-
Bug fix, avoids crashing if the metadata for the objective is missing. This happened in an image from a microscope with a broken objective turret.
After the fix, this is what nd2tool reports:
nd2tool iiJPC526_20240828_001.nd2 --info Warning: Could not find channel/microscope/immersionRefractiveIndex Warning: Could not find channel/microscope/objectiveNumericalAperture 5 FOV in 3 channels: #1 'a594', λ_em=590.0 #FFB300 #2 ' tmr', λ_em=542.0 #00FF00 #3 'dapi', λ_em=432.0 #9900FF Bits per pixel: 16, significant: 16 dx=1000.0 nm, dy=1000.0 nm, dz=300.0 nm NA=0.000, ni=0.000 Objective Name: Uncalibrated Objective Magnification: 1.0X Volume size: 1024 x 1024 x 41 Looping: Dimensions: XY(5) x λ(3) x Z(41) Camera: Andor DU-888 X-9877
If it wasn't obvious that the information was wrong here, we could make a special case when the objective name is set to
Uncalibrated
.In that case it does not make sense to report these values (found in the metadata!)
Objective Magnification: 1.0X dx=1000.0 nm, dy=1000.0 nm
Furthermore we should also prevent nd2tool from reporting these default values:
NA=0.000, ni=0.000
- Updating to Nikon libraries v 1.7.6.
- Total switch to cmake for building and package creation.
- Parsing Camera name and Microscope name from the text section of the metadata. This part of the metadata isn't very well structured so I wouldn't be surprised if it doesn't work for all files. If it doesn't work no info will be printed and that's all.
- Added checks for some unlikely scenarios. Builds cleanly with clang,
i.e., with
scan-build make -B CC=clang DEBUG=1
as well as gccmake -B DEBUG=1
. - Added the --SpaceTx option to save one image per plane according to the SpaceTx convention. This closes [#4]. Please note that for fov_id the name of the nd2 file is used (which can contain underscores) and 0 is used for round_label. In other words, you might need to rename the files (for example with rename in linux) before processing them with starfish.
- Added the --dry option for dry runs which tells the program not to create or modify any folders or files.
- Added the --deconwolf option to generate a script with commands to run deconwolf on the tif files (unless converted with the --composite flag).
- Factored out some code into
nd2tool_util.c
- Experimental support for writing composite images with the --composite flag. It remains to choose color maps that makes sense.
- Added the --coord command line option that will print the position of all images in csv format to the standard output. To be validated together with --shake.
- Experimental (still to be validated and hence not enabled by default) shake detection in the axial direction, enable with the --shake command line argument.
- Slightly improved log file messages.
- Fixed problem with non quadratic image sizes.
- Not saying when it is using the BigTIFF format any more.
- added approximate conversion from emission wavelength to color.