NWAA Labs: Measurement Beyond The Atomic Level

The setting is surreal. As you drive into the Satsop Business Park in rural Elma, Washington (pop. 3500, max), eyes immediately fixate on the looming 481'-tall cooling towers of an abandoned nuclear facility (footnote 1). Remnants of the largest nuclear power plant construction project in the United States (footnote 2), the site was mothballed in 1983, in part due to concerns triggered by reports of what had happened at Pennsylvania's Three Mile Island four years earlier.

The abandoned site's reactor and turbine buildings now house NWAA Labs, a 13-year-old independent laboratory that tests loudspeakers and materials for the audio, acoustics, and construction industries. NWAA Labs' founder is Stanford-educated electrical and mechanical engineer Ron Sauro, 76. Initially, Sauro juggled work at NASA with performing as the Vox organist of the 1963 Gold Record–earning group the Rivieras (footnote 3). Sauro later became a sound system designer/installer for churches and arenas. He began measuring speakers and contributing papers to the Journal of the Audio Engineering Society circa 1990 and opened his first speaker-measuring operation, Western Electro-Acoustic Labs, in 2005.

To say that NWAA Labs is unique is a grand understatement. The lab, which boasts the two largest reverberation chambers in the world, occupies a temperature-stable building designed to withstand a 10-megaton nuclear blast on its roof and a magnitude 10 earthquake. The building rests on an exceedingly stable sandstone layer roughly 10,000' thick. After excavating a 500' × 500' hole to a depth of 300', the Washington Power Supply System (WPSS) dug a 6"-wide trench around the foundation slab. By separating the foundation's sides from the earth's crust, the trench was intended to prevent earthquake damage and transmission of vibration and noise.

Sauro likens the facility's construction to three nested Russian matryoshka dolls. He describes the structure in some detail: The outer-building walls are 5' thick and contain eight layers of 3" rebar. Forty feet inside those walls, a 6" trench surrounds an inner building of identical construction. Inside this building, within another, 1'-wide trench, a 3'-thick circular containment vessel, also of concrete and rebar, was built to hold the reactor. Three nested structures were thus created, each with its own floor, ceiling, and walls. A 40'-deep depression was then excavated underneath the containment vessel and made into a water reservoir. In case of a meltdown, the vessel's contents would drop into the reservoir and the fuel would solidify rather than burn.

In November 2021, I accepted an invitation from acoustical engineer Norman D. "Norm" Varney of A/V RoomService (footnote 4) to explore Sauro's multifloor laboratory. I joined a facility tour with members of ASA, the 7500-member Acoustical Society of America (footnote 5). After this event, I arranged a four-way Zoom and phone conversation with Sauro, Varney, and Stereophile Technical Editor John Atkinson.

When John learned of the chamber's size and its five overlapping modes at 25Hz, he marveled at the labs' capability to conduct extremely accurate power-response measurements.

"At very low frequencies," Ron interjected. "NWAA Labs can measure accurately down to 25Hz because it resides in a floating room inside a floating room, separated from the outside world by roughly 25' of concrete. The room's background noise at 1000Hz is an astounding –43dB!" (This is in reference to a scale where 0dB = 20µPa, the threshold of human hearing astounding, indeed!)

"It enables us to do some interesting transmission loss [TL] measurements, since we have 140dB on the other side of the TL opening between rooms," Ron said. "We average about –17dB in this room if you measure from 25Hz to 10kHz. That gives us almost 160dB of variation for transmission-loss tests. Therefore, when we do TL measurements, we can do full measurement for all frequencies between 25Hz and 10kHz with no notations or compensations."

NWAA Labs serves clients worldwide, with about half of its business coming from Canada (whose main measurement lab is the National Research Council, or NRC, in Ottawa). Anechoic chambers, which are used to simulate a free-field environment devoid of reflections (footnote 6), are few in North America: Paradigm has one in Toronto; Boeing's, the largest in the US, is 65' × 65' × 65' (footnote 7).

