areaDetector R3-6 released

[Mark Rivers via Tech-talk <tech-talk@aps.anl.gov>]

New releases of the following areaDetector modules are now available.

areaDetector R3-6
ADCore R3-6

ADSupport R1-8

These are from the R3-6 release notes for areaDetector:

 

* New submodule repositories

  * specsAnalyser

    * Driver for SPECS Phoibos electron spectrometers, with DLD (Delay Line Detector)

    * From Pete Leicester, Stuart Wilkins and others.

  * ADCompVision

    * Plugin to perform imaging processing using opencv library

    * From Jakub Wlodek and Kaz Gofron.

  * ADPluginBar

    * Plugin to process bar codes

    * From Jakub Wlodek and Kaz Gofron.

  * ADPcoWin

    * Driver for PCO cameras, using their SDK which is Windows only.

    * From Jonathan Thompson and others.

  * ADMMPAD

    * Driver for the MMPAD detectors from Sol Gruner's group at Cornell.

    * From Mark Rivers.

  * ADGenICam

    * New base class for GenICam cameras.  Used by ADAravis, ADSpinnaker, and ADVimba.

    * Not yet ready for production use.

    * From Mark Rivers.

  * ADAravis

    * Designed to replace aravisGigE.  It is simpler because it uses ADGeniICam.

    * Not yet ready for production use.

    * From Mark Rivers, based on aravisGigE which is by Tom Cobb and others.

  * ADVimba

    * Driver for AVT/Prosilica cameras using their new Vimba SDK.  Uses ADGenICam.

    * Not yet ready for production use.

    * From Mark Rivers.

  * ADSpinnaker

    * Driver for FLIR/Point Grey cameras using their new Spinnaker SDK.

    * This is not a new repository, but it now contains a new GenICam branch that uses ADGenICam.

    * GenICam branch is not yet ready for production use

    * From Mark Rivers.

 

 

These are from the R3-6 release notes for ADCore:

 

### NDFileHDF5

  * Fixed issues with chunking.

    * User-defined chunking only supported 2-D arrays, plus the dimension for multiple arrays.

      This is insufficient, we need to be able to support N-dimensional arrays in a general manner.

    * Direct chunk write only worked with 1-D or 2-D arrays. This prevented it from working with RGB images, for example.

    * New records have been added for NDArray dimensions 2-9 (ChunkSize2, ChunkSize3, ... ChunkSize9).

    * The labels on the OPI screen for NumColsChunks and NumRowsChunks changes "Dim0 chunk size" and "Dim1 chunk size",

      added "Dim2 chunk size" to this screen as well.

    * Added a new NDFileHDF5_ChunkingFull screen that shows all 12 chunking related records (10 dimensions, NumFramesChunks, ChunkSizeAuto).

    * Previously the documentation said that if nColChunks and nRowChunks were configured by the user to the special value 0,

      they will default to the dimensions of the incoming frames. However, this was not really true, setting to 0 only worked once.

      If they were set to 0 and a 512x512 array was saved then the chunking was set to 512x512.

      If the user then disabled binning and saved a 1024x1024 array the chunking remained at 512x512, which was not desirable.

    * Added new ChunkSizeAuto bo record to allow always setting the chunking to automatic or manual.

      If ChunkSizeAuto=Yes then the chunking in each dimension is automatically set to that dimension of the NDArray, so

      the chunking is one complete NDArray. The automatic control does not affect NumFramesChunks.

  * Fixed a problem in NDFiileHDF5::calcNumFrames(), it was setting NumCapture to 1 if numExtraDims=0, which is incorrect.

    This caused the autosaved value to be replaced by 1 each time the IOC restarted.

  * Added support for JPEG compression.  There is a new JPEGQuality record (1-100) to control the compression quality.

    This requires ADSupport R1-8 which adds HDF5 JPEG codec filter and plugin.

    JPEG compression can be done in the NDFileHDF5 plugin itself, or in NDPluginCodec writing pre-compressed NDArrays with

    HDF5 direct chunk write.

 

### NDFileTIFF

  * Increased the last used defined tag (TIFFTAG_LAST_ATTRIBUTE) from 65500 to 65535,

    which is the maximum allowed by TIFF specification.

  * The last usable tag is now TIFFTAG_LAST_ATTRIBUTE, while previously it was TIFFTAG_LAST_ATTRIBUTE-1.

  * Fixed bug in memory allocation for user-defined tags that could lead to access violation.

 

### NDPluginStats

  * Removed calls to processCallbacks() when CursorX, CursorY, or CentroidThreshold are changed.

    This is no longer thread safe.

    To see updates when plugin is not receiving NDArrays change the PV and then press ProcessPlugin.

 

### NDPluginColorConvert

  * Added support for converting Bayer color to RGB.  Thanks to Arvinder Palaha for this.

 

### commonDriverMakefile

  * Added ADPluginBar and ADCompVision plugins if defined.  ADCompVision requires WITH_BOOST=YES.

 

### EXAMPLE_commonPlugins.cmd, EXAMPLE_commonPlugin_settings.req

  * Added commented out lines for ADCompVision and ADPluginBar.

 

 

These are from the R1-8 release notes for ADSupport:

 

  * Added HDF5 filter and dynamically loaded filter plugin for JPEG.

  * Fixed memory allocation functions used in blosc and lz4 HDF5 filters.

  * Fixed problem with hdf5PluginSrc/Makefile, only build plugins if SHAREABLE_LIBRARIES=YES.

 

 

The home page is here:
http://cars.uchicago.edu/software/epics/areaDetector.html

The top-level documentation is here:

https://areadetector.github.io/master/index.html

Installation instructions are here:
https://areadetector.github.io/master/install_guide.html

areaDetector release notes are here:
https://github.com/areaDetector/areaDetector/blob/master/RELEASE.md

ADCore release notes are here:
https://github.com/areaDetector/ADCore/blob/master/RELEASE.md

 

Mark