I have been around mics and amps for 50 years and am well acquainted with the unintentional feedback from a mic to an amp/speaker. The mechanism seems simple enough: The mic picks up a sound, the amp amplifies it and sends it out the speaker, and the mic picks up the speaker output and creates an infinite loop until some sound guy kills the mic. But why isn’t it a random tone, the resonant frequency of the room, white noise, or something else, but rather always that ice-pick-in-the-ear tone?
The system will resonate where the gain maximises the feedback signal. There are lots of resonant terms, the most obvious being the distance from speaker to microphone and the frequency determined by that delay. But the system can lock onto a frequency that is a multiple of that, basically putting multiple full waves into the gap. The round trip gain will vary with frequency and the system will tend to find its way to the frequency where the gain is highest. A setup used for voice will probably feedback right in the middle of voice frequencies.
Its not about standing waves and numbers of half wavelengths… because the room has all different distances eg the speaker itself has causes different distances , and it happens whatever angle the speaker is at the to the surfaces in the distance.
There’s four main frequency responses to look at.
-
The microphone
-
the amp and speaker
-
The air
-
The building - the walls,floor, ceiling.
But well they designed 1 and 2 to be suitable to provide a good copy of the sound,given 3 is also kind of constant.
So the most likely culprit is the building …
What frequency noise do you hear from outside the building ? The bass.
The treble reflects internally, the bass passes through to the outside
Its the frequency response… reflect , absorb or transmit ? , in the building materials.
The audio system is not perfectly “flat.”
A typical mike will have a response peak somewhere at the high end, like this:
That will generally set the feedback frequency, unless the speaker’s peak is greater, and at a different frequency. But, one of them will win.
Yup. Round trip gain against frequency is what sets the frequency.
You need to maintain phase at that frequency so the precise round trip delay must be multiple of the period of the feedback frequency.
This is also why you’ll commonly hear feedback start at a lower note and quickly (well under one second) increase to the final stable pitch even as it’s increasing to the final (often unbearable) intensity.
The starting frequency is more or less random, while the finishing frequency is well-determined by the setup. It happens that the typical resonant parameters usually put the finishing frequency pretty high up in the range of frequencies used for speech and music. So “starts low → ends high” is more common than “starts real high → ends lower but still high”.