1
Feedback Theory
Feedback: the howl or ring, happens when sound from a speaker gets back into a microphone, is amplified, comes out the speaker again, and re-enters the mic, looping. It runs away the instant the loop gain (the total gain around that whole mic-to-speaker-to-air-to-mic path) exceeds 0 dB at any single frequency.
That is the key insight: feedback is a property of the whole loop, not a bad microphone. Every fix: mic placement, speaker placement, EQ notches, mic directionality: is just a way to pull that loop gain back below 0 dB.
In practice: A monitor still feeds back even turned down low. The loop gain is still over 0 dB somewhere, usually mic-and-wedge geometry or a room resonance, so you fix the geometry or notch the frequency, not just the level.
Feedback = loop gain over 0 dB at some frequency.
2
Feedback Prevention Hierarchy
Feedback fixes have an effectiveness order, and most people start at the wrong end. The hierarchy, most to least effective: reduce stage volume (the root cause), improve mic technique (closer placement, right pattern), optimize monitor position (into the mic's null), and only then EQ notching.
Notching is the last resort. Exhaust the physical fixes first, they solve the problem, EQ just treats the symptom.
In practice: A band fighting feedback with a loud stage amp and a vocal mic three feet away: turn the amp down and close the mic distance, that fixes 90% before any notch.
Feedback hierarchy: stage volume → mic technique → monitor placement → EQ notching.
3
Source Distance Beats Every Filter You Own
Engineers reach for the graphic EQ when they run out of gain before feedback, and it is almost always the smallest lever in the room.
Gain before feedback is a ratio between source level and open-loop gain. Notching attacks the loop, one narrow band at a time, and each notch costs tone. Moving the microphone closer to the source attacks the numerator, and because level follows the inverse-square law, halving the source-to-mic distance buys about 6 dB. No realistic number of filters returns 6 dB, and distance costs you nothing.
That makes performer technique the largest single variable in the whole feedback equation. A singer at two inches and the same singer at eight inches are two completely different systems, and the difference is bigger than your microphone choice, your wedge position and your entire ring-out combined.
The uncomfortable part is that it is a people problem, not a console problem. It is solved with a conversation at soundcheck, framed as what the performer gets rather than what they are doing wrong, and never in the middle of a song. Everything else, tighter patterns, better wedge geometry, more and quieter wedges, IEMs, is a way of applying the same lever when the performer will not move.
In practice: A singer at eight inches asking for more wedge while you are twelve notches deep: those notches are chasing about 10 dB that a change in mic technique would hand back for free.
Halving source-to-mic distance buys about 6 dB of gain before feedback and costs no tone. No amount of notching competes, which makes mic technique the biggest lever you have.
4
Cardioid, Supercardioid, Hypercardioid: Null Points and Monitor Placement
A cardioid mic's deepest rejection (its null) is straight behind it, at 180 degrees. Tighter patterns are more directional overall, but they buy that by trading the single rear null for two side-rear nulls plus a small rear lobe: a hypercardioid nulls at about 110 degrees, a supercardioid at about 126, and both hear a little from straight behind.
That changes monitor placement: a wedge directly behind a cardioid sits in its null, but the same spot behind a hyper or supercardioid is a pickup lobe. Put the wedge at that pattern's actual null angle, or feedback gets worse, not better.
And read the spec sheet rather than the reputation. The Beta 58A, the mic most often reached for when someone wants "a tighter pattern than an SM58", is a supercardioid: put its wedge at 126 degrees. Assuming 110 because you filed it under hypercardioid puts the wedge 16 degrees off the null, which is exactly the mistake this card exists to prevent.
In practice: Switching a vocal to a Beta 58A for more gain but feedback worsens: the wedge is straight behind it, in the rear lobe. Move it to the supercardioid null near 126 degrees, roughly one wedge-width off dead centre.
Tight patterns null at 110 (hyper) or 126 (super) degrees, not 180. Check which one the mic actually is, then put the wedge there.
