more robust; it can now read some
more slightly broken TIFF files (based on input from Ted Wright,
Bob Klimek, and D. Alan Stewart)

+ Added OS/2 EMX build files (from Andrew MacIntyre)

+ Change "ImageFont" to reject image files if they don't have the
right mode.  Older versions could leak memory for "P" images.
(Bug reported by Markus Gritsch).

+ Renamed some internal functions to avoid potential build
problem on Mac OS X.

+ Added DL_EXPORT where relevant (for Cygwin, based on input
from Robert Yodlowski)

+ (re)moved bogus __init__ call in BdfFontFile (bug spotted
by Fred Clare)

+ Added "ImageGrab" support (Windows only)

+ Added support for XBM hotspots (based on code contributed by
Bernhard Herzog).

+ Added write support for more TIFF tags, namely the Artist,
Copyright, DateTime, ResolutionUnit, Software, XResolution and
YResolution tags (from Greg Couch)

+ Added TransposedFont wrapper to ImageFont module

+ Added "optimize" flag to GIF encoder.  If optimize is present
and non-zero, PIL will work harder to create a small file.

+ Raise "EOFError" (not IndexError) when reading beyond the
end of a TIFF sequence.

+ Support rewind ("seek(0)") for GIF and TIFF sequences.

+ Load grayscale GIF images as mode "L"

+ Added DPI read/write support to the JPEG codec.  The decoder
sets the info["dpi"] attribute for JPEG files with JFIF dpi
settings.  The encoder uses the "dpi" option:

im ="file.jpg")
dpi =["dpi"]  raises KeyError if DPI not known"out.jpg", dpi=dpi)

Note that PIL doesn't always preserve the "info" attribute
for normal image operations.


+ Adapted to Python 2.1.  Among other things, all uses of the
"regex" module have been replaced with "re".

+ Fixed attribute error when reading large PNG files (this bug
was introduced in maintenance code released after the 1.1.1

+ Ignore non-string objects in sys.path

+ Fixed Image.transform(EXTENT) for negative xoffsets

+ Fixed loading of image plugins if PIL is installed as a package.
(The plugin loader now always looks in the directory where the module itself is found, even if that directory isn't on
the standard search path)

+ The Png plugin has been added to the list of preloaded standard

+ Fixed bitmap/text drawing in fill mode.

+ Fixed "getextrema" to work also for multiband images.

+ Added transparency support for L and P images to the PNG codec.

+ Improved support for read-only images.  The "load" method now
sets the "readonly" attribute for memory-mapped images.  Operations
that modifies an image in place (such as "paste" and drawing operations)
creates an in-memory copy of the image, if necessary.  (before this
change, any attempt to modify a memory-mapped image resulted in a
core dump...)

+ Added special cases for lists everywhere PIL expects a sequence.
This should speed up things like "putdata" and drawing operations.

+ The Image.offset method is deprecated.  Use the ImageChops.offset
function instead.

+ Changed ImageChops operators to copy palette and info dictionary
from the first image argument.


+ Additional fixes for Python 1.6/2.0, including TIFF "save" bug.

+ Changed "init" to properly load plugins when PIL is used as a

+ Fixed broken "show" method (on Unix)

1.1 insecure

+ Adapted to Python 1.6 ("append" and other method changes)

+ Fixed Image.paste when pasting with solid colour and matte
layers ("L" or "RGBA" masks) (bug reported by Robert Kern)

+ To make it easier to distribute prebuilt versions of PIL,
the tkinit binding stuff has been moved to a separate
extension module, named "_imagingtk".

*** Changes from release 0.3b2 to 1.0 final ***

+ If there's no 16-bit integer (like on a Cray T3E), set
INT16 to the smallest integer available.  Most of the
library works just fine anyway (from Bill Crutchfield)

+ Tweaks to make drawing work on big-endian platforms.

