The other day, I was asked by a local church to take a look at its sound system and see if I could advise them how to get the final 10% of performance out of it. They felt that, although it was good, it could be better. They were right…well, at least the bit about “it could be better!”
I popped in one evening to take a look and to see if I could do anything to help. Having concluded that I could but they couldn’t afford me to, I arranged to come back the next week with some test gear to give it the once over. Just as I was leaving, the band began rehearsing. I hung out in the doorway for a couple of minutes to see what they sounded like, what instruments were being used and how it was being mixed. The lack of vocal clarity was remarkable: “Wow, that takes skill,” I thought…or perhaps it was the system. I decided that a dose of pink noise was required and I brought some with me the following week, along with one of those annoying sine sweep things that we acousticians like to use!
Now, it is always interesting deciding where to start when doing one of these assessments. For example, should I let rip with the pink noise, the log sine sweeps, listen to some vocals or let full range program reign? I opted for none of these, but decided to put out a band of high frequency, bandpass filtered pink noise because I wanted to see what the coverage was like, particularly at the sides of the room. (A quick eyeballing of the left and right speaker clusters suggested that the high frequencies might not be getting to these regions.)
Before starting, however, I asked the operator if he would bypass all the desk EQ so I knew I would have a level playing field from which I could make my assessment…or so I thought! It was one of those new-fangled digital mixers that had more layers than MPEG4, so I let the expert do it. I then unmuted the input (I could handle that) and set to work, but not before suggesting that the accompanying entourage might like to escape to the relative calm of the room next door and perhaps put the kettle on.
I was interested to find, but not altogether surprised to see, that the entire left-hand seating area was lacking high-frequency sound (good to know that the old adage of “if you can’t see the loudspeaker, then you won’t hear it” still worked). What was nice was that, by listening to the high-frequency signal and measuring the SPL at a given seat, I was able to nicely correlate where the high-frequency sound was dropping off.
The right-hand side of the auditorium suffered from the same fate, although to a lesser degree. What was interesting, and immediately obvious to the ear, was that the right-hand side fills were much louder than the left. (Hearing protection would have been a boon for the listeners on the right-hand side). Grabbing the analyzer showed the difference between the right and left sides to be 10dB.
I then graduated to pink noise and did a general walk around. Now, I have listened to an awful lot of pink noise over the years, and what I was hearing was decidedly not pink! A quick look at the real-time analyzer display (yes, I still use one of those when appropriate) showed that my ears were not deceiving me and that there was a distinct notch in the response occurring at around 1kHz to 3kHz. I finished off this part of the assessment with some sine sweeps and balloon bursts so I could measure the reverberation time characteristics of the room. It wasn’t particularly lively (well, not by my normal standards), clocking in at just under 0.70 seconds. Time for a cup of tea and a chat.
The first topic of conversation was the 10dB discrepancy between the left and right side fills. No one had actually noticed this (apart from those sitting in the target area, perhaps), but then the mix position is just off center to the right. I then tactfully (yes, I can be sometimes!) asked the operator if he had, indeed, bypassed or flattened all the desk equalizers. I received the same affirmative assurance. “Oh,” I said, “then there must be another culprit lurking in the system.” We ventured back into the auditorium and I decided to check the output of the desk directly at the console. Flat the response was not, as can be seen by the plot “exhibit A” that can be found in the accompanying chart.
As you can see, there was, indeed, a large notch, centered at 1kHz, that was affecting more than four octaves of the audio spectrum and sucking out more than 6dB of wholesome audio in the process. OK, so now I modified my question to the sound system operator: “Do you have any effects still operating?” After about 10 minutes of multiple screen searching and menu hunting, I heard a significant “Ah” come from a location near the center of the console.
Now, this was no ordinary run of the mill “Ah,” but one with a deeper meaning. Indeed, it signified that the cause of the audio mineshaft might have been found. And, indeed, it had, for suddenly the sound of virgin pink noise with nascent purity could be heard. (Actually, I exaggerate here, but the new sound was decidedly more pink than the previous version.)
“What did you do?” I enquired. “Bypassed the ‘acoustic enhancer’.” came the reply. Seemingly, buried deep within the ones and zeros, there was a Trojan Horse whose sole purpose appeared to be to destroy the audio signal quality and vocal clarity. The finesse of the beast was, of course, to call itself an “enhancer” rather than a “destroyer.”
With the beast safely caged and guarded by two FIR sentry filters, I re-measured the system response in the main part of the auditorium. It was now approaching flat (or at least the bit that should have been), and amenable to some gentle tweaking. Surprise, surprise, the vocals suddenly had some clarity. So, now, the next time I encounter a digital mixer, I will not only make sure that the desk EQ is set flat but also ensure that no other effects and settings are still lurking to trap the unwary.
I have rarely had to check the frequency response of a mixer when carrying out a sound system check before, but I certainly will be adding this to my mental checklist in the future. Now there’s something to reflect on.