By "magnetic induction", I mean "voltage generated by a changing magnetic field", i.e., invoking Faraday's law. Although an electric field is created as a result, it is not the dominant part, in the sense of sqrt(E/H) = Z_fields is below Zo (impedance of free space), i.e., H is dominant.
Likewise, "electric induction" (or electrostatic) is the voltage generated by a changing electric charge. This is dominant when sqrt(E/H) > Zo.
When the frequency and distance scales are well within 1/4 wavelength, this is strictly near-field behavior, so there's no concept of an "antenna" as a device to interface with propagating (electromagnetic, free space, Z_fields = Zo) waves. Rather, it can be expressed as an equivalent circuit with coupling capacitances or mutual inductances. Another way to put it: the physics can be modeled reasonably well by taking the applicable Maxwell's equations, independently -- the assumption and implication being the speed of light is infinite, or that the system is "static" (thus, electrostatics and magnetostatics, separately). ("Static" in quotes, because we're assuming that time derivatives are nonzero -- otherwise there's no induction! "Quasi-static" is probably a more correct term.)
Putting the physics aside, as long as we can measure the consequences of that analysis -- namely, an equivalent circuit of capacitances and inductances, we can model pretty accurately what's happening.
So that's where it's coming from. At 50/60Hz, we're well within 1/4 wavelength (> 1000 km), pretty much anywhere in the world. So we can model electrostatically induced voltage (high impedances) as capacitive coupling, and magnetostatically induced voltage (low impedances) as inductive coupling. The former only works when there is a [mostly] open circuit, because the impedance of the effect is high; the latter only works when there is a [mostly] closed circuit, because the impedance of the effect is low.
An audio system might have hum introduced in various ways. When plugging in a signal cord, that's most likely electrostatic at work: the cable (and probably the hand gripping the cable) has some 50/60Hz charge on it, which causes the amplifier to BUZZ briefly as the connector mates (assuming RCA connectors, which are made backwards like this
![Wink ;)](https://www.eevblog.com/forum/Smileys/default/xwink.gif.pagespeed.ic.cldandycH0.png)
). (If the cable is connected to powered, ungrounded equipment, the buzz might be explicitly due to capacitance in the power supply circuit.) That voltage gets shorted out when the grounds connect, because the low impedance of the grounds dominates.
However, there may be magnetic induction in the loop formed between amplifiers and sources, if they share multiple ground paths -- creating loops. These loops create voltage drops across the ground conductors in the circuit, so that at any given amplifier input, there may be a voltage difference between signal (no voltage drop!) and ground, which it senses and amplifies.
Ground loop may also be resistive in origin, since incidental AC current flow through any of those grounds can also generate that voltage drop directly, in which case the voltage simply distributes through the loop according to Ohm's law. Like I said, there is usually a difference in voltage between mains neutral and ground, and may be due to induction as well as resistive drop.
Tim