1.0 insecure


- Remove support for ``import Image``, etc. from the standard namespace. ``from PIL import Image`` etc. now required.
- Forked PIL based on `Hanno Schlichting's re-packaging <>`_




-*- coding: utf-8 -*-

The Python Imaging Library

ACKNOWLEDGEMENTS: PIL wouldn't be what it is without the help of:
David Ascher, Phil Austin, Douglas Bagnall, Larry Bates, Anthony
Baxter, William Baxter, Denis Benoit, Jan Blom, Duncan Booth, Alexey
Borzenkov, Jeff Breidenbach, Roger Burnham, Zac Burns, Gene Cash,
Kevin Cazabon, Fred Clare, Greg Coats, Chris Cogdon, Greg Couch, Bill
Crutchfield, Abel Deuring, Tim Docker, Fred Drake, Graham Dumpleton,
Matthew Ellis, Eric Etheridge, Daniel Fetchinson, Robin Friedrich,
Pier Paolo Glave, Federico Di Gregorio, Markus Gritsch, Daniel
Haertle, Greg Hamilton, Mark Hammond, Bernhard Herzog, Rob Hooft, Bob
Ippolito, Jack Jansen, Bill Janssen, Edward Jones, Richard Jones,
Håkan Karlsson, Robert Kern, David Kirtley, Bob Klimek, Matthias
Klose, Andrew Kuchling, Magnus Källström, Victor Lacina, Ben Last,
Hamish Lawson, Cesare Leonardi, Andrew MacIntyre, Jan Matejek, Naveen
Michaud-Agrawal, Gordon McMillan, Skip Montanaro, Fredrik Nehr,
Russell Nelson, Luciano Nocera, Travis Oliphant, Piet van Oostrum,
Richard Oudkerk, Paul Pharr, Andres Polit, Conrado Porto Lopes Gouvêa,
Eric Raymond, Victor Reijs, Bertil Reinhammar, Nicholas Riley, Don
Rozenberg, Toby Sargeant, Barry Scott, Les Schaffer, Joel Shprentz,
Klamer Shutte, Gene Skonicki, Niki Spahiev, D. Alan Stewart, Perry
Stoll, Paul Svensson, Ulrik Svensson, Miki Tebeka, Michael van
Tellingen, Ivan Tkatchev, Dan Torop, Adam Twardoch, Rune Uhlin, Dmitry
Vasiliev, Sasha Voynow, Charles Waldman, Collin Winter, Dan Wolfe,
Ka-Ping Yee, and many others (if your name should be on this list, let
me know.)


+ If PIL is built with the WITH_TKINTER flag, ImageTk can
automatically hook into a standard Tkinter build.  You
no longer need to build your own Tkinter to use the
ImageTk module.

The old way still works, though.  For more information,
see Tk/install.txt.

+ Some tweaks to ImageTk to support multiple Tk interpreters
(from Greg Couch).

+ ImageFont "load_path" now scans directory mentioned in .pth
files (from Richard Jones).


+ The TIFF plugin has been rewritten.  The new plugin fully
supports all major PIL image modes (including F and I).

+ The ImageFile module now includes a Parser class, which can
be used to incrementally decode an image file (while down-
loading it from the net, for example).  See the handbook for

+ "show" now converts non-standard modes to "L" or "RGB" (as
appropriate), rather than writing weird things to disk for
"xv" to choke upon. (bug reported by Les Schaffer).


+ Major speedups for rotate, transform(EXTENT), and transform(AFFINE)
when using nearest neighbour resampling.

+ Modified ImageDraw to be compatible with the Arrow graphics
interface.  See the handbook for details.

+ PIL now automatically loads file codecs when used as a package
(from The Dragon De Monsyne).  Also included an file
in the standard distribution.

+ The GIF encoder has been modified to produce much smaller files.

PIL now uses a run-length encoding method to encode GIF files.
On a random selection of GIF images grabbed from the web, this
version makes the images about twice as large as the original
LZW files, where the earlier version made them over 5 times
larger.  YMMV, of course.

+ Added PCX write support (works with "1", "P", "L", and "RGB")

+ Added "bitmap" and "textsize" methods to ImageDraw.

+ Improved font rendering code.  Fixed a bug or two, and moved
most of the time critical stuff to C.

+ Removed "".  Use "" instead!

+ Improved 16-bit support (still experimental, though).

The following methods now support "I;16" and "I;16B" images:
"getpixel", "copy", "convert" (to and from mode "I"), "resize",
"rotate", and "transform" with nearest neighbour filters, and
"save" using the IM format.  The "new" and "open" functions
also work as expected.  On Windows, 16-bit files are memory

NOTE: ALL other operations are still UNDEFINED on 16-bit images.

