Feedback is a loop. Sound comes out of a speaker, the mic picks it up, the system amplifies it, sends it back out, and the whole thing snowballs into that squeal everyone in the room hates.
Here's the good news: it's one of the most avoidable problems in live sound and recording. Prevention beats cure every single time. The fundamentals haven't changed in decades — but there's genuinely new tech worth knowing about, from automatic suppression to AI voice-separation. Let's get into it.
TABLE OF CONTENTS
What microphone feedback actually is

Feedback is an audio loop. The sound from a speaker gets picked up by a nearby mic, runs back through the amp, comes out the speaker again, and gets picked up again — louder each pass. That escalating cycle is what you hear as a ring or a squeal.
The key thing to understand: it's almost always a setup problem, not a gear problem. The same mic and speaker that howl in one room behave perfectly in another. Feedback depends on the room, the placement, and how much gain you're pushing. Fix the setup and the squeal usually disappears.
Prevent it at the source

The old saying holds up: 1dB of prevention beats 16dB of cure. If you set things up right, you'll spend almost no time chasing feedback later. Here's where to focus.
- Use directional mics. Cardioid and supercardioid mics reject sound coming from behind them. Point the mic away from your speakers — 90 degrees or more off-axis — and you cut off the loop before it starts. Picking the right capsule matters too; our roundup of the best vocal mics for live performance is a good place to start.
- Keep the mic close to the source. The closer the mic is to the singer or instrument, the less gain you need, and less gain means less feedback risk.
- Mind your gain staging. Aim for peaks around -12 to -6 dB. That gives you clean headroom without cranking the system into the danger zone.
- Position speakers smartly. Keep them away from reflective walls and out of the mic's pickup pattern.
- Treat the room. Foam panels, curtains, and rugs soak up reflections that feed the loop.
Less is more here. You don't need every trick at once — a directional mic pointed the right way and sane gain will solve most rooms on its own.
Don't cup the mic
This one trips up a lot of performers. When you wrap your hand around the grille of a directional mic, you basically cancel out the design that was rejecting off-axis sound. A cupped cardioid mic behaves like an omni mic — picking up everything, including the speakers.
It looks cool. It sounds terrible. Make sure your vocalists know to grip the body of the mic, not the ball on top.
Ringing out the room with EQ
Ringing out is a classic technique. You slowly push the gain until the system just starts to ring, find the frequency that's ringing, and notch it out with EQ. Then you push a little more, catch the next one, notch that. Every notch buys you more volume before feedback returns.
Feedback commonly lives somewhere in the 800 Hz to 2 kHz range, so that's a reasonable place to start hunting. But the real culprit is room-dependent — every space has its own resonances. Trust your ears over any chart.
With that being said, treat ringing out as one tool among several, not your whole plan. It helps, but it won't save a badly placed mic.
Automatic feedback suppression is standard now

Automatic feedback suppression, or AFS, used to be a pro-tier luxury you'd rack up separately. Not anymore. It's baked into plenty of digital mixers you can actually get your hands on today.
The most common method is automatic notch filtering. Instead of coloring your sound all the time, it watches for the specific frequencies about to feed back and drops a narrow notch on them only when the system's at risk. Your mix stays clean until it needs help. If you want a deeper technical read, the Wikipedia entry on feedback suppressors breaks down the different filtering approaches.
Most good implementations run two modes. You ring out the system before the gig to catch the obvious troublemakers, then leave live filters running during the show to snag anything new. The dbx AFS built into Soundcraft Ui mixers works exactly this way, and Shure has its own take built into their processing. A lot of the current digital mixing boards for live sound ship with this built in.
Honestly, the benefit is simple: precise detection means less audible degradation than yanking down a whole fader, and it runs without a full-time engineer babysitting every channel.
The new AI voice-separation approach

Here's the genuinely new development since 2023. Traditional suppressors detect a squeal, then notch that frequency. The AI approach flips that — it does continuous spectral separation, pulling the voice out of everything else in real time rather than waiting for a squeal to appear.
The marquee example is L-Acoustics Source Intelligence. It isolates and enhances the human voice while stripping out PA bleed and room reflections, which boosts your gain before feedback and lets performers walk right in front of the PA without setting off a loop. It's not vaporware either — it's been used in the field since Spring 2025 across major tours, Broadway, and large houses of worship.
Now the honest part. Like every current tool, it reacts to feedback energy that's already present rather than predicting it before it's audible. The bigger barrier is integration — getting these models running natively on console DSP is a licensing headache, not a technical one. And aggressive AI processing can push vocals toward a robotic, over-processed tone if it's leaned on too hard.
So it's promising, not magic. Worth watching closely, but the fundamentals above still do most of the heavy lifting.
In-ear monitors kill feedback at the source
The most common feedback culprit on a stage is a floor wedge throwing sound straight at the singer's mic. In-ear monitors solve that by getting rid of the wedge entirely. No wedge blasting the mic means no loop to fight.
The upside stacks up fast: no stage bleed, lower overall stage volume, personal mixes for every performer, and real hearing protection because everyone can monitor quieter. Bands at every level have largely moved this direction, and it's easy to see why. If you're weighing options, we compared our favorites in the best in-ear monitor systems guide.
A few honest caveats. Keep a couple of wedge stage monitors around as backup — if the IEM system drops mid-set, they'll save the show. IEMs can still damage hearing if you crank them, so respect the volume. And wired IEMs remain the pro standard, because even a few milliseconds of Bluetooth latency throws off a performer's timing.
Quick feedback-fighting checklist
- Use a directional mic and point it away from your speakers, 90 degrees or more off-axis.
- Set your gain so peaks land around -12 to -6 dB — clean headroom, no cranking.
- Tell your vocalists not to cup the mic grille.
- Ring out the room before the show and notch the frequencies that ring.
- For tricky rooms, lean on automatic feedback suppression or move to in-ear monitors.
Frequently Asked Questions (FAQs)
What causes microphone feedback?
What frequency is microphone feedback usually at?
Does automatic feedback suppression actually work?
Do in-ear monitors stop feedback?
Can AI really increase gain before feedback?
Final Thoughts
Feedback feels like a mystery when you're standing in front of a squealing PA, but it's really just a loop you can break. Directional mic, smart placement, sane gain, and a quick ring-out will handle almost every room you walk into.
The new AI tools and automatic suppression are worth knowing about, and they genuinely help in the tricky spots. But they're a supplement to good technique, not a replacement for it. Get the fundamentals right first, and the fancy stuff becomes icing.
Some of the links within this article are affiliate links. These links are from various companies such as Amazon. This means if you click on any of these links and purchase the item or service, I will receive an affiliate commission. This is at no cost to you and the money gets invested back into Audio Sorcerer LLC.