CEA-2010: A Good Baseline for Bass?

The controversy

As I started publishing CEA-2010 measurements in S+V about a year and a half ago, I began to receive occasional inquires from engineers about how I was doing the measurements. Like me, they had no results they could compare theirs to, and they wondered if the results they were getting could be repeated by others.

This question came out into the open when Hsu Research introduced the VTF-15H subwoofer. Hsu Research happens to be one of the few manufacturers doing CEA-2010 measurements, and the VTF-15H happened to be reviewed by the only two high-profile publications doing CEA-2010: S+V and Audioholics. The problem was that everybody got different results. (It’s worth noting that despite the differing measurement results both S+V and Audioholics lauded the VTF-15H as an outstanding bargain.)

For example, Hsu’s rating of average Ultra-Low Bass output in the VTF-15H’s maximum extension mode (one port plugged, EQ1 setting selected) was 118.2 dB. Audioholics’ rating was 114.9 dB. My original rating was, if I recall correctly, slightly higher than Hsu’s. (Unfortunately, I lost the original review to a hard-disk crash and can’t get the exact number.) That’s not a huge difference in your living room — it’s unlikely you’d actually push your subwoofer to that level — but to people who like to make their buying decisions by the numbers, every dB counts.

After seeing the disparate results, I checked my measurement system and found that the device I was using to calibrate my levels was off by several dB. I bought myself a new calibrator, ran the measurements again, and got 113.9 dB as the average Ultra-Low Bass output — 4.3 dB below Hsu and 1 dB below Audioholics.

This disparity was noticed by the enthusiasts on AVSForum, resulting in an occasionally testy back-and-forth involving all parties. And it raised the question in my mind: “Whose numbers are correct?”

Dr. Poh Hsu, the founder of Hsu Research, asked me to visit his facility to see how he was doing CEA-2010 measurements. He had recently been visited by Don Keele himself, who according to Dr. Hsu had given the thumbs-up to Hsu’s measurement procedure. Meanwhile, I began corresponding with Audioholics’ Gene DellaSala to find out how his reviewers were doing their measurements.

After corresponding with both men, and realizing that I couldn’t identify major flaws in their measurement techniques, I realized we had a big problem.

Dr. Hsu is pretty smart; he’s been designing great subwoofers for more than 20 years, and “Dr.” refers to the Ph.D he earned at MIT. Likewise, anyone who reads a review or two on Audioholics can tell that DellaSala and his writers possess considerable knowledge and experience in the technical aspects of audio. Given that such extremely competent people achieved significantly different results using CEA-2010, I began to think the standard might not be as rock-solid and repeatable as one would hope.