Tag the 1.9.0 release
This release features a number of improvements for the X-ray tomography community with support for white and dark fields, initial support for using TomoPy in a Docker pipeline, and some additional color presets that were requested after we reworked our presets in the last release. We have made improvements under the hood too, with a more Pythonic interface for the operators to use (the old API was retained to support existing operators), and new API to support X-ray data.
A new crop operator was added which currently crops volume rendered data, it will be extended to support more visualization types in the future. The slice operator was improved to support “thick” slices, i.e. aggregation over multiple slices. We have improved the integration of the documentation within the application, offering more links from dialogs to the relevant help page/section. The documentation has been updated and improved to cover a wider array of use cases. We dropped support for Python 2 many years ago, and continue to support Python 3.7 at this point, with plans to move to Python 3.8 soon. The ParaView and VTK libraries were updated for this release, bringing in a number of fixes and improvements.
This release features a number of improvements including a new color map preset dialog with an improved small form factor, the ability to save custom color maps, and preview all available color maps. The HDF5 support was improved, with a new generic HDF5 file reader, and a DataExchange format reader. The HDF5 formats also support striding before loading for large volumes. A number of major dependencies were updated including Python, NumPy, SciPy, ParaView, VTK and ITK. The data properties panel was improved to show the number of voxels, memory usage, and other important parameters.
This release was made in preparation for the Kitware Summer Courses held in July, and the Microscopy and Microanalysis (M&M) Conference held in Portland, OR. This announcement was delayed a little – it was a busy summer! Chris Harris and Marcus Hanwell ran the short half-day course, with slides available here, and I presented the latest innovations in the Tomviz project at M&M. We are also very pleased to announce a new documentation site, which features content expanded from previous tutorials.
This release features a number of improvements to application startup time, and a splash screen to show that the application is starting up on the occasions where it doesn't start as quickly as we would like. The reconstruction operators were also improved to use less memory, default to 32 bit floating point output, and to execute more efficiently. New operators support improved data wrangling from picking a single component when input data has multiple channels (e.g. RGB) to casting/rescaling data to reduce memory requirements (8 and 16 bit unsigned integers for example).
We have made a number of improvements to the "live" pipeline, where updated reconstructions can be viewed as an iterative reconstruction proceeds, or data can be viewed as projections are acquired. The EMD format saw a number of improvements for saving/loading metadata including lengths/angles. An issue with array ordering was discovered, meaning old EMD files will look corrupt, but all future files will be saved with the appropriate array ordering. We added a transpose data operator that will reorder these old files to enable recovery of data (or you can use an older version of Tomviz to read the file). NumPy files were also improved, and round tripping to/from Python visualizations were tested with both EMD and NumPy. Files, directories and state files can all be opened from the command line now too.
The release also features many other bug fixes and improvements.
This is a bug fix release that fixes a few issues that could cause Tomviz to crash, such as when manually aligning images in a tilt series, or when loading a state file if the data file is not located in the specified location.
Fixed a critical crash affecting macOS Mojave on application start up. Fixed an issue where image stacks would not load correctly from the recent file menu, made a number of speed improvements when loading/previewing image stacks. Reduced application start up times, and added support for loading atomic structure data into the application. Added the ability to lock the aspect ratio when exporting a screenshot.
This release features a number of improvements to the new pipeline that debuted with our 1.3.0 release, including support for live updates of volumes from operators, atomic coordinates emitted from operators, improved support for loading image stacks, and a number of improvements to our data acquisition. Many improvements have been made to the user interface, with some serious speed improvements for histogram calculation and contour determination.
This is a bug fix release that fixes an issue where state files wouldn't restore the file name of the data source, a crash when showing the color legend with no loaded data, ensure visualizations are correctly removed from the pipeline when deleted, and removing a warning when loading state files.
This release features some big changes to the Tomviz pipeline, state file, and related functionality. The pipeline now explicitly shows the input and output data, with the visualization modules attached to the data that is being visualized. These changes have been carried through to the state file, which is now represented using JSON. The pipeline supports live updates (experimentally) for both reconstructions and data acquisition.
The user interface saw a number of updates to improve the default layouts, and make the best use of space. A number of enhancements have focused on improving load times of the application at startup, and when executing the Python environment. Tilt angles can be loaded from a file now, and custom Python operators can be easily added from a menu item. Support for images with multiple components has been added, and separate images can be merged.
Usability has been improved, with manual alignment operations supporting reopening of the dialog to further edit, and a view code button added to Python operators. Image stack loading has seen improvements, basic consistency of image sizes is verified and errors displayed. There are a huge number of bug fixes and improvements that were developed by our team.
This release candidate features a number of major updates to our data processing pipeline, state file, and associated interfaces. Old state files are not supported, and the pipelines in this new version feature a number of differences in how they are presented in the interface. We would appreciate your testing and feedback as we stabilize new features, and work towards the final release.