Back to BasicsSOUND SYSTEM SETUP, TUNING AND OPTIMIZATION 7/01/2006 8:00 AM Eastern
It may not be obvious, but a P.A. system is a precision instrument. Similar to the way that various components contribute to the makeup of a fine musical instrument, each element of a P.A. is critical to the operation and sound of the system as a whole. When the subsystems have been properly designed and implemented, the machine functions smoothly and effortlessly, accurately interpreting the commands of the user. Fine-tuning a P.A., avoiding phase problems while increasing clarity, efficiency and dynamic range, can mean the difference between adequate and great sound. A tweak here and there to a graphic EQ can smooth out harshness in the mids or reduce boom in the bottom, helping reduce listening fatigue. This month, Mix takes a look at some of the basic and not-so-basic things you can do to get the most out of a P.A.
BEFORE YOU TURN IT ON
Placement of speakers within a venue is critical to the quality of sound reproduction. Most important is avoiding feedback from the stage. The house array should be at least five feet forward of the front line of stage mics. If the P.A. is hung (or stacked) too close to the front edge of the stage, then feedback is inevitable. Line arrays may let you shave this distance by a foot or so, thanks to their directional control, but placing conventional boxes close to the performers is a recipe for disaster. In club situations, you may be able to get the house crew to push the house stacks forward (i.e., toward the mix position) to help avoid this problem. If that's not practical, or if the P.A. is flown, then pull the stage monitors and vocal mics upstage a bit to get them farther behind the house stacks.
Loudspeaker cabinets should not be flown or stacked within the arch of a proscenium because the proscenium itself can cause reflections that will interfere with the cabinets' intended dispersion pattern. In Atlantic City, N.J., the House of Blues actually added hang points in front of the proscenium to get its array farther into the room and out of the sonic influence of the proscenium.
Venues with large stages (wider than about 50 feet) present a problem for audience members in the first few rows. People in these seats won't hear vocals from the P.A. because they're too close to the stage and the P.A. is throwing sound to areas behind them. Sometimes the monitors spill vocals into the audience, but this may not be enough, and if the entire band is using personal monitors, then there will be no spill of any sort from monitors into the audience. In these situations, a front-fill is necessary.
The fill can be a compact speaker (EAW JFX200, L-Acoustics 112P, Electro-Voice X-Array Xcn or similar) with a 12-inch woofer and a 2-inch horn. Three such cabinets distributed across the front edge of the stage will usually do the trick. Amplification for these speakers can be fed from a matrix that carries only the vocal channels (and perhaps piano or some lead guitar), as the remainder of instruments usually can be heard from the stage. You'll need to listen from the front rows during soundcheck to determine how loud the fills need to be; bring them up just loud enough to add presence to the vocals — but not loud enough to disturb the main house mix.
Time permitting, every driver in the entire P.A. system should be checked for proper polarity. Obviously, this is more easily facilitated when you're the systems engineer and performing an install, but it's possible to do this even if you're a guest engineer. A polarity checker (more commonly and incorrectly referred to as a “phase checker”) can help, provided you observe two important guidelines: The device must be within inches of the driver and you must be able to perform the check with only one driver operating at a time. If you cannot use a polarity checker under these two constraints, then don't bother.
An oft-ignored aspect of loudspeaker placement (especially at the club level) is the relative position of high/mid/low-frequency drivers in P.A.s employing separate cabinets for these drivers. All drivers should be time-aligned for phase coherence. Time-alignment is physically achieved when the acoustic centers of the drivers are equidistant from the listener. An alignment error may not be apparent to the casual observer, particularly if the high/mid cabs are flown and the low or sub cabinets are ground-stacked.
In club situations, the subs are often downstage of flown mid/high cabs, meaning that arrival time of sound emanating from the high/mid cabinets is different from the arrival time of sound emanating from the low or sub cabinets. This results in phase errors, which are most apparent at frequencies near the crossover point(s). If you cannot physically arrange the boxes for proper alignment, then do so electronically. Just about every loudspeaker processor manufactured these days incorporates a separate delay for each output, so set a small delay to compensate for the difference. The rule of thumb is 1 ms of delay per foot; delays in this context are typically 1 to 3 ms.
