Today, no mid-size or large event goes ahead without acoustic measurement software to tune the P.A. system correctly. We say software because, apart from the dedicated SIM3 hardware, which deserves its own article and which we will cover another time, what engineers use today to align their systems are analysis programs that run on a PC or a Mac.
It goes without saying that with a small laptop and these programs we can dial in a complex sound reinforcement system. The question I put to sound engineers and students is:
Before these programs existed and sound systems were tuned this way, did concerts sound bad?
The answer is no. Concerts sounded very good, in most cases, in the '70s, '80s and '90s. It's true! But if we compare a system that sounded good untuned with the same system once it has been tuned, the latter will sound much better. We know the ear works by comparison, and that's the only way it can decide whether one system sounds better than another. It would be very hard to have an auditory memory with a big enough "buffer" to judge a system we are hearing right now against one we heard days, weeks or months ago.
So let's get to work! What do we need to perfectly tune our live sound system?
- A Windows PC or an Apple Macintosh running OS X
- Measurement software
- An external USB, FireWire, PCMCIA or PCI Express audio interface
- A measurement microphone
Computer
For practical reasons, the computer should preferably be a laptop. It doesn't have to be the most powerful machine on the market, either. The audio interface can connect to the computer in any of the ways we'll cover below, although for practical reasons we prefer an interface with two inputs and two outputs over USB.
At least one of the inputs should be XLR, for connecting the measurement microphone, with its own phantom power. The second input can be XLR or 1/4" jack, preferably line level, and both inputs should have their own gain controls. Two line-level outputs would also be advisable.
Interfaces That Meet the Requirements
Some interface models that meet the minimum requirements to work (for example):
Tascam US122
Edirol UA25
Digidesign Mbox
and many more.
Measurement Microphones
Measurement microphones should, naturally, be omnidirectional condenser mics with a flat frequency response. Any mic that meets those requirements will do. As with the interfaces, here are a few models at very different price points:
Bruel & Kjaer 4007
DBX RTA
Audix TR40
Behringer ECM8000
In a comparison of phase, frequency response, sensitivity and so on among several models, let's just say the price difference never justifies the difference in quality, since they all have very similar responses, although it's true that at high frequencies the shortcomings of some models become obvious.
For the tasks at hand, measuring transfer function, impulse response and SPL, the least expensive mics are enough. Keep in mind, though, that no omnidirectional microphone stays truly omnidirectional once we get above 5 or 6 kHz.
It's also worth stressing that a good measurement microphone will give a better and more reliable high-frequency response measurement.
Measurement Software
None of this gear would be any use without good measurement software, so here are the main programs in use:
Smaart Live (PC and Mac)
Spectralab (PC)
Spectrafoo (Mac)
Mac Foh (Mac)
It's interesting to see that more and more programs of this kind are being developed for the Mac. Why might that be? Even SMAART LIVE, a program that has traditionally run on Windows, is now available for it! Could it have something to do with the direction the company with the apple logo is taking with its new Intel-based Macs, which can run just about any operating system on the market, from Mac OS X to Windows XP and Linux, with impressive stability? Time will tell.
As we did with the acoustic prediction programs covered in earlier issues, we'll go over the main features of each program, focusing on how the most popular one to date, SMAART LIVE, works.
Spectralab
This is the least used of the four programs by engineers, although it is more common in electronics and audio labs, and its user interface has been considerably modernized in the latest version.
* Its main functions are:
Impulse response
RT60
Burst
RTA
Spectrum
Polar plot
WAV file import and export
Spectrogram and sonogram
Transfer function
Oscilloscope
FFT
Signal generator
Audio signal recording
Crosstalk measurement
Phase meter and stereo imaging
3D waterfall response analysis
Transient analysis
This program is relatively easy to use, although the parameter setup is designed for people with at least a moderate level of theoretical knowledge of acoustics and electronics. It is clumsier to use live than the other three, since it has fewer direct functions.
Spectrafoo
This is the longest-established program on the Mac, since the company behind it, METRIC HALO, is of very high quality, prestigious and well known for its top-level audio hardware and software. This measurement program is the veteran of the platform and has a huge number of functions, all of whose windows can be on screen at the same time. It is very practical to use and very useful in the studio, and in its latest version it works well live, too.
* Main features
Impulse response
Bit meter
Signal generator
FFT
Transfer function
Audio signal recording
Spectrogram
Lissajous phase
SPL summary
Envelope
Oscilloscope
Timecode
Peak and RMS level summary
Phase torch for both channels
RTA
Spectrum
Mac Foh
This program had been catching on at a tremendous rate among Mac users for live applications, and as of this writing it is used by many sound engineers around the world alongside SMAART LIVE. The software was designed from the start for tours and live events, and the quality of its graphics is impressive. It is made by and for live sound engineers.
