Tube Preamp Reviews

Robert Harley returned to the LS5 in November 1994 (Vol.17 No.11):
While I was auditioning the $4495 Audio Research LS5 Mk.I preamplifier that I reviewed in August, I took possession of the $3495 Sonic Frontiers SFL-2. As much as I liked the Audio Research, it was clear that the Sonic Frontiers was serious competition. In some respects, the SFL-2 was the winner. The Sonic Frontiers had a cleaner treble, with less grain. In fact, it was the SFL-2's lack of treble grain that pointed out the LS5's trace of edge on instrumental textures. The SFL-2 revealed that the LS5 had a very slight whitish quality to the upper mids and treble. I wouldn't have thought it possible—the LS5 was extremely clean and pure in the treble, but the SFL-2 took treble purity one step further.

The SFL-2 also excelled at soundstage transparency—that "hear-through" quality that makes the music come to life. Soundstage size was bigger through the SFL-2, but the LS5 had a greater ability to change the spatial perspective depending on the recording. On small-scale music—Gary Schocker, Flutist (Chesky CD46), for example—the LS5 presented a tighter, smaller, more intimate perspective. Although the LS5's sense of size on large-scale music was smaller than that portrayed by the SFL-2, the LS5 seemed to have a wider range of spatial perspective. In addition, the LS5 had a greater feeling of bloom. This was perhaps partially the result of the SFL-2's more forward and slightly drier midrange. In the context of my system, the LS5's less forward presentation better suited my taste.

Although the LS5 could be described as warm, full, and rich in the bass, the SFL-2 had a slightly weightier character. The region below about 60Hz was better defined through the SFL-2, with tighter pitch definition. The Sonic Frontiers also had a bit more extension and dynamic impact in the lowermost octaves. In the bass and midbass, however, the LS5 was more articulate, better revealing detail and nuance.

With single-ended sources, the all-balanced LS5 has to be used with Audio Research's $1495 BL2 input converter. I preferred the SFL-2 to the LS5/BL2 combination when using single-ended sources; the BL2 added a slight brightness and dryness to the sound. The SFL-2 clearly excelled in treble smoothness and overall liquidity.

Overall, I was very impressed by both preamplifiers, but if you have single-ended sources, the $3495 SFL-2 is a terrific bargain.—Robert Harley

Robert Harley wrote about the LS5 Mk.II in December 1994 (Vol.17 No.12):
My review of the Audio Research LS5 fully balanced tube preamplifier in the August 1994 Stereophile (Vol.17 No.8), along with my comparisons with the Sonic Frontiers SFL-2 in the November issue (Vol.17 No.11), conveyed my belief that the LS5 is a superb preamplifier, but perhaps not a contender for the State of the Art. At $4495—plus $1495 for the ARC BL2, needed to accept single-ended signals—the LS5's price raises the standards by which it must be judged. There are a lot of great preamps out there for six grand.

As much as I liked the LS5, I had a few reservations about its sound. My primary criticism concerned the preamplifier's treble, which lacked the liquidity I so enjoyed from the SFL-2, instead sounding a touch grainy. The LS5's treble had a somewhat whitish, almost metallic quality compared to the SFL-2.

Other than that shortcoming, I greatly enjoyed the LS5's spectacular ability to present a recording's space, depth, air, and detail. The LS5 threw before me a huge, deep, transparent soundstage. Moreover, the LS5 excelled at presenting a sense of space between the images, rather than fusing them together. It was as though there was a bloom of air around instrumental outlines that made the music more lifelike, less synthetic.

Audio Research apparently felt the original LS5 didn't live up to its potential: the preamplifier was replaced almost immediately after its introduction by the LS5 Mk.II, a fairly serious redesign of the product costing $4995. The LS5's 12BH7 tubes were replaced by Sovtek 6922 (6DJ8) types—the 12BH7s apparently became overly microphonic during shipment. Moreover, the Sovtek 6922 sounds great, and is available in quantity.

Next, the entire gain structure was overhauled. In the LS5, a solid-state switch put an 18dB pad in the signal path when in the 12dB gain setting. This switch remained in-circuit no matter what the gain setting. The Mk.II handles the two gain modes by a completely different method. ARC has eliminated both the solid-state switch and the pad, instead changing the open- and closed-loop gains simultaneously in the same ratio for both gain settings. This resulted in lower noise and more transparent sound. Indeed, making the LS5 quieter was a motivation for the redesign—some customers with very sensitive loudspeakers complained of noise. In addition, the op-amp–based DC servo was replaced by a DC feedback loop.

The LS5's Alps detented potentiometer has been scrapped in favor of a continuous-rotation, motorized Alps pot now used in all ARC preamps except the LS7. ARC claims the new pot not only sounds better, but also offers the ability to be remote-controlled. Remote volume control is a $500 option, either at time of purchase or as a retrofit. Although the pot is intrinsically motorized, the circuitry to drive the pot and the remote itself are what add $500 to the price.

A few other changes were made—power-supply tweaks and different capacitor brands—but the main differences are the new gain topology, different tubes, and elimination of the solid-state switching.

