Once a sound system has to project to audiences greater than a few hundred feet from the stage, delay banks become an important tool. These are speaker clusters that are positioned in the middle of the audience and pointed towards the crowd in the back. Delay banks are used for several reasons:
- They allow you to keep the sound levels directly in front of the stage at a reasonable volume and not deafen those seated down in front while getting the need volume to the audience far from the stage
- High frequencies are attenuated by the distance of the air itself, so that at several hundred feet from the speaker banks there will be a noticeable loss of brightness
- There may be obstacles between the speakers and the audience, putting some of the listeners in an acoustic shadow.
If you look at the diagram, you’ll notice that there are two speakers that the audience in the back can hear. The speakers near the stage are 150 feet away from their ears, while those in the delay bank are only 50 feet away. This produces a distance of 100 feet between the sound paths of the two speaker banks. Since sound travels about 1,100 ft per second, while electrical signals travel at something like 1 billion ft per second, there will be about a 90 millisecond (ms) delay on the sound as it travels through the air as compared to its travel on the wire. If no signal correction is added, this 90 ms difference will sound like a very distinct echo, since you would hear the close speaker first, followed 90 ms later by the sound from the main speakers. This is distracting at best, and can cause a complete loss on intelligibility of the spoken word.
WHAT’S A SIGNAL TO DO?
In order to get the sound from both speakers to arrive at the same time to the listener’s ears, we need to add enough delay to the remote speaker so that it waits for the sound from the stage to catch up. Back in the prehistoric days of audio (before digital processing) various devices such as tape-delay loops, and longs lengths of coiled plastic hose fitted with horn drivers and microphones were used. These devices were hard to adjust, required maintenance, or just plain sounded bad. With the proliferation of cheap digital storage, keeping a few hundred milliseconds worth of sound in storage is easy. The analogue sound is digitized, then sent into storage for a few hundred milliseconds, were it’s then reconstructed into the original analogue format. This is enough time for the sound in the air to “catch up” with it, and the echo is eliminated.
There are a few consideration when selecting and adjusting delay stacks. First of all, pick a digital delay that’s designed to do the job. Yes, you can use an SPX90 in a pinch, but most “digital reverbs” don’t have the bandwidth and noise floor to pass hi-fidelity signal without degrading the sound. I have a Klark-Teknik DN-700 with provides up to 700 milliseconds of delay out 3 separate “taps”. This allows you to feed up to 3 speaker banks, each with it’s own separate amount of delay. Some of the newer BSS units have a temperature probe that automatically changes the delay depending on the air temperature. This allows you to set the delay in the morning and still have it be right in the heat of the afternoon. The reason for temperature affecting the delay is that as the temperature rises, the air gets less dense. As you may remember from high school physics, the speed of sound of various media is dependent on density. So as air temperature rises, the speed of sound slows down. This doesn’t seem like a big deal, but at a few hundred feet from the stage, the needed delay can vary by 10 milliseconds or so from the morning to the afternoon. Now, 10 milliseconds is very noticeable to many listeners. In fact many of my engineering buddies and I regularly tweak delay banks down to the 1 ms level (and sometimes argue over fractions of a millisecond if in the mood).
There are some pretty complicated tools that allow you to set the proper delays, but these tend to be rather expensive and thus relegated to major tours and built-in sound system installations. For most engineers, delays will be set by ear. Here’s the basics. If you have any additional tips, please email them to me. I like to start with a simple click or beep sound. If you don’t have a signal source, then clicking your tongue into a microphone with a long cable will work. First, position yourself very close to the delay speaker bank and a little to the side. This allows your one ear to hear the main speaker, and the other to listen to the delay bank. Now with the sound volume turned down in the delay bank, send your pulse sound to the main bank. It doesn’t have to be very loud, maybe 6 dB above the normal room clatter is sufficient. Now gradually bring up the volume in the delay bank until the volume level approximately matches the sound from the stage. At this point you’ll hear two distinct clicks, with your remote stack happening first and the sound from the stage next. Now start adjusting the delay until the two separate clicks merge into a single sound. If you go too far you may need to back up and start over. Also, you may need to raise or lower the sound level of the remote bank in order to hear the effect. Now, walk directly behind the remote stack and see if the echo is still gone. If not, then add or subtract a few milliseconds from the digital delay and listen again.
Now put on some real music with percussion, the road crew favorite usually being Steely Dan. I then like to “walk through” the delays from the stage on out, listening as I’m walking for any echo zones. This is where you can adjust the actual volume of the delay banks. If they’re too loud, the sound will “wrap-around” the back of the cabinet and cause interference with listeners closer to the stage. The volume should be just loud enough that it’s not really noticeable. Finally, put on some program of the actual music style if possible. Various music styles will need different adjustments techniques. For instance, classical music has lots of string swells and legato playing which can disguise echo problems, but ends up showing up as a fake “phasey” stereo quality. During the live performance, you can tweak the delay for “mono” sound to get it right.
Dealing with delay banks can be tricky sometimes, but they’re often the only way to cover a large crowd with any kind of control. But it really makes you listen and that’s why you started sound engineering in the first place, right?
Copyright Mike Sokol 1997/2016 All Rights Reserved
As Published in TV Tech Magazine August ’97