For ten years now, we’ve been measuring loudspeakers in the anechoic chamber at Canada’s National Research Council (NRC). We picked the place partly because it’s very close to our head office, but mostly because the NRC has long been held in high regard for its speaker-measuring capabilities. They began measuring speakers in the 1970s, when Dr. Floyd Toole, a music lover and audiophile, basically had the run of the place. They made their biggest impact in the audio world in the ’80s, when Toole published his legendary work on the correlation between loudspeaker measurements and subjective evaluation. NRC’s measurements of loudspeakers are considered by many to be the gold standard of accuracy.
The downside is cost. Renting a facility like NRC’s is not cheap, and I suspect that’s one of the reasons other publications don’t try to measure there or in a similar facility. In fact, to the best of my knowledge, the SoundStage! Network is the only publishing company in the world that measures its speakers in a properly equipped anechoic chamber. But that high cost means that we have to limit the number of speakers we measure there. We want the most accurate measurements, but we don’t want to go broke.
This issue of cost came up recently when it was suggested to us that we cut costs by using a lesser place -- perhaps a room, or an open space like a backyard -- so that we could measure more speakers for the same or lower cost. I considered it, as would anyone who runs a business, but ultimately dismissed the notion because of something that happened when we measured our latest batch of speakers at the NRC.
We were examining a speaker that has been on the market for about five years; on the whole, it measures admirably, something that has been revealed when it has been measured by other publications. Our measurements at the NRC, however, picked up an anomaly in the frequency response in the bass region that I’ve seen in no other set of measurements of this model. It caused me some concern, and made me wonder if our review sample was broken. It took me about a day to contact the designer; when I did, he confirmed that the anomaly we were seeing was real. He then commended us for using such a sophisticated facility to measure the speaker -- to the best of his knowledge, no one else had picked up this problem. However, measuring in the NRC’s anechoic chamber allowed us to do so quite easily. We’ll publish the measurements for this speaker soon, alongside the review.
This discovery reminded me of others we’ve made over the years that I can attribute to our use of the NRC’s anechoic chamber. In 2007, we measured the YG Acoustics Anat Reference Main Module loudspeakers, an expensive stand-mount design ($28,000 USD per pair at the time of review). Our measurements showed very flat on-axis frequency response but limited bass response, which you’d expect from a speaker of this size. Most notable in the measurements, though, were the high levels of total harmonic distortion -- more than you see from most speakers, particularly at the price. You need a very quiet environment, such as the NRC’s, to accurately measure distortion; to the best of my knowledge, no other publication based in North America measures THD, I suspect because they don’t have the capability to do so.
Another instance occurred with the JansZen Model One, which also carried a hefty price tag ($30,000/pair at the time of the review). The One’s measurements were quite respectable, given the nature of its hybrid design: front-firing-only electrostatic panels for the mids and highs, and a dynamic woofer for the bass. But our measurements showed a discontinuity in the frequency response at about 55Hz that -- according to designer David Janszen, who was present when we measured the One -- was attributable to a resonance that correlated with the height of the speaker’s enclosure. Accurate measurements in the bass region are difficult to make, and Janszen may not have seen this discontinuity in the measurements he himself took while designing the One. He basically described our use of the NRC’s facilities as looking at a speaker under a microscope.
To me, quality always outweighs quantity. If, in this case, that means measuring fewer speakers, but learning more about those we do measure, then that’s a tradeoff I’m willing to make. It’s also what sets us apart from our competition. In my opinion, the moment we began working with the National Research Council, the quality of our speaker measurements raised the bar for the entire audio industry, a situation that has not changed to this day. The gold standard, indeed.
. . . Doug Schneider
das@soundstagenetwork.com