While you're at it, delay the house stacks to the instrument backline by measuring the distance between the backline and the P.A. Delay the main system by roughly 1 ms per foot. In small clubs, this can make a marked improvement in low-frequency clarity because the P.A. can be brought in-phase with the backline. While it certainly won't hurt to initiate this delay in an arena situation, the ratio of P.A. to stage sound is much higher than in a club, yielding less-dramatic results.
If the opportunity presents itself, don't be afraid to vary the mix position. In spite of the fact that many engineers prefer to be centered between the left and right stacks, this listening position may not be representative of what the majority of the audience hears. P.A.s tend to “focus” in the center of the room, so when you're mixing at center, you'll hear more low end than those folks seated on the sides will. Conversely, setting the mix position closer to one side yields less low end than people in the center seats will hear. Either way, you'll need to make a mental “mix adjustment” to the low-frequency content of your kick drum and bass feed so it's consistent for as much of the room as possible.
MAKE SURE IT WORKS!
Having addressed some of the physical aspects of loudspeaker placement, turn on the P.A. and focus your attention on ensuring that all drivers and amps are operating properly.
Run pink noise through the system and analyze the result with a real-time analyzer. RTAs are no longer the domain of the rich and famous. Handheld models that include SPL metering and noise-generator functions can be had for as low as $300 and are worth every cent, especially when you mix on different systems every day. If you prefer using a computer for audio analysis, check out SIA SmaartLive or Metric Halo's SpectraFoo Complete, both of which include powerful audio diagnostic tools. (One of my favorite diagnostics is SpectraFoo's Transfer function. It can tell you what happens to frequency response and phase of a signal as it is passed from the input of a mixing console through to the output of a drive processor.)
Regardless of what measurement tool you are using, run noise from the generator into a channel with EQ set flat (and no processing) and route it to only the left side. Set the left and right master faders for the same gain and measure the SPL using a slow response time. Pan the signal to the opposite side and again measure the SPL. A difference of more than a dB or two indicates something is wrong. Possibilities include a power amplifier not being turned up all the way, a damaged speaker component or mismatched crossover settings. (More about that later.)
With the signal panned to one side, set the RTA to a slow response and observe the RTA curve. (Your meter or software may let you store the measurement.) Then pan to the opposite side. The RTA should show the same results within a dB or two. If not, use a ⅓-octave EQ to match the frequency responses of the left and right channels of the P.A. If the two don't sound the same, you'll go crazy trying to mix. Most P.A. systems also incorporate a compressor on the left/right bus, so pay attention to the position of the compressor in the signal flow; EQ'ing before the compressor can trigger compression.
Another suggestion for a quick checkup? Turn all faders down and patch a click (from a metronome or a test CD) into a channel on the console. Set the EQ flat, route the click to the left bus, slowly turn up the faders and listen. With the volume still way down, walk up close to the P.A. and listen to each cabinet to make sure it is producing sound. (Have someone baby-sit the console so that no one accidentally turns up the gain.) Repeat for the right side of the P.A.
If the P.A. system incorporates a loudspeaker processor (such as a BSS Omnidrive, dbx DriveRack, etc.), then mute all outputs on both channels. One at a time, turn on each output and listen to the system's sub, low, low-mid, mid- and high ranges, making sure that each range is working for left, right and (where applicable) center channels. If the system does not have a processor, then use a channel EQ to filter out the lows and mids, emphasize the highs and check the HF drivers. Then adjust the EQ to filter out the high and low range and emphasize the mids, etc., each time listening to the output of a particular range to make sure it is operating.
Be aware that playing music CDs through the system may not reveal problems with tired, borderline drivers, particularly in the low end (more on this below), so use a reference-test CD or tone generator to play sine wave tones ranging from 50 to 16k Hz through the system. (Don't use square waves as they mask distortion.) Low-frequency tone will quickly reveal woofers with voice coils that are rubbing, as well as rattling grillework or metal-to-metal contact near speakers. Third-octave tones can help reveal P.A. “hot spots,” which often show up at resonant horn frequencies of 1.6 kHz, 3.15 kHz and 6.3 kHz, and can be easily tamed using a 31-band, ⅓-octave EQ.
