P.O. Box 58126, Raleigh, NC 27658

Help topics

In search of better sound
By Larry Stover soundsys@atlanticaudio.com
Feedback - why it happens - How to prevent it

Feedback - it's every soundman's nightmare and probably the last sound you would ever want to hear in a church or meeting facility. Every sound operator has experienced the problem of feedback although some are still not sure what to call it. I receive calls regularly from distressed soundmen who ask "What is that high pitched squeal coming from the sound system? How can I get rid of it?" The correct terminology is "acoustical feedback" and it is also correctly referred to as re-entrant feedback". This type of feedback is also described as "ringing" or "oscillation". FEEDBACK should not be an accepted part of the program. It can be and should be prevented. It is my project to present here important information about why feedback occurs and how to control it.

Basically feedback occurs when the sound from the loudspeakers is increased in volume to the point that it re-enters the microphone at a volume level equal to the talkers voice. At this point the microphone is sensing (or hearing) the loudspeakers instead of the talker. The signal then begins to race around in a vicious circle from microphone to loudspeaker to microphone and the signal is being amplified continuously as it passes through the electronics. Result - that awful sound, that high pitched squeal, ringing, and oscillation, FEEDBACK.

It does not have to happen. Simple solution - prevent the microphones from hearing the loudspeakers. A simple concept, but I didn't say it would be easy. There are many factors involved, but the most important are these: (1) - Choice of microphones. (2) - Number of microphones. (3) - Choice of loudspeakers. (4) - Acoustical characteristics of the room. (5) - Position and orientation of loudspeakers and microphones. (6) - Choice of electronics - specifically processors and equalizers. Many other factors have an influence but most are not important.

In the overall performance of your sound system no decision will equal the importance of selecting the right microphones. Microphones are the beginning of the " audio chain" (a topic for future consideration). Bad choices at this end of the audio chain can not be offset by having excellent electronics and or loudspeakers. No amount of equalization can make up for a bad microphone. Select microphones, which have excellent "linearity". That means freedom from peaks, or other aberrations (inconsistencies) in the frequency response. A wide frequency response (say 60hz to 15khz) is good but not essential, linearity is essential. Perhaps the most important factor with respect to feedback control will be the pick-up pattern of the microphone. Often referred to as "polar pattern", this specification will explain the microphone's sensitivity to sound coming from different directions.

Omnidirectional microphones have equal sensitivity to sounds coming from all directions. That make Omni's a bad choice for feedback control. Cardioid pattern microphones have reduced sensitivity to sounds coming from the sides or back and greatest sensitivity to sounds coming from in front of the microphone. Most Cardioid type microphone offers excellent feedback suppression making them a good choice for feedback control. Other polar pattern types such as super-cardioid and hyper-cardioid have increased sensitivity to sounds from the sides. Understanding the polar pattern of the microphone becomes even more important when you are deciding on location and orientation of monitor loudspeakers. As much as possible, monitors should be placed in the least sensitive area of the microphone's polar pattern. This information applies equally to cabled and wireless microphones.

Increasing the number of open microphones greatly increases the risk of feedback. Double the number of open microphones and you will increase the risk of feedback by a factor of 4. That means any microphones not in use should be off. That also means that any system that uses several microphones open at the same time will present a greater challenge to control feedback. This emphasizes the need to have made good decisions on the front end of things and to operate the system with great care and in a manner that is consistent with the limitations of the system.

The distance from the talker to the microphone is also an important factor. Close talking allows reduced system gain, which will substantially reduce the risk of feedback. When the distance from the talker to the microphone is doubled the risk of feedback is increased by a factor of 4.

Any system, even the most carefully designed, will at some volume level produce feedback. The question is, will the system produce adequate volume levels for everyone to hear comfortably before it reaches the point of feedback. Is there a safety margin between the level at which feedback occurs and the level required for adequate hearing? In audio terms that safety margin is called "headroom". It is desirable to have substantial headroom (6 to 10 decibels or more) in any system. Too frequently systems have little or no headroom because of design mistakes or improper operation. As an example, consider a system that has 10 good quality cardioid microphones and 1 Omnidirectional Mic. The 1 Omni Mic becomes the limiting factor and no matter how good the rest of the system is there may be inadequate headroom. It is the "weak link" in the audio chain. Get rid of the Omni Mic and suddenly you have a big increase in system headroom. (It's just an example.) Find and eliminate the weak links in your audio chain and you will find improved headroom.

