├ noise
A burst is one cascade. The field goes up by two thousandths, something tears loose, everything that can follow follows, and then it stops. The size of a burst is the number of sites that flipped before it stopped.
Every burst is in the log below, in order, and none are ever removed. Each line carries the time, the field it happened at, the number of sites that went, and the remanence immediately afterwards. That list is the only record this site keeps, and every figure on every page is computed from it plus the state of the wire.
Bursts have no typical size. This is the whole reason anyone still studies the effect. There are far more small ones than large ones, but there is no characteristic scale in between, no average burst that the others cluster around. Plotted with both axes logarithmic, the distribution is close to a straight line, which is what a power law looks like and what a normal distribution does not.
The straight hairline on the histogram is a slope of minus one and a half, roughly what this model predicts at critical disorder. It was drawn from the exponent, not fitted to our data. If the bars settle onto it, that is the model behaving as published. If they do not, that is more interesting than if they do, and it will not be corrected.
The sizes to watch for are the extremes. A run made almost entirely of single-site bursts means the wire is tearing one thread at a time and the disorder is effectively too high. A run containing one enormous burst that takes most of the wire at once means the opposite. Critical disorder is the setting where neither happens and every size in between shows up.
Nothing here is smoothed, averaged or binned before it is recorded. Every burst is logged at the size it actually was, in the order it happened. The figures are drawn from that list afterwards, so anything visible in them can also be found by reading the log directly.
Quiet is the normal state. The wire spends most of its existence pinned, doing nothing, waiting for a field it has not seen. The counter above says how long it has been. A long quiet stretch is not a fault and there is nothing on this site that will shorten it.
Barkhausen's original apparatus made all of this audible, and the sound toggle on this page does the same thing crudely: one click per burst, louder for bigger ones. It is off by default. It is not a feature, it is the original experiment.
├ log
the wire has not moved yet.