HHG energy calibration

• batch calibration for images in a folder
• 2D background substraction (mean value of either x or y in image)
• spectral evaluation for y-axis of image according grating function
• integration of different calibration data (CCD, filter, grating)
• linear interpolation of calibration data
• 1/i dependency for spectral energy range to be integrated
• integration over spatial ROI
• resampling spectral bin ranges from non-linear to linear
• batch process for images in one folder, over certain energy ranges in one image
• approximation of Nphoton/sr @ 0.01 band width for picked energy-ranges over batch list

2-f Binning 'resample'

This project is an application of the so called 2-frequency binning method. 2f binning is commmonly needed when e.g. integrated pictures (spectra) have to be calibrated. Usually the spectra has a non-linear x-scaling, which is a consequence of the grating function. The calibration data set (e.g. measuring the sensitivity of a camera at different photon energies) usually contains of less data points with a variable distance (binning) in the x-axis. In order to do the calibration via broadcasting of arrays, the spectral binning has to be same. Hence, calibration data is linearly interpolated, the measured spectra is linearized - in order to achieve the same bin-size width. Prerequisite is, in order to avoid subsampling and by this alias or an increase of resolution that is not there (Nequist Theorem): the maximum bin-size1.5 existing in our spectral array limits the resulting binsize. The programm checks this, and if the chosen binsize (called self.result_binsize) is to small, it will be set automatically to the maximum bin_size x 1.5 of the array (rounded on one digit accuracy). After this the x-axis and accordingly the y-axis are redistributed to the new binning range.

what it can do:

• opens and further calculates all .tif files from given filepath (16 bit, 2048x2048)
• integrates picture over one dimension (space)
• takes at the right side of the picture a background, substracts mean value
• rescales the x- axis according a function (spectrometer function)
• resizes the x-binning to equi-distant binning (re- sampling)
• needs now 3 calibration files (2 colummn. txt) (e.g. camera, filter, grating)
• can transfer the calibration x-axis from eV to nm
• linearly interpolates the calibration files to the same binning as chosen for the picture (upsampling)
• calculates maximum range (spectrally) from the picture and resizes the calibration range accordingly
• multiplies the calibration to the integrated picture (spectrum)
• delivers the option to save separatly each file
• plots....

what is still not catched:

• stops working if the input binning of the calibration files is smaller than the wanted one
• it has no file dialog
• a more accurate handling of overhang bins (at the moment just delets and by this shifts one or two)
• code needs some cleaning
• code needs more tests

Python 3.7