+ The "paste" method now supports constant sources.

Just pass a colour value (a number or a tuple, depending on
the target image mode) instead of the source image.

This was in fact implemented in an inefficient way in
earlier versions (the "paste" method generated a temporary
source image if you passed it a colour instead of an image).
In this version, this is handled on the C level instead.

+ Added experimental "RGBa" mode support.

An "RGBa" image is an RGBA image where the colour components
have have been premultiplied with the alpha value.  PIL allows
you to convert an RGBA image to an RGBa image, and to paste
RGBa images on top of RGB images.  Since this saves a bunch
of multiplications and shifts, it is typically about twice
as fast an ordinary RGBA paste.

+ Eliminated extra conversion step when pasting "RGBA" or "RGBa"
images on top of "RGB" images.

+ Fixed Image.BICUBIC resampling for "RGB" images.

+ Fixed PCX image file handler to properly read 8-bit PCX
files (bug introduced in 1.0b1, reported by Bernhard

+ Fixed PSDraw "image" method to restore the coordinate

+ Fixed "blend" problem when applied to images that was
not already loaded (reported by Edward C. Jones)

+ Fixed -f option to "" (from Anthony Baxter)


+ Added Toby J. Sargeant's quantization package.  To enable
quantization, use the "palette" option to "convert":

imOut = im.convert("P", palette=Image.ADAPTIVE)

This can be used with "L", "P", and "RGB" images.  In this
version, dithering cannot be used with adaptive palettes.

Note: ADAPTIVE currently maps to median cut quantization
with 256 colours.  The quantization package also contains
a maximum coverage quantizer, which will be supported by
future versions of PIL.

+ Added Eric S. Raymond's "pildriver" image calculator to the
distribution.  See the docstring for more information.

+ The "offset" method no longer dumps core if given positive
offsets (from Charles Waldman).

+ Fixed a resource leak that could cause ImageWin to run out of
GDI resources (from Roger Burnham).

+ Added "arc", "chord", and "pieslice" methods to ImageDraw (inspired
by code contributed by Richard Jones).

+ Added experimental 16-bit support, via modes "I;16" (little endian
data) and "I;16B" (big endian).  Only a few methods properly support
such images (see above).

+ Added XV thumbnail file handler (from Gene Cash).

+ Fixed BMP image file handler to handle palette images with small
palettes (from Rob Hooft).

+ Fixed Sun raster file handler for palette images (from Charles

+ Improved various internal error messages.

+ Fixed Path constructor to handle arbitrary sequence objects.  This
also affects the ImageDraw class (from Richard Jones).

+ Fixed a bug in JpegDecode that caused PIL to report "decoder error
-2" for some progressive JPEG files (reported by Magnus Källström,
who also provided samples).

+ Fixed a bug in JpegImagePlugin that caused PIL to hang when loading
JPEG files using 16-bit quantization tables.

+ The Image "transform" method now supports Image.QUAD transforms.
The data argument is an 8-tuple giving the upper left, lower
left, lower right, and upper right corner of the source quadri-
lateral.  Also added Image.MESH transform which takes a list
of quadrilaterals.

+ The Image "resize", "rotate", and "transform" methods now support
Image.BILINEAR (2x2) and Image.BICUBIC (4x4) resampling filters.
Filters can be used with all transform methods.

+ The ImageDraw "rectangle" method now includes both the right
and the bottom edges when drawing filled rectangles.

+ The TGA decoder now works properly for runlength encoded images
which have more than one byte per pixel.

+ "getbands" on an YCbCr image now returns ("Y", "Cb", "Cr")

+ Some file drivers didn't handle the optional "modify" argument
to the load method.  This resulted in exceptions when you used
"paste" (and other methods that modify an image in place) on a
newly opened file.

*** Changes from release 0.2 (b5) to 0.3 (b2) ***


The test suite includes 825 individual tests.

+ An Image "getbands" method has been added.  It returns a tuple
containing the individual band names for this image.  To figure
out how many bands an image has, use "len(im.getbands())".

+ An Image "putpixel" method has been added.

