Measurements, Maps & Precision

As I wrote in this space last month, test-equipment manufacturer Audio Precision has loaned Stereophile a sample of their top-of-the-line SYS2722 system (footnote 1), which has both significantly greater resolution and greater bandwidth than the Audio Precision System One Dual Domain we have been using since 1989. The reviews you can read in this issue include the first measurements I have performed with this impressive piece of gear, though there are still a number of graphs I produced using our System One. In fact, with the equipment I tested using the SYS2722, I performed duplicate sets of measurements using both the System One and the Miller Audio Research QC Suite in order to get a handle on how close the three systems agreed. (They did on the tests where the SYS2722's improved resolution was not a factor.)

While the SYS2722 is capable of very much more than the System One, at present I am echoing the set of test procedures I have developed over the years as I shift the test regime to the new platform. This is for consistency's sake, and to enable me to get familiar not only with the new system, but also with the Windows XP–based interface for both control of the tests and graphing the results. As clunky as the System One's DOS control program was, it was as comfortable as a beloved but frayed and faded pair of Levi 501s, and I could get it to do all sorts of tricks without having to think overmuch. But even with the help of eBay, keeping and maintaining a stable of older PCs with the necessary AT-bus slots for the interface card is not getting any easier.

This is also a problem I'm facing with the DRA Labs MLSSA software-hardware combination I use for my speaker tests. Which reminds me that it has been many years since I listed the test equipment I use to generate the measurements sidebars that accompany Stereophile's reviews.

When the magazine was based in Santa Fe, New Mexico, both our test lab and the large room I used for speaker testing were in the office building we shared with Stereophile Guide to Home Theater (now footnote 2). However, when we relocated the editorial offices to Manhattan in June 2000, New York real-estate prices made that approach impracticable.

Fortunately, the Brooklyn house my wife and I had bought had a large basement that used to be a dentist's surgery. We reintegrated the basement with the rest of the house and, before moving in, knocked together what used to be the patients' waiting room and a storage room to create my listening room. The surgery at the back of the basement became my wife's office, and I took over what used to be the dentist's lab, complete with gas fittings for Bunsen burners and workbenches for my audio test gear. I had two new 20A circuits installed for the listening room and one for the test lab.

You can find a photo of my test setup here. The Audio Precision System SYS2722 is controlled by a Dual-Core PC running XP, and the Audio Precision System One Dual Domain by a 1997 Pentium MMX PC running DOS. This computer also carries a two-board Prismsound DScope II digital-domain analyzer, which also runs under DOS, and a National Instruments AT-bus card for the original Miller Audio Research Jitter Analyzer. The QC Suite from Miller Audio Research runs on two National Instruments PCI cards installed in a 2000-vintage Pentium III PC running Windows Me. This PC also carries an RME soundcard that I occasionally use to feed 16- and 24-bit digital test signals, via S/PDIF and AES/EBU links, to the components being tested.

This gear is used for testing amplifiers and digital components, in conjunction with a 20MHz Hitachi analog oscilloscope to monitor test and analyzer signals—but I also use a 1987-vintage 8-bit Heath storage oscilloscope to capture squarewaves. (This doesn't have an anti-aliasing filter, so as long as the sample rate is set fast enough, it gives an accurate picture of square waveforms.) I also have handy a Fluke 87 true-RMS multimeter, a Neutrik Minirator portable signal generator, an Old Colony inverse-RIAA adapter, a Heathkit oscilloscope calibrator, a Heathkit high-precision resistor switchbox, two high-power resistive loads, and a plethora of balanced and single-ended cables, plugs, and adapters.

To measure loudspeakers under quasi-anechoic conditions, I use DRA Labs' MLSSA system installed in yet another antique DOS computer, in conjunction with an Outline remote-control speaker turntable, an Earthworks Zero-Distortion microphone preamplifier, and calibrated Earthworks, Mitey Mike, and DPA microphones. For in-room measurements, I use an AudioControl Industrial SA305A 1/3-octave spectrum analyzer. However, I am increasingly using SMUGsoftware's Fuzzmeasure program, which runs on my Apple PowerBook, in conjunction with a FireWire-connected Metric Halo ULN-2 mike preamp and A/D converter.

It is tempting to believe that the file cabinets full of graphs that I produce with this gear can supplant the results of the careful listening performed by the magazine's review team. But the integration of measurements into Stereophile's equipment reviews is not to describe or replace the listening experience—that is, and probably will always be, impossible. As I wrote in November 1990, soon after we started our measurement program, "Those who place their belief in measurements alone should remember that it is still the experience itself that matters, not the description of the experience, no matter how thorough or well-researched that description."

The map is still not the territory.

In memoriam: John Blaine
Back in 1991, I was contacted by John Blaine, a Stereophile reader based in St. Louis, who wondered if I would be interested in featuring some of his recordings on our then-forthcoming Test CD 2. Medically retired at a relatively early age, John had been interested in recording since he bought an Ampex cassette recorder in the 1960s to make tapes for his car. Friends asked him to record a concert at Christ Church Cathedral in St. Louis, and he had been recording there ever since.

I listened to the DAT tape John Blaine had sent me, recorded with a mix of DPA and Schoeps microphones using early Sony digital converters, and was impressed not only by the sense of space that had been captured, but also by the placement of the musicians within that space, as well as the sheer musicality of the sound. I included two of John's recordings on Test CD 2: Schubert's Ave Maria, performed on violin and organ; and an arrangement of "St. James Infirmary" performed by The Brassworks. As well as these works, his 1984 recording of Ruth Slenczynska performing in concert piano works by Haydn, Brahms, Copland, Chopin, and Rachmaninoff is available as an Ivory Classics CD (CD-70902) and has been very positively reviewed.

Sadly, John passed away at the beginning of October 2007. He was an active member of the Gateway Audio Society; a photo of him at the controls of his recorder in Christ Church Cathedral can be found here. John's Memorial service was held in Christ Church on Saturday, October 27, 2007.

Rest in peace, John.

Footnote 1: I purchased the SYS2722 when the loan period ended.—John Atkinson

Footnote 2: merged with in July 2011.