Year of the System 2010: PA Anatomy – The Main Event
Now that Drive is over and we’re starting to settle in on the tuning of the system, I’m going to try to get this going again. So here it is at last, the Mains.
The heart of our system is comprised of two arrays of Meyer Sound MICA’s with 7 boxes per side. The choice of MICA’s dates back to 2005 when they were demoed for the 2005 Drive Conference. The experiences of the staff at the conference have become legend over the last four years, and now that I have heard the system in our room I understand why. I was in all honesty genuinely surprised by how great the boxes sounded, again, simply with an iPod daisy-chained to each box in the array. Through the optimization process, the coverage smoothed out even more. It is a huge win for me to be able to stand in the back corner of the room as far away from the stage as you can get and to still hear things clear as day. Now even if you arrive late and end up standing against the wall, you can listen to clean and clear spoken word that makes up the bulk of our services.
So why line arrays? Line arrays give us wide coverage for our very wide room along with very consistent coverage to the back of the room. One of the greatest challenges of our previous system was definitely room coverage; as you moved to the far outside sections of the room, the coverage dropped off. I think our previous system did very good for its time, but it was 12 years old and did not meet the programming intentions or performance expectations of a modern PA.
Part of the coverage challenge in our room related simply to distance. Since the room gets wide, the far outside corners needed the main loudspeakers to throw much farther to reach them than the center area. The Inverse Square Law dictates that whenever you double your distance you lose 6 dB. So if the PA measures 96 dB at 50 feet from the speaker, it will measure 90 dB at 100 feet. This creates a couple of interesting issues. For starters, we’re losing about 1/3 of our loudness with every doubling of distance. While this might not be an issue for the folks that don’t like the rock music, they probably have a different opinion when it comes to spoken word. Another issue relates to our friend, the Equal Loudness Contours.
The Equal Loudness Contours demonstrate that we do not perceive all frequencies the same as loudness increases or decreases. It’s almost as if there is an internal EQ in our head that adjusts as the volume changes. If we mix at one level or within a consistent dynamic range, the contours don’t really matter to our perception. However, if we mix at one level and our mix is listened to at varying levels, the tonal perception of the mix will be perceived differently as the listeners move farther away from the mix position and their listening level changes. A mix position in the middle of the room might be the best of both worlds in a situation like this, however, think about how many times mix position is put in the farthest possible location. As loudness increases so does our sensitivity to higher frequencies especially in the hurt range around 4k. In this type of situation an even mix in the back of the room might have the same spectral balance up front, however it could be perceived as much brighter where the loudspeakers are closer and louderto the listeners. Conversely, the farther seats–such as our far corners–could sound duller as the level drops.
Line array loudspeakers are bound to the same Inverse Square Law just like any other loudspeaker, however, the way the boxes interact within the array counters some of this and manipulates physics for fun and profit. Basically, the array uses destructive interference to reduce SPL in the nearfield. As distance from the array increases, the array trades destructive interference for constructive interference creating a more consistent SPL level front to back than we can achieve with a traditional loudspeaker system. As the listener moves beyond the point of maximum constructive interference, the line array begins to act like a single loudspeaker and Inverse Square Law returns. However, the bottom line is in our room our new MICA arrays give us much more consistent front to back coverage than we previously had.
And then there’s quite simply the sound. The MICA’s have a sound that we like that I would compare to hi-fi studio reference monitors; you can really hear every nuance, for better or worse, in your sound sources. While this is somewhat characteristic of just about every Meyer Sound loudspeaker I’ve heard, by utilizing line arrays we get a very even response throughout the coverage of the speakers. We also get a nearfield effect throughout the coverage; basically you can stand in the back corner of the room and it sounds like the person talking on stage is right in front of you which is great for the more conversational style of teaching we tend to feature on our stages.
There is still one installment left in PA Anatomy, and I’ll try and get to that sooner than this one.
