Test tones are useful both for optimizing a pair of speakers' setup and the initial diagnosis of their sonic character. But music is what matters when it comes to judging sound quality.
In his book Sound Reproduction: The Acoustics and Psychoacoustics of Loudspeakers and Rooms, Floyd Toole discusses what he called "Circles of Confusion." Toole wrote that recording engineers sometimes make arbitrary choices that affect sound quality (footnote 3). When you insert a new component into your system and listen to a specific recording, therefore, you have no way of knowing whether what you hear is a characteristic of the recording or of the product (footnote 4).
To avoid that paradox, reviewers listen for the aspects of sonic character that are consistent with many different recordings. If, with one recording, the stereo imaging is unstable or the low frequencies loose-sounding, you can't know if those are qualities of the recording or the product being auditioned. But if they persist with many different recordings, it is reasonable to assume that they are characteristic of the product. For this reason, I use as many different recordings as possible in my review auditioning. I also use recordings where I was responsible for the engineering, mixing, and mastering, thus mostly sidestepping the Toole trap (footnote 5).
For judging low-frequency behavior, I reach for a recording I made in 2014 of Jonas Nordwall playing the Toccata from Widor's Organ Symphony No.5 (24/88.2 AIFF file). With the speakers powered by the Parasound monoblocks, the organ's mid- and upper-bass ranges in Portland, Oregon's First United Methodist Church sounded appropriately magnificent—as long as I didn't play the recording too loud. With a reflex-loaded loudspeaker, the woofer is effectively unloaded below the port or passive-radiator tuning frequency; the recording's high-level low-bass content had the BRX's woofer cones reaching what looked like a ½" peak-peak excursion!
I don't seem to be able to get off my recent Rachmaninoff kick. I followed my organ recording with Vladimir Ashkenazy's reconstruction of the composer's first symphony with the Concertgebouw Orchestra (CD, Decca 415-617-2). The bass-drum rolls that lead into the introduction of the theme in the first movement had sufficient weight and were well-defined. Even placed farther out in the room than I suspect is optimal, the GoldenEar BRXes sounded surprisingly full-bodied for a relatively small loudspeaker. The KEFs sounded lighter in weight in comparison.
The trumpets and trombones in this symphony, as well as in the excerpt from my recording of Elgar's The Dream of Gerontius on Test CD 2 (Stereophile STPH004-2), had just the right amount of brassy blattiness. (This term, coined by J. Gordon Holt, describes the sound of the air in the instruments going nonlinear at very high levels, similar to the crackle in the sound of a jet engine.) The balance between the BRX's midrange and treble is well-arranged.
Female voices fared well due to the BRX's pure-sounding midrange. I had recorded soprano Heidi Grant Murphy performing Marc Neikrug's Pueblo Children's Songs at the 1998 Santa Fe Chamber Music Festival; this recording is available on the CD Bravo! (Stereophile STPH014-2). Played on the GoldenEar speakers, Murphy's delicate tone was natural-sounding, without any emphasis or sibilance. Her voice was also stably positioned between the speakers. Similarly, Jennifer Warnes' darker-toned vocals on the title track from Famous Blue Raincoat, her classic album of Leonard Cohen covers (16/44.1 ALAC file, Private Music), sounded uncolored, with a palpable image suspended in space in front of me.
While the articulation of the instrument's notes was excellent, piano was a little forward-sounding compared with my reference KEF LS50s. I had noticed a touch of this character both with Marc Neikrug's Steinway D in Pueblo Songs and with my recording of Robert Silverman performing Liszt's Liebestraum (from Sonata, CD, Stereophile STPH008-2), also on a Steinway D. With recordings that were already forward-sounding, like Keith Jarrett's The Köln Concert (24/48 MQA-encoded FLAC file, unfolded to 24/96, ECM), this quality started to get in the way of the music making, even with the grilles in place. This was especially true when I drove the BRXes with the NAD M10 integrated amplifier. The BRX benefits from high-class amplification, such as the Parasound monoblocks.
For a relatively small loudspeaker, the GoldenEar's dynamics were excellent. Played at a maximum spl of 95dB(C) at the listening seat, the declamatory tune in the third movement of the Rachmaninoff symphony sounded suitably impressive, as did the climax at the end of the Gerontius recording. And while the drum samples in "Fit Song," from Cornelius's Sensuous: la musique de 21° siècle (ALAC file ripped from CD, Warner Japan EVE016), didn't have the chest-crushing impact with the BRXes that they had on the 20-times-the-price, full-range Vimberg Minos that I reviewed in April 2020, the GoldenEar's uncolored mids and clean highs were supported by sufficient upper-bass energy that I almost didn't miss the expensive speakers. Well, not really. But when I took off my critical listening cap and just wanted to enjoy the music, the BRXes served well enough.