+ The Image "point" method can now be used to convert "L" images
to any other format, via a lookup table.  That table should
contain 256 values for each band in the output image.

+ Some file drivers (including FLI/FLC, GIF, and IM) accidentally
overwrote the offset method with an internal attribute.  All
drivers have been updated to use private attributes where

+ The Image "histogram" method now works for "I" and "F" images.
For these modes, PIL divides the range between the min and
max values used in the image into 256 bins.  You can also
pass in your own min and max values via the "extrema" option:

h = im.histogram(extrema=(0, 255))

+ An Image "getextrema" method has been added.  It returns the
min and max values used in the image. In this release, this
works for single band images only.

+ Changed the PNG driver to load and save mode "I" images as
16-bit images.  When saving, values outside the range 0..65535
are clipped.

+ Fixed to work with the new "pilfont" compiler.

+ Added JPEG "save" and "draft" support for mode "YCbCr" images.
Note that if you save an "YCbCr" image as a JPEG file and read
it back, it is read as an RGB file.  To get around this, you
can use the "draft" method:

im ="color.jpg")
im.draft("YCbCr", im.size)

+ Read "RGBA" TGA images.  Also fixed the orientation bug; all
images should now come out the right way.

+ Changed mode name (and internal representation) from "YCrCb"
to "YCbCr" (!)


The test suite includes 750 individual tests.

+ The "pilfont" package is now included in the standard PIL
distribution.  The pilfont utility can be used to convert
X BDF and PCF raster font files to a format understood by
the ImageFont module.

+ GIF files are now interlaced by default.  To write a
non-interlaced file, pass interlace=0 to the "save"

+ The default string format has changed for the "fromstring"
and "tostring" methods.

NOTE: If no extra arguments are given, the first line in
the string buffer is the top line of the image, instead of
the bottom line.  For RGB images, the string now contains
3 bytes per pixel instead of 4.  These changes were made
to make the methods compatible with the "fromstring"
factory function.

To get the old behaviour, use the following syntax:

data = im.tostring("raw", "RGBX", 0, -1)
im.fromstring(data, "raw", "RGBX", 0, -1)

+ "new" no longer gives a MemoryError if the width or height
is zero (this only happened on platforms where malloc(0)
or calloc(0) returns NULL).

+ "new" now adds a default palette object to "P" images.

+ You can now convert directly between all modes supported by
PIL.  When converting colour images to "P", PIL defaults to
a "web" palette and dithering.  When converting greyscale
images to "1", PIL uses a thresholding and dithering.

+ Added a "dither" option to "convert".  By default, "convert"
uses floyd-steinberg error diffusion for "P" and "1" targets,
so this option is only used to *disable* dithering. Allowed
values are NONE (no dithering) or FLOYDSTEINBERG (default).

imOut = im.convert("P", dither=Image.NONE)

+ Added a full set of "I" decoders.  You can use "fromstring"
(and file decoders) to read any standard integer type as an
"I" image.

+ Added some support for "YCbCr" images (creation, conversion
from/to "L" and "RGB", IM YCC load/save)

+ "getpixel" now works properly with fractional coordinates.

+ ImageDraw "setink" now works with "I", "F", "RGB", "RGBA",
"RGBX", "CMYK", and "YCbCr" images.

+ ImImagePlugin no longer attaches palettes to "RGB" images.

+ Various minor fixes.


+ Added experimental IPTC/NAA support.

+ Eliminated AttributeError exceptions after "crop" (from
Skip Montanaro)

+ Reads some uncompressed formats via memory mapping (this
is currently supported on Win32 only)

+ Fixed some last minute glitches in the last alpha release
(Types instead of types in, version numbers, etc.)

+ Eliminated some more bogus compiler warnings.

+ Various fixes to make PIL compile and run smoother on Macs
(from Jack Jansen).

+ Fixed "fromstring" and "tostring" for mode "I" images.


The test suite includes 530 individual tests.

+ Eliminated unexpected side-effect in "paste" with matte.  "paste"
now works properly also if compiled with "gcc".

+ Adapted to Python 1.5 (build issues only)

+ Fixed the ImageDraw "point" method to draw also the last
point (!).

+ Added "I" and "RGBX" support to

+ The plugin path is now properly prepended to the module search
path when a plugin module is imported.

