Automatic audio equalisers have a fundamental flaw in that computers can’t ‘hear’ sound. Anthony Grimani, however, has developed a new method of calibration that achieves better results.

In my capacity as a home cinema calibrator, it falls to me to actually make systems sound good. Many others make essential contributions in the design and installation phases, but the calibrator is ultimately responsible for bringing all the pieces together to produce a harmonious result.

As a calibrator, you have a responsibility to both the client and all those other people who have entrusted their hard work into your hands. You’re there to make everyone look and sound good. It’s a big job.

I’m always looking for improved ways to get closer to the desired result faster (did you know that a truly thorough home cinema calibration can take three days for sound and two for video?) One such recent innovation that I have been curious about is all the automatic equalisation systems available now.

You know what I’m talking about. A product comes with a microphone and some software that takes a few measurements, crunches the numbers and spits out EQ curves. All you have to do is plug in the microphone and hit ‘Go’. While this is certainly a tantalising alternative to days of slogging through multi-layered menus and screens of confusing measurements, the thought has probably crept into your mind that it’s not possible for an automated process to do all this exactly right. And you may have noticed that when you actually run these automated tuning robots, the results are often, well, let’s just say far from stellar.

Well, you’re on target. As much as I’d like to recommend these systems as a way to improve your efficiency and you profit margins, I can’t. That’s not to take away from all the hard work that’s gone into them. They’re a cool idea, a step in the right direction, and have the potential to be a great time and effort saving device. They just need to be refined before we can trust them with something as sensitive as fine-tuning the sound in a truly high-end home cinema. Would you allow a robot to select your wine or prepare your meal at a five-star restaurant?

Here is the root of the problem with automatic systems: the human ear-brain hearing mechanism is a wonderfully complex and sophisticated process. It employs all manner of biological and neurological fi lters to minimise unimportant background sounds and maximise the clarity of the sound you’re focusing on.

The basic fact is that omnidirectional test microphones, sweeps, noises and windowing are not able to measure sound in the amazing way people hear it. And, you can’t correct what you can’t measure.

That’s the problem with automatic systems. They rely on standardised acoustic measurements that are flawed because they don’t match what we actually hear.

If you don’t believe me, you can try it for yourself. Play a test disc with pink noise signals circulating across the various speakers in the room and listen to a system after automatic equalisation. How well does the timbre of the various speakers match? Do the speakers sound alike or different? Does the pink noise sound natural and full-bodied? Or does it have a hollow, hissy, or roaring character? You’ll discover that none of the automatic systems get it entirely right, and some of them don’t even allow you manual control to fix it!

The outlook is bright, however. My company has worked out a new technology for measuring and analysing sound that comes closer than ever to approximating human hearing. It’s called the Sliding Band Integration Curve (SBIC) for now, until the marketing guys come up with a name that sounds cooler and means less.

SBIC is based on the latest psychoacoustic research, which indicates that humans hear predominantly direct sound spectrum above 1kHz and predominantly overall room sound spectrum below 160Hz. In between, there is a gradual transition from direct to reflected sound. In keeping with this, the SBIC is primarily concerned with the direct sound from speakers above 1kHz then transitions slowly down to include more reflected energy at low frequencies.

This can be done in a number of ways, but the most effective we have explored so far is to take measurements both in the relative near fi eld of the speaker and at the listening position. The measurements are then, either manually or electronically, combined on a sliding scale to create a fi nal response.

Does it work perfectly? No. You can’t blindly (deafly?) trust it or any other acoustical analysis system on the market now. But I’ve equalised many, many systems over the years, and this is by far the most accurate technique I’ve employed.

So how do you use SBIC? Is it a hardware box? Software? A plug-in? A set of steps you follow when taking measurements?

Right now, you can use the technique with a simple RTA, omni microphone and graph paper to plot the response. We’re also working with manufacturers to employ an SBIC function in the most popular measurement systems. And, of course, there are always the automatic systems that could build the SBIC into their measurement and correction algorithms.

As for me, I’m just excited about getting one step closer to a measurement system that gives us consistent, repeatable data that correlates to what we hear. It makes the job of the calibrator easier, and that benefi ts everyone involved in the project, including the client!

Chase Walton contributed to this column. MSR is represented in Australia by Wavetrain (www.wavetrain.com.au).