Removing Feedback from Stage Monitors
Just when I thought that club sound systems had evolved past Y2K, I get a gig at a place where the house monitor system is (I’ll be polite here) out-of-date. While the overwhelming trend for stage monitoring has moved heavily toward IEMs (even among local club acts), there are still plenty of performers using wedges. As wedges aren’t going away any time soon, let’s examine the concept of “ringing out” stage monitors.
Wait Just A Minute…
Before you try to electronically correct a feedback problem, be aware of the physical effects of microphone polar patterns and monitor placement on your ability to add gain before feedback. Omnidirectional mics are going to give you headaches (in the form of feedback and leakage) from the word “go,” so try to avoid using them live for vocals. When using cardioid and hypercardioid mics, check to see that the monitor speakers are physically placed in a position where the microphone’s pickup pattern rejects sound. Moving a monitor as little as eight inches can dramatically reduce the amount of feedback.
In the past, ringing out monitors required two people: an engineer to operate the monitor console/equalizer(s) and a person to stand on stage and make silly noises into the vocal microphones. Not so much anymore, since many digital consoles and outboard processors have the ability to be controlled remotely via iPhone, iPad or tablet PC. If you are mixing monitors from front of house and cannot run the console remotely, you’ll definitely need an assistant. It’s likely that there will be more than one stage mix, so the ringing out process has to be performed as many times as there are individual monitor mixes.
You’ll need a separate EQ for every mix, and it’d be handy to have an EQ for every wedge, in case one monitor mix feeds multiple wedges. Speaker components can age differently, so even though two wedges may contain the same drivers, they may not behave the same acoustically if (for example) one of the woofers was recently re-coned. A 31-band EQ is the traditional tool for stage monitors, but a multi-band parametric is equally effective.
Stand (or sit) at the performer’s position facing the audience with the mic and wedges in front of you. The house P.A. and other monitor mixes should be muted. Set the gain on the input channel of the monitor desk or front of house desk if you are running monitors from the house mix position. If there is a compressor on the aux send output from the desk, bypass it for now. Start with all equalizer and gain settings at the “0” point, and then raise the microphone level in the monitor mix while speaking into the microphone (“This is mic number one, isn’t this a lot of fun…”). As the wedge becomes louder, it will start to feed back or “ring.” As soon as you hear a bit of ringing, stop increasing the level on the monitor send, but do not turn it down. We want the monitor just on the edge of feedback. At this point, it’s probably a good idea to not even touch the microphone. Don’t even look at it. If you make the mistake of cupping the mic in your hands, feedback will ensue.
Here is where the graphic EQ comes in handy: we are going to use it to seek and destroy the offending frequency. One EQ band at a time, slowly raise the boost/cut controls above 0 dB until you hear the feedback start to increase. Note the word slowly in that last sentence. If you simply start whacking up the boost/cut controls, you will induce nasty feedback that could harm the gear as well as your ears.
Typically, there’s more than one frequency that rings, so it may take some trial and error using the EQ to zero in on the fix. Experienced engineers can identify the offending frequency more easily, and I promise you’ll get better after you do this a few thousand times. A third-octave or (better yet) sixth-octave spectrum analyzer can help you identify the feedback frequency until you have enough experience to recognize them.
There will be times when a 31-band graphic EQ is too blunt an instrument to solve your problems — like when the feedback frequency falls exactly in between two “standard” frequencies on the 31-band graphic. That’s where the parametric EQ or notch filter comes in handy and you can use a somewhat different approach.
Set one of the mid-frequency bands to a narrow-to-medium bandwidth (“Q”). Don’t go too narrow, or it will become difficult to find the ringing frequency. Increase the gain by 4 or 5 dB. Slowly sweep the frequency control until you hear the ring increase, then stop. Make the bandwidth more narrow, bring the gain up another few dB, then sweep again, but don’t sweep as far across the frequency range as the first time — stay in the range where you first heard the ringing. This helps increase the accuracy of your frequency selection so that your EQ affects the timbre of the audio as little as possible. When you find the frequency that causes the monitor to ring, use the gain control to cut that frequency. I suggest you cut only as much as you need to defeat the ringing. A notch filter can be swept in a similar manner, except that “true” notch filters won’t allow you to boost.
A-E-I-O-U
The person speaking into the microphone should vocalize different consonant and vowel sounds. A good place to start would be with “t’s” and “s’s” which can cause high frequency ringing, but don’t ignore long “a” and “e” sounds — they can be sources of feedback in the midrange and low-middle frequencies. If the performer wears a hat or theatrical headgear during the show, you’ll need to do the same, because they affect feedback.
Once the ringing has been reduced or eliminated, raise the level of the monitor. Chances are high that another frequency will start to ring, so repeat the process. Sometimes reversing the mic’s polarity (“phase”) can reduce ringing. When you get to the drum wedges, keep them tight by cutting the area between 250 and 350 Hz on the kick drum, and try the polarity switch if the toms are ringing.
One thing that helps control wedge feedback is using EQ and filters to reduce the overall bandwidth of the signal being reproduced. There’s no point in making an amplifier try to reproduce 40 Hz for a small wedge with a 12-inch woofer, so often you can pull down the lowest two or three bands on the graphic EQ. Ditto for the very high frequencies: no one is going to hear 20 kHz in a live situation. This increases the efficiency of the amp and gives you more clarity in the mixes.
Once you’re happy with the overall level of the wedge, you can add a bit of compression to keep the drivers safe from harm. Use a low ratio and a fairly high threshold so that you are not constantly compressing the mix, because doing so will cause more feedback when the performer stops singing and the compressor “lets go” of the gain reduction.
Unfortunately, the ringing out process is one of diminishing returns. As gain of the monitor is raised higher and higher, the likelihood of feedback will increase, regardless of how much EQ is applied to the signal. The idea here is to be able to run the wedges at very high SPL’s, but not actually do so. Just because your Ferrari can do 200 MPH doesn’t mean you should drive that fast down a city street — so don’t forget to turn the mix down after you ring it out. Hopefully, the SPL where feedback becomes unavoidable will be far higher than the stage performer’s desired level. If not, you’re in big trouble — but that’s another topic…
Steve “Woody” La Cerra is the tour manager and front of house engineer for Blue Öyster Cult.