+ Added "draw" method to the ImageWin.Dib class.  This is used by
Topaz to print images on Windows printers.

+ "convert" now supports conversions from "P" to "1" and "F".

+ "paste" can now take a colour instead of an image as the first argument.
The colour must match the colour argument given to the new function, and
match the mode of the target image.

+ Fixed "paste" to allow a mask also for mode "F" images.

+ The BMP driver now saves mode "1" images.  When loading images, the mode
is set to "L" for 8-bit files with greyscale palettes, and to "P" for
other 8-bit files.

+ The IM driver now reads and saves "1" images (file modes "0 1" or "L 1").

+ The JPEG and GIF drivers now saves "1" images.  For JPEG, the image
is saved as 8-bit greyscale (it will load as mode "L").  For GIF, the
image will be loaded as a "P" image.

+ Fixed a potential buffer overrun in the GIF encoder.


The test suite includes 400 individual tests.

+ Improvements to the test suite revealed a number of minor bugs, which
are all fixed.  Note that crop/paste, 32-bit ImageDraw, and ImageFont
are still weak spots in this release.

+ Added "putpalette" method to the Image class.  You can use this
to add or modify the palette for "P" and "L" images.  If a palette
is added to an "L" image, it is automatically converted to a "P"

+ Fixed ImageDraw to properly handle 32-bit image memories
("RGB", "RGBA", "CMYK", "F")

+ Fixed "fromstring" and "tostring" not to mess up the mode attribute
in default mode.

+ Changed ImPlatform.h to work on CRAY's (don't have one at home, so I
haven't tried it).  The previous version assumed that either "short"
or "int" were 16-bit wide. PIL still won't compile on platforms where
neither "short", "int" nor "long" are 32-bit wide.

+ Added file= and data= keyword arguments to PhotoImage and BitmapImage.
This allows you to use them as drop-in replacements for the corre-
sponding Tkinter classes.

+ Removed bogus references to the crack coder (ImagingCrack).


+ Make sure image is loaded in "tostring".

+ Added floating point packer (native 32-bit floats only).

*** Changes from release 0.1b1 to 0.2 (b5) ***

+ Modified "fromstring" and "tostring" methods to use file codecs.
Also added "fromstring" factory method to create an image directly
from data in a string.

+ Added support for 32-bit floating point images (mode "F").  You
can convert between "L" and "F" images, and apply a subset of the
available image processing methods on the "F" image.  You can also
read virtually any data format into a floating point image memory;
see the section on "Decoding Floating Point Data" in the handbook
for more information.


+ Fixed the tobitmap() method to work properly for small bitmaps.

+ Added RMS and standard deviation to the ImageStat.Stat class.  Also
modified the constructor to take an optional feature mask, and also
to accept either an image or a list containing the histogram data.

+ The BitmapImage code in ImageTk can now use a special bitmap
decoder, which has to be patched into Tk.  See the "Tk/pilbitmap.txt"
file for details.  If not installed, bitmaps are transferred to Tk as
XBM strings.

+ The PhotoImage code in ImageTk now uses a Tcl command ("PyImagingPaste")
instead of a special image type.  This gives somewhat better performance,
and also allows PIL to support transparency.

+ ImageTk now honours the alpha layer in RGBA images.  Only fully
transparent pixels are made transparent (that is, the alpha layer
is treated as a mask).  To treat the alpha laters as a matte, you
must paste the image on the background before handing it over to

+ Added McIdas reader (supports 8-bit images only).

+ PIL now preloads drivers for BMP, GIF, JPEG, PPM, and TIFF.  As
long as you only load and save these formats, you don't have to
wait for a full scan for drivers.  To force scanning, call the
Image.init() function.

+ The "seek" and "tell" methods are now always available, also for
single-frame images.

+ Added optional mask argument to histogram method.  The mask may
be an "1" or "L" image with the same size as the original image.
Only pixels where the mask is non-zero are included in the

+ The "paste" method now allows you to specify only the lower left
corner (a 2-tuple), instead of the full region (a 4-tuple).

+ Reverted to old plugin scanning model; now scans all directory
names in the path when looking for plugins.

