The vast majority of my listening time with the REF600M was spent with them tethered to a pair of Spatial M3 Turbo SE loudspeakers ($2650). These open-baffle horn designs have a sensitivity of 94dB so they require very low-noise amplification. With the REF600M powered up I could hear only the faintest hiss if I put my ears within a few inches of the Spatial’s coaxial driver. I ran the M3/REF600M combination full-range with no crossovers limiting their bass extension. I combined the Spatial M3s with a pair of JL Audio Fathom f112 subwoofers set for a 45Hz crossover with a 24dB roll-off.
I also used the REF600Ms with a pair of AV123 X-Statik open-baffle loudspeakers modified by Skiing Ninja with an external crossover and No-Rez interior damping material. Again I ran the X-Statiks full-range and coupled one Velodyne DD 10+ subwoofer set to 50Hz crossover with a 12dB-per-octave roll-off.
The problem with saying that a solid-state or a switching amplifier is “tube-like” is that readers will inevitably think that you are referring to its warmer harmonic balance or softer top octaves. But the REF600M is not tube-like in that way. No, its tube-like characteristics are in its soundstaging and dimensional capabilities. To put it bluntly I’ve never experienced a switching amplifier that is as spatially accurate or three-dimensional as the REF600M.
Michael Morgan, who is a world-class on-location recording engineer, came to visit me during the review period. He brought with him some of his fine recordings. He immediately noticed the REF600M’s imaging capabilities. Through the REF600Ms all of his recordings were mapped out in the soundstage so clearly that you could easily and instantly locate every instrument or vocalist. Also the layering and depth retention was as three-dimensional as I’ve heard from any recording on any system. My own recordings also sounded more three-dimensional than with any switching (and most linear solid-state) power amplifiers I’ve had in my system. Only the Pass Labs X150.3 had the same level of dimensional accuracy.
One of the principal sonic failings laid at the feet of most switching amplifiers is their harmonic threadbareness. In timbral neutrality I’d place the REF600M on the darker, richer side of the razor’s edge, primarily due to its more nuanced lower midrange. Compared to the Bel Canto 300M amplifiers that I often use for the rear channels in my 5.1 system the REF600M had a less mechanical and more relaxed harmonic presentation. Not only was I immediately struck by the REF600M’s richer tonal palette, but also by its superior dimensionality and by spatial characteristics that were noticeably superior to the 300M.
The REF600M’s relaxed harmonic character was also a result of its less forward upper midrange. Compared to one of my reference power amplifiers, the April Music Eximus S1, the REF600M was less hi-fi-like and more natural. The midrange’s leading edges didn’t jump out, but instead remained within the confines of the music. Usually when an amplifier is more “relaxed” image specificity suffers as a result, but the REF600M still outpointed the Eximus S1 when it came to dimensional accuracy and depth retention. Simply stated, music sounded more like the real thing through the REF600M than it had via any switching amplifier I’ve heard before.
Does the REF600M do everything better than its competition? Not quite. Its mid- and low-bass response was not as vibrant and impactful as several of my reference power amplifiers. While not soft or overly fluffy in a tubey way, the REF600M did not produce the same amount of dynamic punch or bass slam as either the April Music Eximus S1 (used in bridged mode as dual mono amplifiers) or my freshly refurbished Pass Labs X150.3. (The reason for this refurbishment was due to increased noise from aging capacitors and resistors after almost 15 years of continuous use.) After its return from the factory the Pass was almost as quiet in terms of noise (without a signal) as the REF600M, with only a slightly higher hiss level and zero hum.