* Its main functions are
Plug-in module support
RTA
Spectro
Impulse response
Transfer function
Digital speaker processor control
MIDI control
DMX control
CD and DVD recording and editing
WLAN control
Simultaneous analysis of multiple incoming signals through the audio interface
The functions and features of this program are constantly being updated, and we keep finding more modules and plug-ins. It is without doubt the most complete of them all. It is the most practical and fastest of the programs covered in this article, and it is easy and clear to use.
Smaart Live
Finally, we come to SMAART LIVE, the most widely used program in live sound. Its latest version runs on both PC and Mac, which is why it is destined to be one of the most used in the world. The program was born in the mind of an engineer at Meyer Sound who left for JBL and tried to create a software version of SIM, which at that time was the most powerful audio analysis hardware ever built for live sound applications. After JBL, he spent some time at EAW until Mackie bought the latter.
* Main functions
Transfer function
Impulse response
RTA
Spectrum
SPL level summary
Spectrogram
Noise criteria
RT60
Digital speaker processor control
SPL meter
Signal generators
This is the program whose use is most debated in Internet forums, including the one on its own website. So below we'll explain how to use it in practice for live work, and specifically the most important operation in system tuning: the transfer function.
Connecting the Rig for a Transfer Function Measurement
Below you can see the correct hookup. The measurement mic goes into one input of the audio interface, with phantom power on, and supplies the measurement signal. The other input carries the reference: the pink noise, usually generated by the program itself, so that output is connected to the mixing console and sent back out on a pre-fader aux. That way we are ready to compare the reference signal against the measurement.
Recommended Values and Parameters
COH TH: Coherence threshold
AVG: Number of averages for the trace
DB +/-: Shifts the trace up or down to normalize the display
MAG TH: Magnitude threshold used for the transfer function
WEIGHT: Weighting curves
SMOOTH: Smoothing
SR: Sampling Rate
FFT: The maximum value in Spectrum and FPPO in Transfer Function; this is the FFT time value and determines low-frequency resolution
TC: The FFT time constant, which changes automatically when you change the FFT
FR: The FFT linear resolution constant, which changes automatically when you change the FFT
Procedure for Measuring and Tuning a System
Set the Reference and Measurement Gain to the Same Level
1. Place the measurement mic at the distance where you want to do the tuning (the mix position).
2. Send an audio signal to the full-range system only.
3. Capture the time difference between the signal on the reference channel and on the measurement channel (AUTO SMALL), then enter the value it gives you.
4. Go to the TRANSFER FUNCTION window. It should show the frequency response trace and the phase response trace.
5. Store this trace in a memory.
6. Look at the phase trace in the low-frequency region (for example, from 50 Hz to 150 Hz), where it slopes downward. That slope is the delay per frequency produced by
The behavior of the loudspeaker
The cabinet design
The HPF and LPF filters used: corner frequency and filter order (dB/octave)
The internal position of each driver within the cabinet
7. Mute the full-range system and turn on the subwoofer.
8. Compare the subwoofer's phase trace with the full-range memory. The phase trace with the shallower slope in the crossover region must be delayed so that the slope is the same over that frequency range. When the phase slopes match in the crossover region, you have the same time "per frequency"
9. This leaves only three possibilities:
A) The two traces are identical and overlap: all is well.
B) The slope of the subwoofer phase trace is shallower than the slope of the full-range phase trace (in other words, the sub is early relative to the full-range). In that case, use any of the following three options:
- Add electronic delay to the sub to increase the phase slope.
- Increase the slope of the sub's LPF, which increases the delay time.
- Physically move the sub back to increase the phase slope.
If necessary, you can combine the options above to match the sub's phase slope to the full-range memory.
C) The slope of the sub's phase trace is steeper (in other words, the sub is late relative to the full-range). In that case, use any of the following three options:
- Reduce the slope of the sub's LPF. This reduces the delay time.
- Physically move the sub forward to increase the phase slope.
If necessary, you can combine the options above to match the sub's phase slope to the full-range memory.
10. Turn on the sub and the full-range system at the same time. If everything has been done correctly, the overlapping crossover region should "sum."
Note: To verify this, you can invert the polarity of the subs, and when you do, the crossover region should cancel. The correct polarity is the one where it sums. And once these steps are done, our P.A. system is tuned!
