Speed of Sound is approximately 1130 feet per second in air

(the actual speed is affected by the temperature)

Therefore if you have an 1130 Hz tone, it will complete one full cycle in one foot.

Now suppose that you have a tone generator feeding a speaker,
with two microphones as shown below.

A) Distance affects the Phase

1130 Hz

If the microphones are both the same distance from speaker, they would be In Phase and would add together. The resulting tone would be twice the level (6 db) of either tone.

1130 Hz

Similarly, if the second microphone was one foot further away from the speaker, the two sources would still be In Phase and would again add together.

1130 Hz

If the second microphone was only six inches further away from the speaker than the first microphone, the two sources would now be Out of Phase. This would cause the tones to cancel.

For the next section return to the previous setting
(ie. microphone #2 is one foot further from the source than microphone #1).

B) Frequency affects the Phase

565 Hz

At 565 hz (1130 hz / 2) the tone will now complete a full cycle in two feet. As seen in the following example, the two tones now arrive Out of Phase and thus cancel.

1695 Hz

At 1695 hz (1130 hz x 1.5) the two tones also arrive Out of Phase and cancel.

2260 Hz

However at 2260 hz (1130 hz x 2) the two tones arrive In Phase and thus add.

This effect (Comb Filtering ) can be shown to repeat all the way up the frequency band. The following graph shows the resultant Gain verses frequency. Note that when the two signals are equal, if they are exactly in phase they add 6 db but if they are exactly out of phase, they totally cancel

In an actual situation, the effects would probably not be as pronounced, since the levels from the two microphones would seldom be exactly equal.

One good example where this comes into play is when two microphones are (mistakenly) placed on each side of the lecturn, with the idea that they will pick up the audio regardless of which way the speaker turns. This will result in poor sound quality. As the speaker turns his head, one mic can be closer that the other, thus introducing the comb filtering.

3:1 Rule

The 3 to 1 rule will effectively minimize the Comb Filtering. This rule states that the distance between adjacent microphones should be at least 3 times the distance from the microphones to the sound source.

Because Microphone B’s level is approx 10 db down, the amount that it will add or cancel is insignificant and thus hardly noticeable.

The Cardiod Microphone

A Cardiod Microphone is more sensitive toward the front. As the frequency decreases, it is less directional

Effects with a full Choir

A similar effect (Comb Filtering) will occur with one microphone receiving the same sound from two sources.

A common example of this is shown below.

If the microphone had been closer, the difference in the direct path and the reflected path would have been greater and thus the reflected path’s reduced level would have had less effect. Also the reflected source volume would have been less if the floor had been carpet.

Methods of correction:

1. Keep the vocal audio mix low into the monitor
2. Hand hold or place the microphone closer to the singer

While the monitor helps the singer, as the monitor’s gain is increased, the resulting vocal will be more muffled. Although not popular with the performers, using music only on the monitors (without vocal) will also minimize Comb Filtering.

Often, when trying to improve the monitoring for the performers, the house audio suffers.

InEar Monitoring can eliminate this situation

In-Ear Monitoring

This article was prompted after I attended several concerts in which the music was excellent, however the dialogs were difficult to understand. Most of the production crews knew the script so well, that they were unaware of the problems.

If you asked the audience, they would probably say that they thoroughly enjoyed the music. If however, you were more specific and asked them about the script, they probably would be unable to answer.

The "Comb Effect" of excessive use of stage monitoring would mush the dialog, so that the audience would have difficulty following the script.

If the concerts are trying to tell a story, with excessive Comb Filtering, they often will miss the goal and end up providing enjoyable music only.

Summation:

Ideas to reduce the undesired effect of Comb Filtering

1. Reduce the number of paths from the same audio source
   • Fewer microphones
   • Reduce possibility of reflections
2. Reduce the relative amplitude of the additional paths
   • Increase the difference in path lengths
     (thus the secondary path will have more attenuation)
   • use absorbent material
   • use directional qualities of the microphones

The following sites assisted in this article

• Calculations of attenuation over distance ... www.mcsquared.com/dbframe.htm
• Calculations of distances ... www.pagetutor.com/trigcalc/trig.html

This article is also published in "Professional Sound"

( part 1 - Dec 2006,   part 2 - Feb 2007 )     www.professional-sound.com

26-Dec-06

Number of visitors

Return to top

Tutorials: Home  Before Audio Feedback  Effects of Audio Phasing  Amplifier Gain
 70/25 Volt Audio  Dimming Stage Lamps  In-Ear Monitoring