Hi there, I’m looking for any info on reducing microphone feedback in PA speakers.
As I understand it inductance plays a role in amplifiers (as vague as that is- I don’t know in which part of the amplifier inductance is an issue). I’m hoping my ICEpower 125ASX2 has been designed with this in mind.
The thing I may be able to change, however, is the drivers I’m using.
Are there TS parameters which relate to feedback? The ability of the suspension to arrest the driver? Voice coil inductance?
Tweeters seem to have the worst problem with feedback, I’m guessing this has to do with the frequency range they cover (which is increasingly directional as the frequency rises). But are there other factors? I’m guessing 1.4” exit 3” diaphragm drivers are a bit more feedback resistant due to their higher diaphragm mass and greater mid range capability, but not sure.
Any ideas welcome as I have tried various Google search terms and found nothing relating directly to this question.
As I understand it inductance plays a role in amplifiers (as vague as that is- I don’t know in which part of the amplifier inductance is an issue). I’m hoping my ICEpower 125ASX2 has been designed with this in mind.
The thing I may be able to change, however, is the drivers I’m using.
Are there TS parameters which relate to feedback? The ability of the suspension to arrest the driver? Voice coil inductance?
Tweeters seem to have the worst problem with feedback, I’m guessing this has to do with the frequency range they cover (which is increasingly directional as the frequency rises). But are there other factors? I’m guessing 1.4” exit 3” diaphragm drivers are a bit more feedback resistant due to their higher diaphragm mass and greater mid range capability, but not sure.
Any ideas welcome as I have tried various Google search terms and found nothing relating directly to this question.
Normally, it matters where you have the mics pointed, how directional they are, how many are turned on at once, where the speakers are pointed, and the frequency response around the acoustic feedback loop (speakers, room, mics), etc.
If the mics and speakers are chosen appropriately and set up appropriately, the usual thing is to then use notch filters to reduce the volume level of frequencies that tend to feed back the most.
There are some books that explain all this stuff pretty well. For one example: https://www.amazon.com/Sound-Reinforcement-Handbook-Gary-Davis/dp/0881889008
Here's another: https://www.amazon.com/Sound-System...0e-b577-fad2c1e89d0f&pd_rd_i=0415731011&psc=1
If the mics and speakers are chosen appropriately and set up appropriately, the usual thing is to then use notch filters to reduce the volume level of frequencies that tend to feed back the most.
There are some books that explain all this stuff pretty well. For one example: https://www.amazon.com/Sound-Reinforcement-Handbook-Gary-Davis/dp/0881889008
Here's another: https://www.amazon.com/Sound-System...0e-b577-fad2c1e89d0f&pd_rd_i=0415731011&psc=1
I was aware of all of those factors, but some speakers are less feedback prone than others.
From experience using a graphic EQ in a live setting, and having been instructed at university, you can only do so much.
Generally speaking I find a maximum of two or three notches and -3 or 4 dB on each, plus a bit of high shelving makes any real difference. Any more and what you’re doing is just turning the level down (which may initially appear an advantage) or distorting the response of the microphone.
The different speakers I’ve owned or hired have different GBF characteristics, so I’m mainly interested in differences in the design of the woofer and tweeter.
Doesn't that have a lot to do with frequency response and directivity? Both speakers and mics have horizontal and vertical polar patterns, e.g.
The patterns can change depending on local boundary conditions too. For example, the face of a person standing close behind a mic can act a reflector which bounces sound from the speakers back into the sensitive side of the mic. Therefore, conditions that cause feedback can change dynamically. IME they also change from when a room is empty to when it gets full of people. Nowadays I believe it is possible have a lot more ability than we used to have to assess and adjust system behavior in real time (e.g. active beam steering). However, sophistication may come at some cost. What you can do may depend on what type of sound reinforcement you are trying to do, what kind of budget you are trying to work with.
The patterns can change depending on local boundary conditions too. For example, the face of a person standing close behind a mic can act a reflector which bounces sound from the speakers back into the sensitive side of the mic. Therefore, conditions that cause feedback can change dynamically. IME they also change from when a room is empty to when it gets full of people. Nowadays I believe it is possible have a lot more ability than we used to have to assess and adjust system behavior in real time (e.g. active beam steering). However, sophistication may come at some cost. What you can do may depend on what type of sound reinforcement you are trying to do, what kind of budget you are trying to work with.
Yes, but I don’t think it’s only that.
What I’m working on is a monitor, so it’s on-axis (well, about 180 degrees off axis from the microphone, but as far as the polar pilots of the speaker is concerned). I know different pick-up patterns make a significant difference, but that’s not what I’m trying to correct, because the speaker is supposed to work in multiple situations.
I think maybe I mentioned a couple of the factors in my original post. I got the inductance idea from George Krampera from KV2. He explains that amplifier design has a lot to do with it, one of the factors being inductance; in a feedback resistant amplifier, the output stops immediately as soon as the input signal stops. Otherwise the amplifier itself is a feedback generator as the signal has a tail after the input signal stops.
I’m guessing the ICEpower amplifier is pretty good in this regard as it’s certainly not the cheapest board, and is used in a variety of PA speakers and guitar amplifiers.
I was wondering if there are similar characteristics in a driver, and I’m pretty sure the suspension stiffness will be one of them. But wondering if there are further details.
What I’m working on is a monitor, so it’s on-axis (well, about 180 degrees off axis from the microphone, but as far as the polar pilots of the speaker is concerned). I know different pick-up patterns make a significant difference, but that’s not what I’m trying to correct, because the speaker is supposed to work in multiple situations.