Previous Post
Next Post
You bring up a very interesting point in terms of the sound booth placement. The problem I’ve noticed (not just our church, but visiting many other churches) is that the sound booth is either back in some corner where it is difficult to mix. Or, like many modernly designed mega-churches, it’s smack in the middle of prime real-estate seating. And usually taking up the equivalent to 40 chairs. Surely there’s a better way?
Watching TV I had a thought. Why not just put the sound booth in some isolation room and mix on studio monitors? Assuming the acoustics of both the large room and the mix room have been set up properly, what you hear from the studio monitors will sound very close to what comes out of the line arrays. So, in essence you are hearing what those sitting in the audience are.
Besides, using Virtual Soundcheck to mix your rehearsal in a studio for Sunday morning is not too different, eh?
Good questions, JB.
Mix position is a touchy subject for just about everybody. In a modern setup, there is no reason why FOH should take up 40 chairs. A huge sell in the touring world in the early-early days of digital consoles was the smaller footprint provided which allowed promoters to sell more seats. For most Church world administrators, there is a cost associated with every seat in the room and more seats are generally a good thing which is why the location of the gear often takes a backseat.
I look at it this way, I believe that the mix position should ideally be in the location that gives the mix engineer the best approximation of the listening situation in the rest of the room. Typically this is in the center of the room maybe 1/2 to 2/3 back. I want to be in the center because 9 out of 10 times it’s the one location where every frequency adds together so the mixer will hurt when anyone else hurts. The minute you start moving off-center, your propagation times between loudspeakers becomes asymmetrical and induces comb-filtering so now your mix engineer might not hear something someone else will; this could be a painful frequency or it could be feedback. When you look at the big shows with the big name big chair mixers, they sit in the middle of the room for a reason. When you compromise your mix position, you compromise your mix; this can and will happen, but it’s just important that everyone is aware of what can happen when you do do that before you make a decision.
There are a couple issues that come with putting the mix position in another room. For starters, you have a completely different acoustic space. You can work all you want at making them similar, but depending on the acoustics and size of your room, sound coming from the stage will play a part in the listener’s perception, and I believe the guy mixing for them should be able to account for that. In a bigger room you could probably get away more with a “studio” mix location from an acoustics standpoint, but most churches aren’t big enough to where they can get away with it.
Then there’s the whole challenge of equal loudness. The dynamic range in our services is about 30-40 dB in the room ranging from the mid- to upper-90’s for music down to the low-60’s to low-70’s for spoken word. I don’t know about anybody else, but I could never sit in a studio or backroom with a couple of nearfields in front of me blaring at 97 dB; that would tear my face off. I probably spend more time in the mid- to upper-70’s on an A scale meter if you measured in the studio. So right away we have a problem thanks to our friend Equal Loudness because there’s now a 20 dB difference between the mix reference level and listening level for music which is going to create tonal perception differences. Even if we can counter that, there is still a challenge in the overall dynamic range.
We’ve found after years of having our studio mixing on Sunday, that there is a natural tendency to mix everything in the studio at the same relative loudness level. However, if we mix everything at the same level in the studio and feed it out to the room, now our music will be at 97 dB along with our spoken word and video playback, and that doesn’t work. I can’t explain why 30-40 dB of dynamic range in our room works and feels natural, but it does.
Now, in reality someone who is motivated enough could probably figure out a way to get around all this stuff and mix in another room. Buford Jones did it on a Faith Hill tour that was in arenas I believe, and if memory serves I think he did it just to prove that it can be done. But there is still one factor that you lose if you have a guy sitting in an isolated room, and that’s the audience interaction. Every engineer I respect mixes off the audience, and the only way to feel that out properly is to sit in the room with them.
As far as Virtual Soundcheck goes, we never do it in a studio; it is always done on the PA at performance levels. There are things we lose like cymbals spilling off the stage, but I find it is very easy to make any necessary modifications when the band is back on the stage. Plus after mixing in our rooms for 3 1/2 years, we more often than not naturally account for this when we’re using virtual soundcheck.
Dave — I noticed the new line arrays from the very first moment of Drive 2010. They sound awesome!
