Radioactive atoms decay at random, at a rate that statistically fits a decaying exponential. It is characterised by a 'half life' which is the time at which a large group of the atoms will have lost half the original number, emitting particles like alpha or beta, and creating the same number of atoms of a new element.
The escaping particles were largely detected by Geiger tubes, although many other technologies exist for this- photographic film, cloud chambers, semiconductor detectors.... but many of these require later analysis. Geigers give an audible click for every detected particle, so you know what has happened as it happens. You can also show each detection as an upward tick on an analogue meter. Smoothing out the individual kicks ( 'integrating') creates a visible 'rate meter' rather than a counter.
Several of us have been involved in a discussion of writing code to analyse the blips/clicks of a Geiger counter. Of interest are the individual pulses, their rate, and the average interval and distribution statistics. For old times' sake I did the opposite- recreating the lovely orange glow of neon tubes.
Dekatrons have 30 electrodes arranged as 3 connected sets of ten. By suitably phasing the drive the glow transfers to the nearest phase1 electrode then to phase2, before relaxing back to the count electrode, but one place advanced.
Once divided down to rates below a few per second, an electromechanical counter can cope.
I added about 10 lines to Rod's wave analysis program. ( I included saving the amplitude as a csv file- which loads into a spreadsheet, but the OpenOffice spreadsheet was very slow in displaying this long file. Bit of a dead end)
You certainly need to average the bursts of sound- I used calculating absolute values and averaging each new value with the three previous ones.
This show the +/- waves made unidirectional with 'abs('.
..
and this show the smoothed result which is now easier to analyse.
