Pims

This release is the same as v0.4, it only has an improved readme.

This is a bugfix release recommended for all users.

- Fixed imageio \"infinite reader\" bug (PR 415)
- Added installation instructions for imageio-ffmpeg and moviepy (PR 417)

This is a major release recommended for all users. Your existing code to
open files may need to be edited slightly -- see the first change
listed below.

- API: all readers do not support the keyword arguments
`process_func`, `dtype` and `as_grey` anymore. Please consult the
documentation on Pipelines on how to convert videos. (PR 250)
- New built-in `as_grey` pipeline for convenient conversion to
greyscale (PR 305)
- New built-in `crop` pipeline for making a cropped sequence (PR 247)
- Major speedup for random access in PyAVReaderIndexed (PR 340)
- Bumped Bioformats version to 6.5.x (PR 301)
- Added instructions for installing/updating Bioformats (PR 346)
- Enhanced support for metadata in the CINE format (PR 317)
- Enhanced documentation for PyAV video support (PR 334)
- Added multidimensional capabilities to ImageIOReader (see
"multidimensional" documentation) (PR 320)
- Added support for nd2reader as alternative reader for Nikon nd2
files (PR 272)
- Removed fallback to scipy.misc.imread, which was removed from scipy
(PR 359)
- Fixed compatibility with PyAV 0.4.0 and newer (PR 300)
- Fixed compatibility and bugs in PyAV export (PR 283, PR 313)
- Fixed opening of PyAV videos with audio (PR 322)
- Fixed compatibility with newer versions of tifffile (PR 314, PR 339)
- Fixed response to missing ImageIO (PR 333)

This is a hotfix release to adjust to a change in our dependency, six, that made the library un-importable in Python 2. There are no other changes.

API: N-dimensional readers now smartly use get_frame methods; depending on the bundle_axes settings, the reader optimizes which get_frame methods is optimal to use. Readers that derive from FramesSequenceND will need to call FramesSequenceND.__init__() on initialization, and also will need to register get_frame methods using _register_get_frame(method, axes). (PR 227)
API: Swap elements of frame_shape in SpeStack to match frames’ shape. (PR 241)
API: Swap elements of frame_shape in PyAVVideoReader to match frames’ shape. (PR 251)
API: Reimplement the PyAVVideoReader (or: Video) into a reader that uses the frame timestamps and reader frame rate to compute the frame index. The new reader is named PyAVReaderTimed and the names PyAVVideoReader and Video refer to it. For the case when the video misses timestamps, the old implementation is available under PyAVReaderIndexed.
API: The video exporter (export) takes more arguments. The argument 'format' has been renamed to 'pixel_format', while 'format' now refers to the container format. (PR 257)
Fixed filename sorting when list is provided explicitely to ImageSequence (PR 252)
Fixed plot_to_frame with savefig.bbox == 'tight' (PR 248)
Added a reader that wraps ImageIO ImageIOReader (PR 233)
Added a reader that wraps MoviePy MoviePyReader (PR 233)
Added a reader for single images ImageReader and ImageReaderND (PR 249)
Added a reader that stacks readers into a multidimensional reader ReaderSequence (PR 249)
Added a video exporter based on MoviePy (PR 233)
Added BioformatsReader.metadata.fields that lists all metadata fields. (PR 230)