Measuring speakers properly takes time, effort, an SUV (or truck), and quite a bit of money. All of these came into play on Thursday, April 18, 2013, our latest speaker-measuring day.

One of the big differences in the way that we do speaker measurements versus the way most other publications do them (if they do them at all) is that we use a properly equipped lab and an anechoic chamber -- it's all in building M-37 at Canada's National Research Council (NRC). It's the same place that Dr. Floyd Toole did a lot of his groundbreaking research work in the 1970s and '80s, correlating loudspeaker measurements with subjective listening impressions and basically giving the Canadian loudspeaker industry its starting point. It's also the same place that Paul Barton of PSB Speakers still uses to refine his latest designs, as do others. In contrast, most other publications resort to the amateur way of doing things: measurements in some room, on a driveway, or even in a backyard. To us, a chamber is really the only way to go about these things properly, even though the rental costs amount to a pretty penny.

We're fortunate because NRC's campus is only a few kilometers from my home, so we have been measuring speakers in M-37 since the year 2000. But before any speaker goes into the chamber, it makes a stop by my place to be broken in and checked for obvious defects. Doing this alleviates quite a few problems up front because, occasionally, speakers do arrive with defects, which are often the result of damage during shipping. This also allows me to familiarize myself with the speakers, even if the setup isn't ideal.

Three speakers to be measured

The picture above shows the three speakers going to NRC on the 18th (left to right): Amphion's Argon 7L, Canton's SLS 780 DC, and Grand Cru Audio's Essentiel, which are from Finland, Germany, and France, respectively. (The Tannoy speaker behind them, which comes from Scotland, was too big and cumbersome to be taken in the vehicle that day.) Batching the speakers into groups that range anywhere from three to five helps us make our speaker-measuring process more efficient, which allows us to measure more speakers in the chamber than if we had to rent the chamber for each one.

Below is a small gallery showing my trip to M-37 on the 18th, which happened to be a very rainy day.

Doug Schneider
Publisher, The SoundStage! Network