TA-software

Software for the transient absorption setup.

pyTA

Measurement setup and data acquisition.

Thanks to the OE group at Cambridge for some of the original code.

hdf5-converter

Conversion of data files (.hdf5) to useful things.

usage

Open an anaconda command prompt and cd to the TA-software folder. Then activate the environment and launch the software by running:

conda activate pyTA
cd pyTA
python pyTA.py

Then to use the hdf5 conversion tool, run:

cd ..
cd hdf5-converter
python hdf5-converter.py

When finished run conda deactivate.

development

Things to fix throughout the python code are denoted with @todo.

key things to implement

• Set up and test NIR detectors for sub-ps TA
• Set up and test NIR detectors for nsTA - current have a $\tau$ flip-related bug, among other issues
• B-matrix referencing (as an option?)

important things to fix

• Fix the quality control algorithm in dtt.py so that bad data is properly rejected but we don't get stuck in a loop of retaking the data point. Not sure yet what the solution is.
• Output correct metadata.txt file (e.g. seems to think we're always using the short delay stage currently), and output more things (the more information the better)
• Properly fix weird bug in saving which motor COM port was last used (seems to be due to putting a string into the last_instance_values.txt). Currently doesn't save it at all as a hack-fix!
• Fix $\tau$-flip related bug associated with NIR detectors and electronic delay. Not sure what the solution is yet; may be due to hardware, not software...
• Fix sharp features seen using electronic delay at late time delays. As above, could be a hardware thing...

nice things to have

• Have dark correction shots x in the Acquisition tab as well
• Log scaling of kinetic plot
• Move the hdf5-conversion tool into a new tab on the main software panel
• Put in an option for converting to .ufs files in the hdf5-conversion tool
• Show rough time remaining for an experimental run
• Random and Bilinear stepping order (currently only Linear)
• More options for the 'Exponential' time point model (how many points before time zero, initial spacing)