NWAA Labs' "turbine room," which is 650' long, 350' wide, and 80' high—large enough to hold four NFL football fields—is a true free-field space. The roof area is a whole acre. As John noted, it's so large that there's no need to gate impulse responses in this space because reflections are so attenuated after 160ms.

John was fascinated by the room's huge, curved microphone array, which contains 19 matched Earthworks M30 (ANSI Type 1) measurement mikes. The array is 4.1m in radius and is suspended 15' above the floor. Microphones are mounted within 1/16" of a central point, one every five degrees. Each mike is electronically compensated for distance to one sample at 48kHz.


This converted nuclear power plant is now used for measuring loudspeakers.

The diffuser Ron was measuring at the time of our interview sat on a pedestal connected to an LS-360 turntable that could move in increments of one-tenth of a degree. At the mike array's typical distance of 4.1m from whatever is being measured, Ron can accurately measure phase response at 10Hz.

John asked whether at a 4.1m distance it was possible to measure a large speaker and get proper integration of drive-unit outputs. Ron replied, "We know what our limits are. The spacing of drivers cannot exceed 48"." John, who acknowledged his envy, replied, "The big problem with loudspeaker measurements in general is you need to be further away than the largest dimension of the loudspeaker. That's fine for an LS3/5a but impossible for a significantly larger loudspeaker." Ron responded, "We can measure subs all the way down to 25Hz when we support them above our 5'-tall wedges. The wedges you see around the measurement pedestal prevent reflections from the floor. Our biggest sub has been something like 570–600lb. We also do long line arrays with different cabinets, as in a PA system. We use a forklift to drop heavy speakers into place."

When NWAA Labs opened, they focused mainly on pro-speaker measurements. As time progressed, they also began measuring audiophile brands including Bowers & Wilkins and Genelec. "You name it, we've done it," Ron said. "I've measured approximately 300 brands—probably a little over 3000 speakers—over the last 10 years. Speaker measurement is probably about 20% of our business."

Loudspeaker measurement
After acknowledging NWAA's fantastic ability to measure the far-field directivity, magnitude, and time-domain response of the device under test, John asked, "In your opinion, Ron, which of the measurements best correlates with what people hear? When they say that a speaker sounds 'good,' are they talking about the flatness of response, the evenness of dispersion, the coherence of the time-domain presentation, or the absence of distortion and resonance? Which is most important? My experience with subjective-listening tests is that people will agree if a loudspeaker has elevated low frequencies, but half will say that's a bad thing and half will say it's a good thing. They agree on the perception but disagree on the value judgment."

Ron offered no simple answer. "AES, ASTM, and all the groups that help develop standards have been working for 25 years to come up with a standard that says, 'This sounds good,' and we have not even come close to agreeing. All we can do is objectively describe what we can measure in terms of frequency response, phase response, radiation patterns—the things that don't change, depending upon the flavor of the speaker."

Norm, whose A/V RoomService focuses on resonance and acoustics, noted that even though all currently measurable aspects of speaker response interact with each other to produce the sound that reaches our ears, acousticians typically measure speaker response by playing a swept sinewave or pink noise. After that, they examine one parameter at a time. Our ear-brain system, however, performs multiple calculations simultaneously, accesses our experience, and then passes value judgments.

"This is why you have to spend some time learning how to look at the different measurements and determine how they all integrate with each other," Ron said. "You have to integrate them in your brain and understand that if you take this particular parameter and combine it with this other parameter, it's going to give this type of result. That takes time and experience. You can't just throw a graph out there and expect that somebody who has absolutely no experience integrating these measurements will understand what they're seeing."


Ground view showing the Free Field Lab and the Reverb Room.

John agreed. "As I wrote years ago, to understand measurements, you have to look at all of the measurements simultaneously. In the measurement sidebars in Stereophile, I try to characterize what is happening in ways similar to what you guys are describing and then look at measurements which are so out of the ordinary that they may have an effect on sound character.