+ Added PIXAR raster support.  Only uncompressed ("dumped") RGB
images can currently be read (based on information provided
by Greg Coats).

+ Added FlashPix (FPX) read support.  Reads all pixel formats, but
only the highest resolution is read, and the viewing transform is
currently ignored.

+ Made PNG encoding somewhat more efficient in "optimize" mode; a
bug in 0.2b4 didn't enable all predictor filters when optimized
storage were requested.

+ Added Microsoft Image Composer (MIC) read support.  When opened,
the first sprite in the file is loaded.  You can use the seek method
to load additional sprites from the file.

+ Properly reads "P" and "CMYK" PSD images.

+ "pilconvert" no longer optimizes by default; use the -o option to
make the file as small as possible (at the expense of speed); use
the -q option to set the quality when compressing to JPEG.

+ Fixed "crop" not to drop the palette for "P" images.

+ Added and verified FLC support.

+ Paste with "L" or "RGBA" alpha is now several times faster on most

+ Changed to initialize the image to black, as described
in the handbook.  To get an uninitialized image, use None as the

+ Fixed the PDF encoder to produce a valid header; Acrobat no longer
complains when you load PDF images created by PIL.

+ PIL only scans fully-qualified directory names in the path when
looking for plugins.

+ Faster implementation of "save" used when filename is given,
or when file object has "fileno" and "flush" methods.

+ Don't crash in "crop" if region extends outside the source image.

+ Eliminated a massive memory leak in the "save" function.

+ The GIF decoder doesn't crash if the code size is set to an illegal
value.  This could happen since another bug didn't handle local
palettes properly if they didn't have the same size as the
global palette (not very common).

+ Added predictor support (TIFF 6.0 section 14) to the TIFF decoder.

+ Fixed palette and padding problems in BMP driver.  Now properly
writes "1", "L", "P" and "RGB" images.

+ Fixed getpixel()/getdata() to return correct pixel values.

+ Added PSD (PhotoShop) read support.  Reads both uncompressed
and compressed images of most types.

+ Added GIF write support (writes "uncompressed" GIF files only,
due to unresolvable licensing issues).  The "" script
can be used to create GIF animations.

+ Reads 8-bit "L" and "P" TGA images.  Also reads 16-bit "RGB"

+ Added FLI read support.  This driver has only been tested
on a few FLI samples.

+ Reads 2-bit and 4-bit PCX images.

+ Added MSP read and write support.  Both version 1 and 2 can be
read, but only version 1 (uncompressed) files are written.

+ Fixed a bug in the FLI/FLC identification code that caused the
driver to raise an exception when parsing valid FLI/FLC files.

+ Improved performance when loading file format plugins, and when
opening files.

+ Added GIF animation support, via the "seek" and "tell" methods.
You can use "" to play an animated GIF file.

+ Removed MNG support, since the spec is changing faster than I
can change the code.  I've added support for the experimental
ARG format instead.  Contact me for more information on this

+ Added keyword options to the "save" method.  The following options
are currently supported:

format	option		description
JPEG	optimize	minimize output file at the
expense of compression speed.

JPEG	progressive	enable progressive output. the
option value is ignored.

JPEG	quality		set compression quality (1-100).
the default value is 75.

JPEG	smooth		smooth dithered images.  value
is strength (1-100).  default is
off (0).

PNG	optimize	minimize output file at the
expense of compression speed.

Expect more options in future releases.  Also note that
file writers silently ignore unknown options.

+ Plugged memory leaks in the PNG and TIFF decoders.

+ Added PNG write support.

+ (internal) RGB unpackers and converters now set the pad byte
to 255 (full opacity).

+ Properly handles the "transparency" property for GIF, PNG
and XPM files.

+ Added a "putalpha" method, allowing you to attach a "1" or "L"
image as the alpha layer to an "RGBA" image.

+ Various improvements to the sample scripts:

"pilconvert"  Carries out some extra tricks in order to make
the resulting file as small as possible.

"explode"	(NEW) Split an image sequence into individual frames.

"gifmaker"	(NEW) Convert a sequence file into a GIF animation.
Note that the GIF encoder create "uncompressed" GIF
files, so animations created by this script are
rather large (typically 2-5 times the compressed

"image2py"	(NEW) Convert a single image to a python module.  See
comments in this script for details.

