Dimensional Focus: Horizontal Plane

Previously, we looked at a real-world example of a design which benefited from extra attention paid to the vertical (front to back) plane with simple horizontal plane coverage requirements. Now we will examine the opposite situation: a venue design in which the horizontal coverage will require a special focus.

Our venue has “Auditorium” in the name, but it’s a poor descriptor of its audience geometry: it’s a rectangle sunken inside a larger rectangle, the whole thing equal in scale to a small arena. If you imagine draining a large swimming pool and holding a concert inside it, you’ve got some idea of the situation (Figure 1). (“Swimming pool” is also a rather apt characterization of the reverberant properties of the space, much to the chagrin of the local residents who attend shows there.)

Figure 1 - This is not a skatepark

Mix position is at the rear center of the floor, immediately in front of the vertical wall. We are not scared of this wall, as it is acoustically treated, but even if it were threatening, our options for dealing with it are quite limited as we need to cover the balcony seats above, and mix position below.

These “wider than they are deep” venues tend to pose a unique challenge in terms of center gap. We have a low trim and a short throw compared to similar capacity small arenas which means we are likely to encounter a center gap, and with mix position being only about 66 feet from the stage, we need to make sure our mix engineer doesn’t end up in the light coverage off the end of a horn. Figure 2 shows the section view and splay strategy for 12-box main hangs (truncated from 16, which would over-narrow the low-mid dispersion due to the short throw distance), nothing to see here. Our concern is in the horizontal plane.

FIGURE 2 - Bog-standard mains splay strategy

The plan view prediction shows us that our concerns were justified - coverage at mix position is light in the HF range (Figure 3, left). Can’t have that!

Pointing the mains straight forward works great, until it doesn’t. “But we always do it that way.” I find convention a poor reason to continue doing something if there’s a compelling reason to depart from it. A 4° toe-in fixes the issue (Figure 3, right).

FIGURE 3 - Left - mains straight ahead gives light coverage at mix. Right - a 4° toe-in fixes the issue. (2.5 kHz)

Note that due to the venue’s aspect ratio, the mains don’t cover around the corners and onto the side seating as they would in a typical arena, so our side hangs have more work to do (and our mains toe-in just made it worse). This model definitely needs corners :) Note that more wide boxes would have been an alternative solution here, but as we’ll see now, they make themselves more useful at the top of the side hang.

Our side hang is also truncated from 12 boxes down to 8, as it doesn’t have nearly as much vertical coverage to take care of it. The side will mostly concentrate its energy on the upper seating level, and only supply a tiny bit of help on the edges of the floor level, so 8 boxes is appropriate and gives us a similar splay density (impact points per foot) as the mains. The default side hang configuration is 6 wide boxes over 6 longs, and we’ve scratched the bottom cart of 4, so we have 6 wides over 2 longs. The wides are tremendously helpful to reach up to the rear rows of the upper seating despite the short throw. (Figure 4)

FIGURE 4 - Left - Side hang only. Right - Mains + Side (2.5 kHz)

As an aside, the non-orthogonal angle of impact means we must overshoot a bit with our side hangs if we want to make sure we’re covering up to the last row for the entire coverage width (see Figure 3 in Saturday In the Park for why). The simple landing strip section view of splays has limitations, and this is one of them.

Now we’ve covered to the 180 line, but management has asked what would be needed to open the upper seating tiers all the way to the wall. The answer: more speakers! Now we’ve got a 270 hang, we just need to figure out where to put it. The promoter has advised us that the upstage wall in that upper seating area is hostile, and they would only like to open those seats if we can find a way to cover them without painting the wall with excess HF.

Whenever additional radial zones or hangs are added to a sound system, the overall direct:reverberant ratio can suffer because the percentage of loudspeakers in the rig that are covering any given listening position goes down, and the percentage of loudspeakers pointed somewhere else goes up. Such is the nature of increased ticket sales.

