The Space Between

Flying subwoofers directly behind a flown full-range array is now a well-established approach in large-scale live sound reinforcement. Some people, when learning of the technique, enquire as to whether any of the output from the subwoofers is blocked by the mains. The typical response is that the subwoofers emit low frequencies with long wavelengths (50 Hz = 22.5 ft / 6.8m) that can simply bend around the main hang and continue on its way, rather than being occluded.

I realized during a recent iteration of this explanation that I had never tested it myself, and as far as I could remember, had never seen any data in support of this explanation. Is this one of those bits of audio “common knowledge” that has been accepted as true but has a more complicated reality behind it? Is the presence of the main hang truly completely transparent to the subwoofer array? Maybe there is some sort of attenuation effect, most likely towards the upper end of the subwoofer range, but it’s a good example of the Magnitude Fallacy in that it’s not worth worrying about? Or maybe there’s a more significant occlusion that we’ve been overlooking. Of course, it is scientifically sound to state that a 20 ft wavelength will not be deterred by a 3’ wide line array in front of it, and our ears seem to confirm this, but it wouldn’t be the first time a seemingly valid idea isn’t borne out by data.

Once the question had occurred to me, the only thing to do was to test it for myself. This presents some logistical challenges as you need to be in an environment where you can lower the main hang once the PA has been flown, measure the subs unoccluded, and then raise the mains again, all without disturbing anything else on stage that might be placed beneath the PA. Also, since the main PA looms are typically draped over the top of the sub array in such a configuration, bringing just the mains back down from trim would cause a cabling crisis. A recent show afforded the opportunity to get this data, and my PA tech graciously agreed to run the mains looms alongside the sub array and independently to the cable pick, rather than over the top of the subs, to facilitate me bringing the mains in and out independently.

The test is a simple relative comparison: measure the response of the subwoofers with, and without, the mains blocking them. Then compare the traces. If the presence of the main array causes any level loss, we’ll see it in the measurement.

Figure 1 shows the rig as deployed. The downstage rigging points for mains and flown subs are separated by 6 feet. The KSL cabinets are 24 inches deep, the subs are tipped slightly upward and the main hang has curvature so the physical space between the rear of the main hang and the front of the sub hang ends up being about 3.5 to 4 feet which is fairly typical for my designs.

FIGURE 1 - The system deployment as measured for this test. Special appearance by the downstage lighting truss cable pick.

I chose two mic locations out in the venue, directly on-axis with the right Main/Sub hang, approximately 200 feet (green) and 150 feet (pink) away (Figure 2). I also took a measurement using the mic located on venue center at mix position about 110 feet away (orange). Mix position is visible as the grey tent structure in the photos below.

FIGURE 2 - Measurement locations at approximately 200 feet (green) and 150 feet (pink) from the PA. I took an additional measurement from mix position located in the grey tent at left (orange). 

Measuring the right flown sub array from these three positions - with the main hang in place in front of the sub array, but muted - yields the transfer function data shown in Figure 3.

FIGURE 3 - Sub array response with mains occluding them. 

In all three cases I have added some excess measurement delay to bring the phase traces into more readable orientations, for reasons that will soon become clear. The phase deviation from the FOH trace above 63 Hz is not unexpected - remember this is a cardioid subwoofer array, so we are bound to see the effects of the pattern coming into effect as we move off axis. Such is outside of our focus here - we are interested in how these responses will (or perhaps will not) change when the main hang is lowered out of the way (Figure 4) and the sub array is measured again from the same locations.

Figure 4 - the main array is lowered out of the way so the subs can be re-measured without the occlusion

I’ve overlaid the “with mains” and “without mains” traces as a direct comparison to see the effects of the presence of the main array in front of the subs. Here I’ve split the traces into three separate plots for clarity (Figure 5). The original (mains in place) traces retain their original colors, while the lighter colored traces in each plot are the measurements taken without the mains in the way.

Figure 5 - response of the flown sub array with and without the mains hang in front of the array, from three audience locations

So what can we learn? The magnitude traces tell us that the conventional wisdom about the energy simply flowing around the mains hang, rather than being blocked or attenuated, seems to be accurate, as the measurement reveals no significant change in level between the two traces. Where we do observe a measurable change is in the phase trace. Once the obstruction of the main hang went away, the sub energy showed up just a tad bit earlier at all three measurement positions, as indicated by all phase traces. This makes sense - the energy takes a direct route to the listening position instead of having to take a longer path bending around the main array.

This is the same reason why using simple point-to-point distance measurement doesn’t quite work for timing on cardioid sub arrays - the actually box-to-box arrival times are a bit longer due to the wraparound paths. In this case, the difference between the two scenarios was equivalent to 1.06 ms of delay, the addition of which re-aligns the sets of phase traces. Of course the exact value is of only limited academic interest and likely to be highly situation dependent, but the concept here is the valuable bit - empirical verification of the conventional wisdom that we can safely fly our subs behind our mains without loss of output.