IntClean : Alpha Release
Closed Jan 11, 2024
66% complete
Prototype for internal stakeholder testing, for 'look and feel'.
Features added for cube imaging.
- Mask handling now uses images with 1/0, with hand-drawn regions immediately translated to a mask image. This allows the use of automasking.
- Added automasking support by calling task_deconvolve(niter=0) prior to each task_tclean() call, so that control return…
Prototype for internal stakeholder testing, for 'look and feel'.
Features added for cube imaging.
- Mask handling now uses images with 1/0, with hand-drawn regions immediately translated to a mask image. This allows the use of automasking.
- Added automasking support by calling task_deconvolve(niter=0) prior to each task_tclean() call, so that control returns to the GUI after mask generation but before the deconvolution iterations begin.
- Simplified the keyboard shortcuts for mask editing/copies. Visual toggle for masked/unmasked regions
- Added controls to switch across Stokes planes.
- Added the final restoration tclean at the end of the iteration sets.
- Performance improvement (via the bit-depth of color values sent for rendering). Tested for image sizes of 2048x2048. To do : Test on larger images (16kx16k) and decide if explicit downsampling is or is not required (and add to the to do list).
- Bug fixes for startup/restart/timeouts, etc.
- Populated more iteration control options from the GUI
- Added colormap support.
- Improved visual feedback when adding/subtracting a mask but before committing the change.