To address this as responsibly as possible, we will depart from the traditional “semi-coupled” 270 hang location, hung near the mains, subs and sides, and instead slide the hang upstage, so we can hit the additional seating section head-on, which minimizes energy impacting the upstage wall as compared to a more typical hang placement. We’ll also use all 80° cabinets to keep the horn pattern tight. The short throw actually helps us here. We can land the hang to keep all the needed seats in the “sweet spot” (say, within -3 dB points of the horn) and roll off outside of that. In general, a listener seated at the -6 dB point of a horn, without an adjacent source to pass the coverage baton, is experiencing a very different show than the listeners seated on axis, and I try to avoid it when possible. This is, like so many other aspects of this work, a balancing act. (Figure 5)

FIGURE 5  - Left: 270 hang. Right: Mains/Sides/270 (2.5 kHz)

And now: the subwoofers. The default configuration for this rig (7/side flown) presents two issues. One, the line is too long and there is too much narrowing for maximum uniformity, again due to the “stout” room dimensions. Removing one and flying 6/side gives us all of the floor within the same 3 dB color bar, and most of the upper level within the next 3 dB down.

Second, and more importantly, since the main hang is now shorter, it won’t curl under the subwoofer hang as usual but instead collide with it. We could space them out more, by moving the subs further upstage (not optimal, as it increases LF wash on deck) or push the mains further downstage over the audience (which requires more curvature and makes it harder to cover the first few rows, also not optimal). Solution: hang 6 subs instead of 7. We also buy ourselves a little extra wiggle room by increasing the trim height on the subs by a box.

Finally, our subs are cardioid, which actually starts working against us a bit in the 270 area. We can get those listeners a little more LF by toeing the subwoofers out. I usually avoid this technique in more standard arenas because it can put too much sub energy into the side hang area where the seats are already much closer to the arrays than the listeners in coverage of the mains (maybe the subs are -6 dB at 90° off-axis due to the cardioid pattern, but if the seats are 65 feet away instead of 180, that’s a 10 dB proximity gain and only 6 dB attenuation via the cardioid pattern, leaving the sides 4 dB hot with sub energy). In this case, however, the squat aspect ratio of the room means no significant change out front and on the sides, a bit less energy on stage, and a bit more energy at the mid point of the 270 hang coverage, so we’ll take it. (Figure 6, note the change in the yellow isobar).

FIGURE 6  - It ain't much, but it's honest work. 

Now we have to answer the important question: can we actually hang it this way? Prediction is a wonderful frictionless, ideal environment, free of the shackles (heh) of structural steel and beam locations. The original Main, Sub and 270 locations were chosen because they can all be accomplished by simple deadhang rigging points. Luckily in this circumstance we can accomplish our off-kilter mains and subs without adding more bridles to our tour rigger’s workload.

Usually I seek to accomplish this via choker hitches (wrapping the beam like a P instead of an upside-down U) so the steel drops to one side or the other of the beam. By converting the mains and subs points to on-off and off-on chokers respectively, the arrays will toe out of straight alignment. (Choker hitches have more limitations than a regular deadhang basket hitch, so they are not appropriate for all situations, so this call needs to be made by the qualified rigger on site.) In this case, the points were made to the upper steel, and dropped to alternating sides of the lower steel, giving me the toe I was looking for without the capacity de-rating or beam torsion effects of a choker hitch. Luckily our tour rigger is accommodating to my design eccentricates (either that or he’s a glutton for punishment) and our FOH engineer doesn’t seem to mind mixing on a PA that looks like it has been involved in an earthquake. (Subs toed out. Mains toed in. Science!)

The 270 location presents an interesting wiggle when it comes to alignment: due to its “decoupling” from the main hangs and placement closer to the target listening area, it will arrive at the seam much later than the side hang. This would mean delaying the side hang significantly, which in turn means delaying the mains and subs significantly. Does it make sense to push the entire PA noticeably back in time for all listeners (including our intrepid mix engineer) to smooth a seam for a few dozen seats? This is one of the choices we must make as systems engineers. In this case I decided the answer was “no” and after listening to the rig, our FOH engineer agreed. The seam ended up being out of time by about 7 ms and neither of us found it to be overtly objectionable when walking the system, so we left it as is. Once it’s aligned we are able to achieve an encouraging level of tonal consistency, which will be our safe harbor in the sea of reverberant conditions (Figure 7) (not all traces were taken with the subwoofers on, but no normalization was applied to them for this screenshot). The pink buildup at 60 Hz in the 270 hang was countered with EQ (it’s a different model of loudspeaker than the rest of the rig, and not cardioid).

FIGURE 7 - Trace overlays for the aligned system. No offsets applied. 

And maybe - just maybe - that might be enough to perk up the ears of locals with low expectations for the room.

The author would like to thank tour rigger Patrick “Lizard” Dashiell for his expertise and assistance with this deployment.