Sound: The LS5 Mk.II is clearly an improvement over the LS5. Although both preamplifiers have a close family resemblance, the Mk.II was more musical, the biggest difference being in the treble. The Mk.II's treble was smoother, cleaner, more refined, and better integrated with the rest of the spectrum, rather than sounding slightly separate from the music. In addition, the treble was less forward and incisive in the new design—a quality that better complemented the preamp's overall spaciousness. I also heard less grain, manifested as less spittiness on vocal sibilants, a more velvety violin sound, and a cymbal sound that was more like burnished brass and less like chromium.

The mids were also better portrayed in the Mk.II, sounding warmer and richer, with less forwardness. A slight edginess to brass instruments was ameliorated. Overall, the Mk.II was more harmonically coherent. The music had a greater sense of ease and flow through the Mk.II.

Nonetheless, I still wouldn't characterize the LS5 Mk.II as sweet, laid-back, or tubey. The preamp still has a trace of forwardness in the treble, and may not be quite as smooth and clean as the Sonic Frontiers SFL-2's high-frequency reproduction.

I also heard a marginal increase in transparency and soundstage focus—already the LS5's strong suits. Finally, there was less sonic disparity between the two gain modes. The original LS5 sounded so much better in the 30dB position that the 12dB mode wasn't worth using; the Mk.II made the 12dB gain mode sound as good as the 30dB mode.

The LS5 Mk.II made a synergistic match with Audio Research's VT150 power amplifiers. The VT150s' ultra-liquid mids and treble were a perfect complement to the Mk.II's highly detailed rendering. System matching is therefore vital in achieving a musically satisfying result; listen to the LS5 Mk.II in your system before making a final decision.

Conclusion: With the Mk.II revision of the LS5 preamplifier, Audio Research has taken an excellent preamp and made it truly world-class. The Mk.II expands on the LS5's strengths, and corrects the product's musical shortcomings. The more liquid mids, cleaner treble, and less incisive sound were all welcome changes. Fortunately, the Mk.II retained the LS5's ability to portray great space and depth, and, particularly, to reveal a wonderful sense of bloom around image outlines.

Overall, the LS5 Mk.II is a reference-quality preamplifier, and one with which I've spent many an enjoyable hour.—Robert Harley

Steven Stone listened to the LS5 Mk.II in July 1995 (Vol.18 No.7):
The Threshold T2 handled frequency extremes well. Highs lacked any etched quality, while low-frequency information was clean and well-defined. Only when I compared the T2 directly to the Audio Research LS5 Mk.II could I hear a very slight amount of additional high-frequency grain through the former. Low frequencies were slightly tighter on the T2 compared to the LS5, but the LS5 did have slightly more bass authority and bloom. Compared to the LS5, the T2's midrange was ever so slightly leaner, with the LS5 having a hair more lower-midrange energy.

The question is, of course, whether the T2 is subtracting harmonic information or the LS5 is adding extra harmonics. After much listening and thought, I concluded that the T2 did subtract a smidgen of harmonic juice from the midrange and bass, and added a hair of grain to the upper frequencies. The LS5 also had a more three-dimensional soundstage presentation. Both instruments and soloists had a certain dimensional palpability with the LS5 that was missing through the T2. The front of the soundstage through the LS5 began several feet behind the front of the speakers instead of at the speaker grilles, as with the T2.

A final difference between the LS5 and the T2 was that I was better able to hear into dense musical passages with the Audio Research. This was surprising, since transparency is a strong suit of the T2. Remarkably, the LS5 consistently surpassed it in rendition of inner details. Bear in mind that the LS5 is a purist design, lacking even a channel-balance control or any sort of display. The LS5 does have a remote, but it sports only Volume Up, Volume Down, and Mute. In terms of ergonomic ease, comparing the T2 to the LS5 is like comparing a modern Lamborghini to a vintage Bugatti.—Steven Stone

Steven Stone returned to the LS5 Mk.II in February 1996 (Vol.19 No.2):
Compared to other preamplifiers I've used recently, the Carver Research Lightstar Direct acquitted itself very well. The Carver had less personality than the Threshold T-2, lacking that robust midbass "slam" and ever-so-slight bit of top-end grain. The Carver also had somewhat better low-level detail, and a more neutral overall harmonic balance. Both preamplifiers were emotionally involving, and both did an excellent job of conveying the music's feeling. The Threshold reminded me of the Wheaton Triplanar Ultimate tonearm, with a more dynamic, even Rabelaisian, presentation. The Carver was more like a Clearaudio/Souther arm, with better low-level detail and more precise imaging, but less dynamic bravado.

Compared to the Audio Research LS5 Mk.II, the Carver had a slightly smaller soundstage coupled with a somewhat less holographic rendition of dimensionality. The Audio Research sounds a bit larger than life, and has a warmer, richer harmonic balance, especially in the lower midrange. The high-gain Audio Research has a much higher noise floor than did the Carver, which was silent as a grave. But the Audio Research's amazing dimensionality, liquid midrange, and explosive dynamic abilities give it an unforgettable sonic signature that may not be entirely neutral (as I discovered in my bypass tests), but is so musically satisfying that one is inclined to overlook its quirks while reveling in its sonic glories.

In some systems, the Audio Research's 33dB of maximum gain will make it indispensable. Anyone who is scared away from the LS5 II because of its lack of single-ended inputs should know that I've used it with a wide variety of single-ended sources via Neutrik and Boulder RCA/XLR adapters, with excellent results. It's still my favorite active preamplifier despite its low-level noise, lack of balance control, primitive remote, and excessive gain. Gorgeous sound has to count for something.—Steven Stone