I think maybe I mentioned a couple of the factors in my original post. I got the inductance idea from George Krampera from KV2. He explains that amplifier design has a lot to do with it, one of the factors being inductance; in a feedback resistant amplifier, the output stops immediately as soon as the input signal stops. Otherwise the amplifier itself is a feedback generator as the signal has a tail after the input signal stops.
I’m guessing the ICEpower amplifier is pretty good in this regard as it’s certainly not the cheapest board, and is used in a variety of PA speakers and guitar amplifiers.
I was wondering if there are similar characteristics in a driver, and I’m pretty sure the suspension stiffness will be one of them. But wondering if there are further details.
From experience, feedback is not directly related to speaker design. Feedback is dependent on the acoustic properties of the microphone and speaker location, the dynamics of the room. First establish the frequency of the feedback - work from there. Every cubic foot of a venue has different acoustic characteristics. You may think you've a perfect set-up but feedback can start by a vocalist lowering a mic 6".
That's complete, utter nonsense.
The directivity and orientation of both the speaker and mic, and distance between them,
are the primary factors for acoustic feedback.
The situation is not that I’m trying to reduce feedback at a gig, I know how to do that. It’s to do with the acoustic/electro-acoustic design of the speaker. There must be differences, otherwise that aspect would be left out in the design of a D&B M2 or MJF-210, which I find hard to believe.
Feedback is feedback, right? An oscillator is a feedback device where the feedback gain and phase shift at some frequency are just right for oscillation to occur. With a speaker, it projects sound waves into space and or pressurizes space, let's say. It does that a bit differently depending on direction and frequency. Well, if we want to get more deeply into it, the state of a speaker is in part dependent on its recent history: cone position, velocity, acceleration, cone vibrational modes, etc. We can simplify that and just say there is always a feedback path through the air from a speaker back into a mic. Always. The question is what how much sound pressure, at what phase, at some frequency, is getting back into the mic. At some point the conditions required for oscillation will occur. That's when we get the unwanted squeal we call 'feedback.' However, in reality there is always feedback, just not always producing a squeal that is really an oscillation of the feedback system. Does that make sense?
listen from 4 minutes.
Plus, as I have explained now at length, I’m not interested in how the equipment is used, but the other factors involved, small though they may be by comparison with those you mention, of which I’m well aware.
Again- I’m pretty certain D&B, Meyer, L-Acoustics do something to apprehend the issues I’ve mentioned.
How else do you think there is any difference between the feedback rejection performance of various wedges? As I point out, the speaker is pointing directly at the area the microphone is picking up (yes the microphone is the other way round, but you still get feedback in the mid and low frequencies in systems to do only with the proximity of microphones to the speakers and not the direction they’re pointed in).
Point source verses line array has to do with the shape of the wavefront launched into the air. Conventional speakers can be roughly approximated as a point source (if you are far enough away from it), whereas line arrays approximate line sources. The wavefronts expand and propagate through the air differently depending on the shape of the source (with line arrays its is also possible to do some beam steering by adjusting the relative timing of the different speaker cones, but that just approximates a source with a different shape, that is to say, it effectively no longer a perfect line shape). One could also have a planar source, such as the electrostatic flat panel speakers I use. Its basic physics stuff.
That doesn't change the basic principle of a feedback system that can oscillate.
That doesn't change the basic principle of a feedback system that can oscillate.
Last edited:
Yes as I say I’m aware of all that.
But it is known in circles interested or involved in live sound that speakers have different properties with regards to GBF.
For example, Dave Rat who does the Chilli Peppers shows spends time creating videos on YouTube analysing these differences.
He says no.1) it’s to do with point source vs line array (that’s a tangible example of what I’m indicating actually, but never mind)
no.2 it’s to do with control of the speakers, by which he means the motion of the diaphragm of a driver (‘speakers’).
It changes the directivity of the speaker is all. That can be done electronically if you have enough speaker cones distributed over a large enough area. As I said before, doing that changes the relative timing of each speaker, which then approximates a different source shape for the overall collection of speakers, which then produces a differently shaped wavefront in the air. Still just basic physics.
No. As a scientist and engineer, I do not need to hear someone say that an amplifier causes acoustic feedback.
That is wrong.
Of course, jerking the tone controls up or down can change the behavior of acoustic feedback.
But that is not what you/he are claiming. Electrical feedback in an amplifier is part of how it works.
That has nothing whatsoever to do with acoustic feedback.
Last edited:
Regardless, the basic facts remain the same. A perfect amplifier equally amplifies all signals within the specified spectrum. A perfect loudspeaker system equally reproduces all frequencies. A perfect microphone responds equally across the audible spectrum.
There are no variables in perfect equipment.
All variables are environmental; including crowd density, humidity, room temperature, and atmospheric pressure. Ergo, bespoke studios don not have feedback issues.
I’ve used a lot of different systems, some of it is to do with sensitive frequencies, probably to do with the natural response of the drivers or the geometry of horns used, a lot of it is to do with the way the speakers are pointed (though I cite the example of a monitor speaker), some of it is to do with the way the speakers are directed, some of it is to do with the room.
But some of it is definitely to do with the amplifier and drivers, otherwise the top manufacturers charging £2000 and up for a single passive speaker would be rather out of a job.
There’s no need to denounce something as ‘utter nonsense’ as soon as you see it.
As a scientist you’ll know when you look at things carefully they change, and other factors come into the mix that you simply had not considered.
But some of it is definitely to do with the amplifier and drivers, otherwise the top manufacturers charging £2000 and up for a single passive speaker would be rather out of a job.
There’s no need to denounce something as ‘utter nonsense’ as soon as you see it.
As a scientist you’ll know when you look at things carefully they change, and other factors come into the mix that you simply had not considered.