"One of the things I fall over all the time is the effect of resonances. Floyd Toole and Sean Olive have said that the higher the Q (footnote 8) and the higher the frequency, the less likely the resonance will be excited and the less likely that it will affect sound quality. I find that some resonances have a low enough Q and a low enough frequency that I hear them when I'm doing the measurements. Yet, other people out there don't mention the resonance-induced coloration that I'm hearing."

Ron replied that each frequency range has a different driver set and a different far-field point. This distance, called critical distance, is where the direct sound from the speaker is equal in magnitude to the sound of the reverberant field.

John shared an anecdote. "Many, many years ago, I visited the home of the late Alastair Robertson-Aikman, the founder of SME, who had a very big listening room—something like 35' × 25', with a 15' ceiling. He equalized the sound at the listening position to be flat. But the highs were so exaggerated—tilted up like crazy—because he was equalizing the reverberant field."

"There are so many people who think they're experts," Ron lamented. "They take a Radio Shack level meter and say, 'See, it measures equally all the way around, so we have good coverage.' I look at them and say, 'No, you just happen to be measuring the reverberant field far enough out to where it won't show up all your mistakes.' ... The only way a speaker owner can properly use measurement for setup is to sit within the area where the direct sound of the speaker reaches them, and sound isn't affected by the reverberant field."

Drowning in sound
"When you're listening to sound in a room, there are two different kinds of sound," Ron said. "If you're within the critical distance from the speaker, you will hear the direct sound of the speaker. If you're beyond the critical distance, you will hear the reverberant field, aka the soundfield that is developed by the room.

"The best way I can describe the room's soundfield is by analogy. Let's pretend that a bathtub is the room, the bathtub drain is the equivalent of the absorption in the room, the input pipe is the speaker or sound source that adds energy to the room, the water is the reverberant soundfield, and the air is the direct soundfield.

Footnote 1: More information can be found here.

Footnote 2: Intended as a 3.9-gigawatt nuclear power plant that didn't need active pumps for cooling, it was on course to be the only completely passive Series 2 reactor in the United States. Instead, it was eventually given to Grays Harbor County, which turned it into a business park.

Footnote 3: He also played Hammond B3 briefly with Santana and the Doobies.

Footnote 4: Varney has used NWAA Labs for testing over the years.

Footnote 5: See youtu.be/JDyvmgMCuNY.

Footnote 6: "Anechoic," in Greek, means free from echo.

Footnote 7: The famous one at Bell Labs' Murray Hill, New Jersey, facility—now part of Nokia—where much important science was done, is now used to create art. See youtube.com/watch?v=wqCjE2WzhBk.

Footnote 8: Q stands for Quality Factor—it is related to the amplitude of a resonance and the frequency range that it covers.—John Atkinson


Anton's picture

Thank you!

Other than the LS-360 turntable, did he mention which electronics [associated equipment] they prefer?

(Hidden joke disclaimer.)

remlab's picture

You captured it all perfectly. Best primer on the loudspeaker/room/listener interface I have ever read. Wow!

rt66indierock's picture

Very nice article, I’ve had a fun over the years pointing out the room is the most important thing. Thanks for showcasing the people doing the heavy lifting.

mark_o's picture

I've driven by many times and had no idea what was happening at Satsop. Very interesting article, even without the local angle.

SteveG's picture

I’ve driven by many many times, on the way from Olympia to parts West. Never had any idea that this facility exists.

Dr Z's picture

jason, i really enjoyed your very interesting report — nice to see such a facility outside of canada for a change :) but i'll bet you a beer that no structure constructed by mere mortals could withstand a 10 MT blast (nuclear or chemical) "on its roof". the IVY MIKE and CASTLE BRAVO tests were both of approximately such yields and a ten-ish-foot height of burst; they each resulted in craters roughly a mile and a quarter wide and two hundred feet deep. admittedly, they were both in wet sand, so a conservative reduction of the crater diameter and depth would be by a factor of two to allow for solid rock — any building on whose roof that firecracker went off wouldn't be in kansas anymore…