CHOICE OF LOUDSPEAKERS

Your choice of loudspeakers will greatly influence the overall performance of the system and the amount of headroom available. It is important to select loudspeakers, which will provide as much linearity as possible. If loudspeakers do not provide adequate coverage or have substantial peaks (lack of linearity) in frequency response it will greatly increase the risk of feedback.

Loudspeakers and microphones are called "transducers" because they convert energy from one form to another. Acoustical to electrical energy (microphones) and electrical to acoustical energy (loudspeakers). If linear transducers are selected from the beginning it will put less burden on the other components in the system and the system operator.

In some cases the room may be the greatest cause of feedback. The amount of sound absorbing materials in the room will be a definite factor. If most of the surfaces in the room are hard it will cause reflections that contribute to peaks in the frequency response. That will result in poor sound overall and also increase the risk of feedback.

POSTITION AND ORIENTATION OF TRANSDUCERS

The distance between microphones and loudspeakers an/or monitor loudspeakers is another important factor. Greater distance results in lowered risk of feedback. Orientation is important also. Loudspeakers should not be aimed at microphones. The microphone's polar pattern should be considered when deciding on placement of loudspeakers, especially monitors.

I have seen some rooms where the loudspeakers were positioned behind the talker. That means the output of the loudspeaker was directed into the microphone. In that situation system gain (volume) will be extremely limited. Poor sound qualities as well as feedback are likely to be a problem. In this case the position and orientation of loudspeakers and microphones may be the weak link in the audio chain. Move the loudspeakers to a position forward of the talker and microphone and suddenly a large increase in headroom occurs.

Yes, you can improve the overall sound quality and reduce the risk of feedback by the use of correct equalization. A system equalizer properly set up will definitely improve headroom. Usually that would mean using a 1/3 octave equalizer and set-up by instrumentation. Normally this requires a 1/3-octave spectrum analyzer, a pink noise generator, and a trained technician experienced in this type of set-up. I know a few technicians who say they can set a 1/3-octave equalizer by ear. They probably can also change a flat tire with a soldering iron.

It is important to note that improper equalization is usually far worse than no equalization. A point that is frequently overlooked is that each group of loudspeakers should be separately equalized. That means that mains, monitors, remotes, rear fill loudspeakers, and/or time-delayed loudspeakers should each have there own source of equalization. To do otherwise would mean that some loudspeakers would receive a totally incorrect frequency adjustment since the types of loudspeakers are different and each is in a different acoustical environment.

Today there exists a new element of feedback control. It's called the automatic processor. It automatically finds and eliminates the frequencies that cause feedback. And it does this in a very precise narrow-band manner (1/10 octave) that will remove the feedback without destroying the quality of the sound. It requires no technician or special instruments to set-up. Use of this processor will normally result in an improvement of 6 to 9 decibels in available headroom. This type of processor has been available only since 1992. The device I have used with excellent results in over 100 system installations is the Sabine FBX-901, a single channel device. This Unit is now available, as the Sabine FBX-1020. This same technology is available in a dual channel device, the FBX-2020.

For additional information on this subject or sound systems in general you may want to obtain the instructional video program "UNDERSTANDING SOUND SYSTEMS' (1 Hour, 48 Minutes, $39.95 + 3.95 shipping and handling.) Order from Atlantic Audio Systems, P.O. Box 58703, Raleigh, NC 27658. 1-800-701-7686.

The writer, Larry Stover has over 20 years experience in designing and installing sound systems for churches and meeting facilities (over 2000 systems).

Back to Main Page

Email with your questions to sheila@atlanticaudio.com

This page is maintained by Laura Stover. © 1997,1998