5
Multi-Mic Feedback: The NOM Rule
The NOM rule (Number of Open Mics) is a hard physical fact: every doubling of open microphones costs 3 dB of gain-before-feedback. A band with eight open mics has 9 dB less headroom than a soloist with one.
That is why muting is not optional housekeeping, it is feedback physics. Fewer open mics directly buys back gain.
In practice: A band with ten open mics where the vocal can barely come up before ringing: mute every mic not being played, going from ten active to five recovers about 3 dB.
Every doubling of open mics costs 3 dB of gain-before-feedback, mute the unused ones.
6
Gain Before Feedback
Gain before feedback is how much level you can get out of a system before it howls, and it is a budget you can actively increase. The biggest lever is mic placement: a mic close to its source needs far less system gain, which buys enormous feedback margin. Beyond that: use more directional mic patterns, move speakers away from mics, keep unused mics muted (every open mic costs you), and notch the worst frequencies.
Note what does not help: turning the monitor up. That raises loop gain, it spends the budget, it does not grow it.
In practice: A performer cannot hear themselves and you are already near feedback. Do not push the send, move their mic closer to their mouth or instrument so it needs less gain to begin with.
Close mics and tight patterns buy margin; turning up spends it.
7
Feedback-Prone Venues: Churches, Gyms, and Hard-Walled Rooms
Some venues are inherently feedback-prone, churches, gyms, hard-walled rooms. The cause is shared: hard parallel surfaces making strong reflections, long reverb times, and sometimes focusing geometry like domes that concentrate sound back at the stage.
The strategy is acceptance plus control: keep SPL lower (less room excitation), use directional speakers aimed only at the audience, mute aggressively, keep the mix dry, and accept that monitor volume will be tightly limited.
In practice: A stone church with a 3-second reverb: push IEMs over wedges, aim directional speakers strictly at the seats, run moderate SPL, and keep the mix dry, the room supplies its own reverb.
Hard reverberant venues: lower SPL, directional speakers, aggressive muting, a dry mix.
8
Every Notch Costs Tone: Knowing When to Stop
A ring-out feels like free headroom and is not. Every notch is a transaction: you buy gain before feedback and you pay with a hole in the response, cut out of the same band the voice lives in.
Three or four narrow cuts are ordinary tuning and largely inaudible. Past roughly six you have removed enough of the vocal range that the wedge goes loud without going clearer, and the performer starts reporting the specific complaint that gives it away: it is loud and I still cannot hear myself. That is not a request for more level, it is a description of a response with a dozen bites taken out of it.
So treat notch count as a diagnostic rather than a score. A high count means there is an upstream cause you have been paying to work around: mic technique and distance, wedge geometry relative to the microphone null, a hard surface behind the performer, the wrong capsule, stage volume, or too many open mics. Find it, fix it, flatten the graphic and ring out again from scratch.
And keep every cut you do make narrow, because a wide filter removes far more programme material for the same headroom.
In practice: A singer telling you their wedge is loud but they cannot hear themselves after twelve notches: the notches are the problem now. Find the upstream cause, flatten the graphic, and start over.
Notches buy headroom and cost tone. A high count is a diagnostic pointing upstream, not a job well done.
9
HPF on Everything
A high-pass filter (HPF) removes the lowest frequencies from a channel. In live sound you engage it on nearly every channel, vocals, guitars, fiddle, even kick and bass set just below their useful range, because almost no source needs that sub-low energy, but every channel picks some up: stage rumble, foot noise, mic handling, HVAC, wind.
On its own each channel's low junk is small. Across a whole console it sums into a muddy mess that eats headroom. The HPF stops that buildup before it starts.
In practice: The mains sound muddy even though every instrument sounds clean soloed. That is low-frequency buildup across unfiltered channels, engage HPFs on all of them and the mix clears up.
HPF nearly every channel, low junk sums into mud.
These lessons are the teach side of knowledge cards from the Live Sound deck. In the daily plan each one is followed by a quiz, spaced over weeks, so it stays known rather than read once.