"player"	If multiple images are given on the command line,
they are interpreted as frames in a sequence.  The
script assumes that they all have the same size.
Also note that this script now can play FLI/FLC
and GIF animations.

This player can also execute embedded Python
animation applets (ARG format only).

"viewer"	Transparent images ("P" with transparency property,
and "RGBA") are superimposed on the standard Tk back-

+ Fixed colour argument to "new".  For multilayer images, pass a
tuple: (Red, Green, Blue), (Red, Green, Blue, Alpha), or (Cyan,
Magenta, Yellow, Black).

+ Added XPM (X pixmap) read support.


+ Added MNG (multi-image network graphics) read support.  "Ming"
is a proposed animation standard, based on the PNG file format.

You can use the "player" sample script to display some flavours
of this format.  The MNG standard is still under development,
as is this driver.  More information, including sample files,
can be found at <>

+ Added a "verify" method to images loaded from file.  This method
scans the file for errors, without actually decoding the image
data, and raises a suitable exception if it finds any problems.
Currently implemented for PNG and MNG files only.

+ Added support for interlaced GIF images.

+ Added PNG read support -- if linked with the ZLIB compression library,
PIL reads all kinds of PNG images, except interlaced files.

+ Improved PNG identification support -- doesn't mess up on unknown
chunks, identifies all possible PNG modes, and verifies checksum
on PNG header chunks.

+ Added an experimental reader for placable Windows Meta Files (WMF).
This reader is still very incomplete, but it illustrates how PIL's
drawing capabilities can be used to render vector and metafile

+ Added restricted drivers for images from Image Tools (greyscale
only) and LabEye/IFUNC (common interchange modes only).

+ Some minor improvements to the sample scripts provided in the
"Scripts" directory.

+ The test images have been moved to the "Images" directory.




+ Fixed filling of complex polygons.  The ImageDraw "line" and
"polygon" methods also accept Path objects.

+ The ImageTk "PhotoImage" object can now be constructed directly
from an image.  You can also pass the object itself to Tkinter,
instead of using the "image" attribute.  Finally, using "paste"
on a displayed image automatically updates the display.

+ The ImageTk "BitmapImage" object allows you to create transparent
overlays from 1-bit images.  You can pass the object itself to
Tkinter.  The constructor takes the same arguments as the Tkinter
BitmapImage class; use the "foreground" option to set the colour
of the overlay.

+ Added a "putdata" method to the Image class.  This can be used to
load a 1-layer image with data from a sequence object or a string.
An optional floating point scale and offset can be used to adjust
the data to fit into the 8-bit pixel range.  Also see the "getdata"

+ Added the EXTENT method to the Image "transform" method.  This can
be used to quickly crop, stretch, shrink, or mirror a subregion
from another image.

+ Adapted to Python 1.4.

+ Added a project makefile for Visual C++ 4.x.  This allows you to
easily build a dynamically linked version of PIL for Windows 95
and NT.

+ A Tk "booster" patch for Windows is available.  It gives dramatic
performance improvements for some displays.  Has been tested with
Tk 4.2 only, but is likely to work with Tk 4.1 as well.  See the Tk
subdirectory for details.

+ You can now save 1-bit images in the XBM format.  In addition, the
Image class now provides a "tobitmap" method which returns a string
containing an XBM representation of the image.  Quite handy to use
with Tk.

+ More conversions, including "RGB" to "1" and more.


+ Where earlier versions accepted lists, this version accepts arbitrary
Python sequences (including strings, in some cases).  A few resource
leaks were plugged in the process.

+ The Image "paste" method now allows the box to extend outside
the target image.  The size of the box, the image to be pasted,
and the optional mask must still match.

+ The ImageDraw module now supports filled polygons, outlined and
filled ellipses, and text.  Font support is rudimentary, though.

+ The Image "point" method now takes an optional mode argument,
allowing you to convert the image while translating it.  Currently,
this can only be used to convert "L" or "P" images to "1" images
(creating thresholded images or "matte" masks).

+ An Image "getpixel" method has been added.  For single band images,
it returns the pixel value at a given position as an integer.
For n-band images, it returns an n-tuple of integers.

