# What is Channel Number? (In gamma spectroscopy)

by CraigH
Tags: channel, gamma, number, spectroscopy
 P: 185 Really quick question. In a graph like this: http://en.wikipedia.org/wiki/File:Cs137_Spectrum.PNG Source What does the Channel Number represent? Also, just to make sure, am I right in saying the count rate is $\frac{\delta m}{\delta t}$ ? where m is the cumulative number of scintillations and t is time Edit: I've just noticed the graph says "number of counts" and not "count rate". Some graphs do though, so can I change my second question to: Is a count a single scintillation?
 Admin P: 21,591 Channel number is a band of energy (ΔE). There is a total energy, and one divides that into a number of channels. A count is one (detection) event or scintillation. The systems captures the counts, and the count rate is simply the difference in accumulated counts divided by the counting interval.
 P: 185 Thanks Astronuc. So the scintillation detectors are sensitive to the energy of the flash, but why do we divide this into discrete energy bands instead of just having energy on the horizontal axis?
Mentor
P: 11,038

## What is Channel Number? (In gamma spectroscopy)

Have you ever used histograms before? How else would you plot a set of e.g. 10000 experimentally-measured energy values in a way that shows how they are distributed versus energy?

After you've made a histogram of your data, then you can find a smooth curve that at least approximately "fits" the data, and plot that curve along an energy-axis that doesn't have "bins" or "channels", but that's a somewhat different beast.
 Admin P: 21,591 Actually, some systems may record energy. http://www.iaea.org/OurWork/ST/NE/NE...ectrometry.pdf IAEA-TECDOC-1363 Guidelines for radioelement mapping using gamma ray spectrometry data http://www-pub.iaea.org/mtcd/publica...e_1363_web.pdf
 P: 185 jtbell, ah yeah of course. Sorry I didn't think before posting that question.
Mentor
P: 10,511
 Quote by jtbell Have you ever used histograms before? How else would you plot a set of e.g. 10000 experimentally-measured energy values in a way that shows how they are distributed versus energy?
Make a Gaussian distribution out of every measured value, add them up.
It is possible to plot data without histogram bins, it is just impractical in many cases. Unbinned fits use very similar methods.

 Related Discussions High Energy, Nuclear, Particle Physics 4 Nuclear Engineering 12 Nuclear Engineering 2 Introductory Physics Homework 13 Advanced Physics Homework 1