0When it comes to installing high-end audio systems, do you deliver on your promises? Anthony Grimani looks at whether or not you’re being true to your word.

Why does someone hire an audio integrator? There are plenty of DIYers out there, and tons of big-box retailers just waiting to sell gear at discounted prices. So why would someone pay a premium for an integrator?

One answer is that the person has plenty of money, but no time or interest. He/she wants the result but someone else to do the work. But, you may be surprised to find that customers who hire an integrator also expect to get better results because the integrator is an expert who can consistently deliver a high-end product.

Are you?

To find out, we have to define high-end audio. There is a persistent misnomer that ‘high-end audio’ is synonymous with ‘audiophile gear’, that if you sell clients nice stuff, then install it so it looks tidy and pretty, you’ve provided them ‘high-end audio’. Not so.

There is no high-end audio without the right frequency response, and you’re never going to achieve that by simply putting gear in a room. There are frequency-dependant amplitude phase errors that result from the sum of the equipment and the room. As long as those exist, there can be no high-end.

Think of it another way: buying a Ferrari doesn’t automatically make you a champion race car driver, does it?

Measuring and correcting the equipment/room response is absolutely mandatory for high-end audio, but here’s the rub. Even with the advanced gear and technology available today, the certainty of accurately measuring the audible frequency response is still too low.

This is true whether you use an automatic or manual method. Why? The simple version is that there is no follow-the-numbers approach that achieves a sonically pleasing and consistent result in all rooms with all equipment. One technique may achieve a consistently good result in one room every time, but a terrible result in other rooms. Another technique may randomly achieve a good result in a variety of rooms, but you’ll get five different results if you apply it five times in a row in the same room – some good and some bad!

So what is an integrator to do? I wish I had a magic-button solution for you, but I don’t. What I can offer you is some insight into the system that I’ve developed over decades of measuring and correcting high-end rooms, giving it my dead-level best to achieve both excellent and consistent results.

The basic concept is to use a multi-step process of checks and balances. In other words, don’t just use one technique all the time and blindly trust that it’s doing the right thing. Use multiple techniques and cross-check them until a consensus is reached.

The best results are still achieved by using good quality manually-adjusted digital equalisers. You could also start out with what the automatic system does, if you have one. But do not push the buttons and forget. You aren’t finished when it has run its course.

Next, bring in several manual measurements systems (such as REW) with spatial averaging to check the results of the automatic system. If you don’t have an automatic system (or its filters can’t be adjusted manually after the fact), this is the time to rough in manual EQ. Run some amplitude response charts with your analysers and compare the results. Hopefully, they aren’t wildly different. From this point, you have to use a bit of intuition and experience to figure out why discrepancies exist, what needs to be corrected and what doesn’t.

This is where the next cross-check becomes invaluable. Get yourself a pair of ER4s reference in-ear transducers from Etymotic and a portable music player with pink noise. Once you’ve got a decent-looking response on your analysers, use your ears to compare the timbre of pink noise from the speakers vs. the Etymotics (put them in and out of both ears – don’t do the one-ear-in, one-ear-out thing). If you’ve never done this, you’ll be amazed how easy it is to hear how the speakers are wrong. You’d never be able to tell without the reference point.

But why is it that a measured response that looks great on the screen of an analyser may actually sound wrong? It comes down to time-domain issues in both the direct sound from the speakers to the listener area, and in the reflected sounds bouncing off the walls. The vast differences in speaker off-axis responses, and reflected patterns will all sound different, while possibly measuring the same! So now you have to continue fiddling with EQ until the timbre matches – while observing the analyser data. Don’t let your ears make a change that is obviously completely stupid. If you feel like you need to do that, something is wrong somewhere else. Troubleshoot that before proceeding.

Finally – after the analysers, Etymotics and your ears all agree – break out a few music and movie clips that you know by heart: tracks you’ve heard on every system as far back as you can remember. Don’t use them to correct the response; use them to verify that your corrections are accurate and consistent. If the end result sounds wrong, go back to the analysers and Etymotics and figure out what you are seeing and hearing with them that might cause what you hear with program material.

Once again, the pink noise reference with the Etymotics is key. You know that’s going to be the same every time in every room. If you match each system to that, you’re going to have consistent sonic results.

Yikes! This sounds like a lot of complicated, expensive, extra work occurring entirely after the point where you normally pack up your tools and go home. I won’t lie; it’s not easy and it’s not something to which a lot of people are naturally inclined. It is, however, the only way we have right now, in 2016, to achieve truly high-end audio.

I want to challenge those of you who call yourselves audio integrators to take the final step and really do your job well. If you follow the path I’ve outlined, you’ll be able to promise your clients excellent, consistent results and achieve them every time. It won’t just be your word on it. You’ll be able to prove it. How do you measure the value of that?