+ An Image "getdata" method has been added.  It returns a sequence
object representing the image as a 1-dimensional array.  Only len()
and [] can be used with this sequence.  This method returns a
reference to the existing image data, so changes in the image
will be immediately reflected in the sequence object.

+ Fixed alignment problems in the Windows BMP writer.

+ If converting an "RGB" image to "RGB" or "L", you can give a second
argument containing a colour conversion matrix.

+ An Image "getbbox" method has been added.  It returns the bounding
box of data in an image, considering the value 0 as background.

+ An Image "offset" method has been added.  It returns a new image
where the contents of the image have been offset the given distance
in X and/or Y direction.  Data wraps between edges.

+ Saves PDF images.  The driver creates a binary PDF 1.1 files, using
JPEG compression for "L", "RGB", and "CMYK" images, and hex encoding
(same as for PostScript) for other formats.

+ The "paste" method now accepts "1" masks.  Zero means transparent,
any other pixel value means opaque.  This is faster than using an
"L" transparency mask.

+ Properly writes EPS files (and properly prints images to postscript
printers as well).

+ Reads 4-bit BMP files, as well as 4 and 8-bit Windows ICO and CUR
files.  Cursor animations are not supported.

+ Fixed alignment problems in the Sun raster loader.

+ Added "draft" and "thumbnail" methods.  The draft method is used
to optimize loading of JPEG and PCD files, the thumbnail method is
used to create a thumbnail representation of an image.

+ Added Windows display support, via the ImageWin class (see the
handbook for details).

+ Added raster conversion for EPS files.  This requires GNU or Aladdin
Ghostscript, and probably works on UNIX only.

+ Reads PhotoCD (PCD) images.  The base resolution (768x512) can be
read from a PhotoCD file.

+ Eliminated some compiler warnings.  Bindings now compile cleanly in C++
mode.  Note that the Imaging library itself must be compiled in C mode.

+ Added "", which converts BDF fonts into images with associated
metrics.  This is definitely work in progress.  For info, see description
in script for details.

+ Fixed a bug in the "" module.

+ Fixed a bug in the netpbm save hack in ""

+ Fixed 90 and 270 degree rotation of rectangular images.

+ Properly reads 8-bit TIFF palette-color images.

+ Reads plane separated RGB and CMYK TIFF images.

+ Added driver debug mode.  This is enabled by setting Image.DEBUG
to a non-zero value.  Try the -D option to "" and see what

+ Don't crash on "atend" constructs in PostScript files.

+ Only the Image module imports _imaging directly.  Other modules
should refer to the binding module as "Image.core".

*** Changes from release 0.0 to 0.1 (b1) ***

+ A handbook is available (distributed separately).

+ The coordinate system is changed so that (0,0) is now located
in the upper left corner.  This is in compliancy with ISO 12087
and 90% of all other image processing and graphics libraries.

+ Modes "1" (bilevel) and "P" (palette) have been introduced.  Note
that bilevel images are stored with one byte per pixel.

+ The Image "crop" and "paste" methods now accepts None as the
box argument, to refer to the full image (self, that is).

+ The Image "crop" method now works properly.

+ The Image "point" method is now available.  You can use either a
lookup table or a function taking one argument.

+ The Image join function has been renamed to "merge".

+ An Image "composite" function has been added.  It is identical
to copy() followed by paste(mask).

+ An Image "eval" function has been added.  It is currently identical
to point(function); that is, only a single image can be processed.

+ A set of channel operations has been added.  See the "ImageChops"
module,, and the handbook for details.

+ Added the "pilconvert" utility, which converts image files.  Note
that the number of output formats are still quite restricted.

+ Added the "pilfile" utility, which quickly identifies image files
(without loading them, in most cases).

+ Added the "pilprint" utility, which prints image files to Postscript

+ Added a rudimentary version of the "pilview" utility, which is
simple image viewer based on Tk.  Only File/Exit and Image/Next
works properly.

+ An interface to Tk has been added.  See "Lib/" and README
for details.

+ An interface to Jack Jansen's Img library has been added (thanks to
Jack).  This allows you to read images through the Img extensions file
format handlers.  See the file "Lib/" for details.

+ Postscript printing is provided through the PSDraw module.  See the
handbook for details.