I think one of the key benefits of line arrays is their precise coverage. The ones we use have a 10-degree vertical spread, meaning we can aim them very directly at the crowd and not have a lot of energy wasted on walls or floors or ceiling (creating all kinds of acoustic nightmares). Plus, with more boxes aimed at the back and fewer at the front, you can really get very uniform loudness. We even run the ones pointing to the front rows with “shading” enabled which basically reduces the high output by 3 dB to compensate for the effect you mentioned in your post. Line arrays rock!
Hi Dave,
That was a really good post on the line arrays. I’m in Australia, and we are nearly at the stage of construction of our church building (literally months away). We are definitely going to use Meyer’s, but our consultant has recently suggested that Arrayable speakers might be better for our auditorium. Based on pictures of the NP auditorium, I estimate ours to be half the width (so not as wide), but do have the arch design.
Did you guys ever consider arrayable speakers? Basically the idea here (for our auditorium) is that arrayable speakers give a more uniform spread of the sound waves (in the horizontal plane), as opposed to ‘spikes’ at different (higher) frequencies using line arrays.
The use of delayed speakers for NP’s balcony is interesting, given that the line array should account for that space. So assuming you did not have any obstructions for the upper speakers to throw to the balcony, would you have still gone with delayed speakers?
Cheers,
Tim
Tim, I’m going to assume by arrayable speakers, you’re referring to horizontally arrayed speakers sometimes referred to as trap[ezoidal] boxes. I did consider a more traditional system using horizontal clusters of speakers. We looked at both JM-1P’s and MSL-4’s over the last couple years. There are pro’s and con’s to both approaches. In a traditional horizontal array setup, it can actually be harder to get a uniform horizontal spread due to the pattern overlap between boxes in a cluster. Things also get complicated if you have a cluster that is flown, and you start tilting or aiming it down because now you’re actually starting to rotate some of your horns in the vertical plane which will alter their coverage. This isn’t something that should be cause for alarm, though. Horizontal/traditional/distributed/whatever systems DO work; I think it’s just important to understand the challenges that come with any type of system because line arrays are no magic bullet, either.
I think Meyer’s trap boxes(MSL-4’s, UPQ’s, JM-1P’s, etc.) do a great job when properly arrayed horizontally, however not all boxes perform the same. Some manufacturers might leave coverage gaps in the seam between boxes, and others have a bit of destructive interference in the overlap. It’s just very hard to maintain coherency where the patterns of clustered boxes overlap. In a perfect world, we would have trap boxes with coverage patterns that run right down a perfect line so that each listening position hears one speaker, but that probably isn’t going to happen so manufacturers make every effort to make that overlap as transparent as possible, and some do a better job than others.
Something to keep in mind that I hear from touring system techs is that the more splay you put in the array, the more your coverage breaks up creating those spikes or beams. Flatter vertical/line arrays work better. It’s important to understand that modern line arrays behave as a line source in the low frequencies, but as you reach the upper-mids and high frequencies the boxes operate more as point-source boxes. However, the difference between line arrays and traditional trap setups is that when you do have pattern overlap in the upper frequencies using a modern line array, the wavefront remains coherent minimizing destructive interference. Line arrays are really an interesting animal because adjustments made in one part of the array–even at very high frequencies that in theory shouldn’t overlap with other boxes–have consequences throughout the coverage pattern. Multiple measurement mics come in very handy when tuning line arrays to get a wide perspective of what’s happening within the coverage of the entire array.
It’s hard for me to speculate what we would have done if there weren’t obstructions between the arrays and the balcony. Based on what I saw modeled, I probably still would have advised against line arrays for the balcony, BUT it’s something that I would have liked to actually hear before making a decision. However, when we flew the first array, we had an 8th box in it(sightlines required we drop the 8th–another model vs. real world situation), and the coverage was definitely smoother with 8. I think 7 boxes covering the floor sound great, but I do know that 8 was better. I would be nervous, though, if we had taken 2 of those 7 and pointed them at the balcony so that the floor was only covered by 5. I think it would have worked, but over time I don’t think we would have liked it.