If a particular frequency range is especially hot, turn your attention toward the output controls on the loudspeaker processor to turn down (for example) the drive to the mid- or high-frequency amplifiers. (Do this before going for the EQ.) You may also need to tweak a crossover frequency (or slope) if there is too much energy due to overlap between the mids, high-mids and highs. A slow sweep tone played through the system should sound smooth and consistent through each range, with no hot spots.
INVESTING IN CDS
Many engineers use a favorite CD to tune P.A. systems. They become very familiar with a certain piece of music and they know how it should sound over a variety of different sources, such as headphones or studio monitors. Try listening to the piece of music through the P.A. and applying EQ until it “sounds right.” There's nothing wrong with this approach, but keep in mind that a CD — with its processing, compression and mastering — is not representative of the transients encountered with amplified live musicians. An alternative method employs using a vocal microphone for equalizing a system. (Carry your own mic for consistency and hygiene.) Talk into the mic to excite the resonant frequencies of the P.A. and then use EQ to correct them. In this case, you're using the same tools that you use in your live mix: compressors, equalizers and mics (as opposed to a CD).
Another technique calls for plugging a CD player into two channels of the desk and bypassing the channel EQ and the system graphic EQ. At the power amps (not the crossover), turn down the subs and the tweeters and listen to the midrange only. Slowly add the subs and tweeters into the midrange until you think it sounds balanced. If it's still necessary, use a graphic EQ for correction, but keep in mind that many engineers feel that any graphic EQ distorts the phase relationship of the input signal.
MIND YOUR BOTTOM
The majority of sound systems are set up with the L/R bus feeding signal to a drive rack (or crossover), which in turn divides the audio signal into various bands and routes signal to various amps in a multi-way P.A. In this type of system, it is possible to send a signal to the subwoofers that has no business being there, such as a hi-hat or lead vocal mic. Hopefully, the engineer has the smarts to use a highpass filter to remove whatever low-frequency crud might make its way into a hi-hat mic in the first place — thus preventing that signal from ever reaching the subwoofer.
However, if you really want to mind your bottom, here's an alternative approach: Use one of the console's auxiliary outputs as a subwoofer send. This concept removes unwanted audio from the subwoofer simply by virtue of the fact that you turn up the subwoofer aux only on the channels that need to be in the sub. Channels such as kick drum, floor tom, bass guitar and synth get fed to the sub aux, while channels such as lead vocals, hi-hat and the triangle microphone do not. The aux output designated as the subwoofer feed is usually routed to a crossover or lowpass filter, from the filter to a compressor and then to the subwoofer amplifiers.
In lieu of a “proper” crossover, I've seen engineers run the sub aux out from the console to a single-channel, 31-band EQ with all of the high-frequency sliders (say, those 125 Hz and above) pulled down all the way — thus acting as a filter for the high frequencies. Make sure that this aux is set to post-fader, or your low end will become disproportionate every time you move a fader. But be aware that when you mute the main L/R outputs of the system, the sub aux will not be muted and your audience will hear low-frequency rumblings from the subs.
While we're on the topic of low end, don't forget to take advantage of the highpass filters provided on each input channel. During soundcheck, audition each input over the P.A. (not headphones) one at a time, turn on the highpass filter and bring up the cut-off frequency until you can hear the low end start to drop out. Then back it off a bit. This will keep unwanted sounds such as mic stand rumble from ever reaching your low-frequency amps, preserving clarity in the bottom end and maintaining system headroom.
Tuning a P.A. system doesn't have to be rocket science, and the payoff is well worth the effort. The most important tools you need — your ears — are free. With planning and diligence, you can make the most of a subpar system or bring a state-of-the-art system to its highest level of performance.
Steve “Woody” La Cerra is Mix's new sound reinforcement editor.