This past weekend, I had the opportunity to teach a class on the above topic at NorthWest MinCon in the Seattle area. We had a a great group show up, and it was a lot of fun talking about the process of designing PAs. While I can’t condense the entire class into a single blog post, I thought it would be good to hit some of the highlights. Here are four design principles I employ when looking at a sound system design.
Even coverage throughout the seating area
We really want to see the same level and very similar frequency response in every seat in the house. Achieving that goal—perfectly—is nearly impossible. So we have to give ourselves a range. We typically want to shoot for a ±3 dB level in every seat in the house. That means the back of the house is likely to be no more than 6 dB quieter than the front row. And depending on your room and PA, it might be the sides, front corners or even center that is quieter. But the overall variation should not be more than 6 dB. With a good design and proper speaker selection, this is a very achievable goal. Often, we can get it down to ±2 dB.
Minimize overlapping sources
Another key design principle for me is to minimize overlapping sound sources. Again, this is easier with some designs than others, but as much as possible, I want one speaker to cover a seating area. The exception to this rule is a stereo PA design, but that’s a whole different kettle of fish, and I’m taking a pass on that one for now.
The issue with overlapping sources is that you will almost always get some kind of phase cancellation, which can really mess with our even coverage goals. Delay speakers are one situation where we will have some overlap, and it requires careful attention during the commissioning process to make sure they sound even, natural and are in time.
Keep sound in the seats and off the walls
This seem so simple, and yet, I’ve seen many, many designs where speakers are pointed at walls and not the seats. This is bad. In fact, in my last church, 8 of the 12 speakers were pointed at walls and not seats; as a result, 80% of the seating area was off-axis of the PA. The resultant sound was terrible. We had variations of 12-18 dB in our coverage, and the sound field in the seats was almost all reflected, not direct. Intelligibility was terrible and it simply sounded bad. No amount of EQ or DSP will fix that. Basic rule of thumb—point the speakers at seats, not walls.
More direct than reflected sound
A high direct to reflected sound ratio means the sound will be very clear. If you’ve ever been in a venue where it seemed like the speaker’s voice was right in front of your face, that system had a high direct to reflected ratio. The problem with high levels of reflected sound is that it diffuses the sound field, causes some nasty cancellations in the form of comb filtering and makes it really hard to hear what is said or sung. We can minimize reflected sound with a judicious use of treatment, and that should be part of the sound system design. But we also want to choose our speakers carefully, and position them properly to keep as much direct sound from hitting the walls—causing reflections—as possible.
There is of course more to the story, but we’ll let this serve as an introduction. I’ll touch on some additional speaker system design points in upcoming articles.
This post is brought to you by Nemosyn. The guitar player practices, the keyboard player practices, the vocalists practice, how does a sound guy practice? Nemosyn record. Practice. Perfect your mix. Visit their website at Nemosyn.com.
This will likely be a quick one. One of the most common problems I hear in mixes, especially with less experienced engineers, is that when the reverb gets turned up, the vocals get muddy. The solution is simple; high pass the reverb. Got it? Ok, we’re done.
Channel Strip High Pass
Well, not quite. I’m going to give you a few ways to high pass your reverb. Perhaps the easiest way is to simply turn up the high pass filter on the reverb return channel. Most of the time, we’re bringing the reverb in on a regular channel, and it has full processing. So go ahead and engage that HPF and dial it up. You also might find that taking a little out of the low-mids also helps. But we’ll come back to that in a moment.
Effects Level High Pass
Some reverb effects have a high pass filter (and even a low pass) built in. You can also use the function there. Why use it at the effect level instead of the channel strip? Well, the main reason I prefer to do it there is snapshot control. It’s not that I can’t snapshot the EQ of my return channel—that’s easy enough—it’s that I’m already recalling the effect setting with each snapshot. For me, it’s easier to recall the effects, and then recall only fader movements on my individual channels. It’s easier and quicker for me to scan down the recall settings and make sure I’m not recalling any EQ or anything on any of the channels than it is to make sure I’m grabbing EQ on the effects return channels. Again, it’s not that we can’t do it, it’s just easier to do the effects.
You may also have some different controls in the effect unit. High dampening is another one that is very useful for taming some of the high end sizzle that can make reverb stand out instead of blend in.
The question should be asked, how high to set the HPF? Well, probably higher than you think. I typically start with my high pass somewhere around 250-300 and go from there. It’s not uncommon for me to go to 400, especially when I’m layering effects. This is all by ear, of course, there is no set formula. However, what you’ll find is that the higher you run the HPF up, the more reverb you can use without it clouding the vocals. It seems counter-intuitive, but you can actually get more reverb by taking it away. At least the low end.
Give it a try this week. I bet your mixes will sound better.