Imagine this case:
We have two instruments with the same characteristics, and we scan the same sample in the same cup for both instruments. As you can see is the case we use to create standardization files, but is also the case when we create repeatability files.
We have two NIR files with the same samples in the same order and we want to subtract every spectra of one sample scanned in one instrument from the spectrum of the same sample scanned in the other instrument, so we get the difference spectrum. If we do this for all the spectra we get a NIR file with all the difference spectra.
This cannot be done easily in Win ISI , so we can to use Excel: First I export the NIR files to Excel, so I creating a book and in one of the Sheets (Sheet 1) I paste the spectra of the samples scanned in Instrument 1, and in the second Sheet (Sheet 2) I paste the spectra of the same samples scanned in Instrument 2. So now we can create a new Sheet (Sheet 3) with the difference of the same sample scanned in both instruments and extend the formula to all the rest.
We can save Sheet 3 as a TXT file and convert it to a NIR with the Convert tool of Win ISI, so we can see the difference spectra for all the samples. These difference spectra are what the repeatability file use, giving reference value of cero for every constituent in order to minimize the instrument variance when developing the calibrations.
in the next plot you can see the difference spectra in Excel:
And in this one the plot after exporting to Win ISI: