What System Optimization Means To Me

What System Optimization Means To Me

A colleague of mine owns a production company and had a gig yesterday for a government accounting agency. The setup is simple: a podium mic and a couple of RF handhelds that everybody needs to be able to hear and understand. The trick is that the event is inside a literal aircraft hangar, probably the most acoustically hostile environment I have encountered.

So this is a busy time of year for those guys and their crew is spread a little thin. The guys working this event are hard workers and good dudes but they’re not audio techs and they’re a bit green. I keep getting calls asking about relatively basic stuff like how to turn on phantom power on an X32. Everything is feeding back, and tensions are rising, and after the fifth call (asking me to walk them through ringing out a podium mic over the phone…) I got the TD on the phone and said, “hey, do you need me to come up there?” I had another event, so I had about 30 minutes to fix things up.

The basic setup looks like this:

I tend to work backwards through the signal flow, outputs before inputs. No sense ringing out a mic if the PA isn’t working correctly. So the first thing I do is jog around the perimeter of the room and deal with the loudspeakers. A few problems:

  • Each box (they’re active) has a different gain setting on the amp, simply because nobody checked. I set them all to a reasonable nominal level for a speech program.

  • The aims are goofy. Using the front pair of boxes to try to get all the way to the back means high SPL at the podium, which is why they’re having feedback issues. I’m going for a zoned approach – dividing the audience area into four quadrants, each getting its own loudspeaker, aimed through the middle-middle of its own section. That looks like this:  The toe-in aim on the front pair closes the center gap nicely and also sends less energy towards the back of the room.

  • They also had both boxes per side daisy chained off the same drive signal, which means the folks sitting in the back would be getting two direct arrivals because the rear pair aren’t delayed, which makes intelligibility worse, not better. So I asked the crew to run direct lines from the “delay” pair over to the console.

At the console, I then patched L/R to the front pair of boxes, and patched the delays onto MTX1/2. I used the onboard oscillator to send pink to all four loudspeakers and added delay to the matrix outs until the localization shifted back towards the front. In lieu of an analyzer, this usually tends to get me within a few ms of the correct value. (“Correct” here is a slippery term because of the aim. The timing won’t hold for more than a few seats. But the intelligibility difference vs no delay at all is night and day.)

I grabbed an RF handheld and walked around, confirming that the coverage was far more even over the space.

The loudspeakers they were using are sufficiently close to a reasonable curve so I left the main EQ alone given the time constraint.

I rang out the podium mics, and was able to get about 6 dB more gain before feedback.

Then I fired up my RF analyzer and yikes:

Red bands are local DTV exclusions and that nonsense around 520 is coming off the BTR RF comms rack. Even if you don’t have mic freqs sitting on top of that, the comm system was just inches from the mic receivers and was probably overloading them. I had the guys move the comms rack backstage behind the drape line. The RF mic Rx should really be back there too, closer to the mics in use, but we didn’t have time to re-cable everything. I did a coordination for their seven frequencies taking the BTR into account, and walked the Tx around, checking for dropouts. It wasn’t perfect but it was functional (once I fixed the loose antennas on the Rx rack).

Finally I set up two more MTX outputs from the desk for a hearing-impaired feed and a teleconference feed. I showed the guys how they had zone control via the matrix faders but their whole show would follow the main LR mix so they didn’t have to worry about messing with it.

I also stressed to the TD that the money he spent on me bailing out his event would be better spent having me train his guys. Apart from coordinating RF and setting delays, everything else I fixed was something that was pretty simple and would have been noticed by guys with a little training who were paying attention, even if they’re not audio techs. A lot of small mistakes and oversights compounded to create an untenable situation.

So my point here is that, to me, all this is “system optimization.” Literally, getting the audio system to an optimized state, or functioning the best it can given the circumstances. This is taking a wholistic view that includes everything from mics to loudspeakers and everything in between. This is not about owning an expensive analyzer or a fancy reference mic. It’s about basic process and attention to detail. The stuff we do with the analyzer and the reference mic is just the cherry on top, and if everything else was done properly, should be relatively minor.

Leave a Reply

Your email address will not be published. Required fields are marked *