An area where the little BRXes did match the big Vimbergs was stereo imaging. The spread of the vibraphone image, which I had recorded in stereo with a pair of Shure cardioid mikes and arranged to go from far left to just left of center when I mixed the Jerome Harris Quartet's performance of Duke Ellington's "The Mooche" from Stereophile's Editor's Choice CD (STPH016-2), was clearly defined. Similarly, the image of Billy Drummond's drums, which I had used in the mixdown as a backdrop that extended the full width of the soundstage, was accurately placed between and behind the GoldenEars.
I finished my auditioning of the GoldenEar BRXes with Periphery, a relatively new David Chesky recording of Indian musician and singer Priya Darshini (24/192 FLAC). Chesky's promotional copy for this album claims it "was recorded in stunning high definition fashion using a new recording methodology which captures crystal-clear vocal performance and the immaculate acoustics of the church in which the album was recorded." In a conversation, Chesky told me that Periphery was recorded with a ribbon microphone and a pair of omnis in a giant church. "Your room should open up to infinity when you play this back," he promised. With the GoldenEar BRXes driven by the Parasound amplifiers, it did!
Conclusion
I very much enjoyed my time with the GoldenEar BRX. While it benefits from being powered by high-quality amplification, its sound quality features an uncolored midrange, clean high frequencies, and enough mid- and upper-bass energy to satisfy, coupled with stable, accurate stereo imaging.
A couple of weeks before this issue's deadline, I received an email from Sandy Gross, letting me know that a decade after the company's premier he was stepping down as GoldenEar's president emeritus at the end of May 2020. This great-sounding, elegant-looking, affordable bookshelf loudspeaker is a fitting swan song for Sandy.
Footnote 3: See Sean Olive's discussion here. Footnote 4: See my discussion of this problem regarding stereo imaging accuracy here. Footnote 5: Mostly because, while I know what engineering choices I made, even sound engineers are at the mercy of their monitoring equipment, which, however carefully chosen, is never 100% transparent.
Female voices fared well due to the BRX's pure-sounding midrange. I had recorded soprano Heidi Grant Murphy performing Marc Neikrug's Pueblo Children's Songs at the 1998 Santa Fe Chamber Music Festival; this recording is available on the CD Bravo! (Stereophile STPH014-2). Played on the GoldenEar speakers, Murphy's delicate tone was natural-sounding, without any emphasis or sibilance. Her voice was also stably positioned between the speakers. Similarly, Jennifer Warnes' darker-toned vocals on the title track from Famous Blue Raincoat, her classic album of Leonard Cohen covers (16/44.1 ALAC file, Private Music), sounded uncolored, with a palpable image suspended in space in front of me.
While the articulation of the instrument's notes was excellent, piano was a little forward-sounding compared with my reference KEF LS50s. I had noticed a touch of this character both with Marc Neikrug's Steinway D in Pueblo Songs and with my recording of Robert Silverman performing Liszt's Liebestraum (from Sonata, CD, Stereophile STPH008-2), also on a Steinway D. With recordings that were already forward-sounding, like Keith Jarrett's The Köln Concert (24/48 MQA-encoded FLAC file, unfolded to 24/96, ECM), this quality started to get in the way of the music making, even with the grilles in place. This was especially true when I drove the BRXes with the NAD M10 integrated amplifier. The BRX benefits from high-class amplification, such as the Parasound monoblocks.
For a relatively small loudspeaker, the GoldenEar's dynamics were excellent. Played at a maximum spl of 95dB(C) at the listening seat, the declamatory tune in the third movement of the Rachmaninoff symphony sounded suitably impressive, as did the climax at the end of the Gerontius recording. And while the drum samples in "Fit Song," from Cornelius's Sensuous: la musique de 21° siècle (ALAC file ripped from CD, Warner Japan EVE016), didn't have the chest-crushing impact with the BRXes that they had on the 20-times-the-price, full-range Vimberg Minos that I reviewed in April 2020, the GoldenEar's uncolored mids and clean highs were supported by sufficient upper-bass energy that I almost didn't miss the expensive speakers. Well, not really. But when I took off my critical listening cap and just wanted to enjoy the music, the BRXes served well enough.
I very much enjoyed my time with the GoldenEar BRX. While it benefits from being powered by high-quality amplification, its sound quality features an uncolored midrange, clean high frequencies, and enough mid- and upper-bass energy to satisfy, coupled with stable, accurate stereo imaging.
Footnote 3: See Sean Olive's discussion here. Footnote 4: See my discussion of this problem regarding stereo imaging accuracy here. Footnote 5: Mostly because, while I know what engineering choices I made, even sound engineers are at the mercy of their monitoring equipment, which, however carefully chosen, is never 100% transparent.
