Display MoreI though this would be the case too. But here is the background spectrum Wyttenbach posted.
Remarkably smooth for a spectrum that has not been averaged! And the little wiggles that do appear here are no more than about 0.5 counts from peak to trough whereas the wiggles in the background spectra are about 4 times larger. So I don't think that this background account for features in the background-subtracted spectra.
Another thing I don't see in this background spectrum is the set of "most active discrete lines" that Wyttenbach describes in the text as 63.3, 92.4 and 92.8 keV. Wytttenbach has not provided a good explanation for this disappearance.
Due to the low counts per second with this system a background spectra takes a while to come into focus.
This is why I was wondering about channels and bins earlier. To get an idea of how much of a average is obtained over 10 minutes. I was also wondering how a more sensitive detector might work for background subtracting, but then normalization of CPS (CPM) becomes a probable point of argument.