Sooloos Music Server System Measurements
Of the three components that comprise the Sooloos system, it is the Source:One that outputs the audio signal. I measured the sample that had been auditioned by Kalman Rubinson; confusingly, although this identified itself to the system as a two-channel, single-zone Source:One, its front panel was labeled Source:Five, which I understand is a multizone version.
Internally, the relatively large chassis is populated with a switching power supply, an Ethernet network card, and an industrial single-board PC, into which are plugged a RAM daughterboard and a PCI soundcard. This latter, a high-performance Hammerfall HDSP 9632 from German manufacturer RME, is capable of decoding data sampled at up to 192kHz with a 24-bit word lengthit will hardly break a sweat handling CD-format data in the Sooloos system. Flying leads connect the digital and single-ended analog outputs of this card to the RCA jacks on the Source:One's rear panel. Setup was straightforward, once I realized I had to connect the system to my Internet router when I turned it on for the first time.
I examined the Source:One's measured behavior using Audio Precision's top-model SYS2722 system (see www.ap.com and "As We See It" in the January 2008 issue), as well as our Audio Precision System One and the Miller Audio Research Jitter Analyzer. (My results, of course, reflect the performance of the RME card; I understand that future production of the Source:One may well use a soundcard from a different company.) I ripped my library of CD-format test tones to the Sooloos Store using the slot in the front of the Control:One's base. Although my test signals are burned on a CD-R, when the Control:One accessed the Sooloos metadata library on the Internet, it identified the disc as a Norah Jones album; I assume the number of tracks and their timings must have been similar enough to confuse the Sooloos metadata library.
The maximum output level at 1kHz was 3.35V, 4.5dB higher than the 2V CD standard, sourced from a very low output impedance of 49 ohms at low and middle frequencies. The source impedance rose to 82 ohms at 20kHz, but this is still a very low value; there won't be any compatibility problems. The frequency response was perfectly flat from 10Hz to 20kHz (fig.1), and channel separation was excellent, at 100dB LR and 116dB RL, at 1kHz.
Fig.1 Sooloos Source:One, frequency response at 12dBFS into 100k ohms, 44.1kHz data (right channel dashed, 0.5dB/vertical div.).
The spectrum of the Source:One's analog output while it decoded data representing a dithered 1kHz tone at 90dBFS, using a swept 1/3-octave bandpass filter, indicated excellent linearity, with an absence of any harmonic spuriae, but uncovered some lower-frequency artifacts at 86Hz and its odd harmonics (fig.2). It's possible these were the result of ground leakage from the switching power supplythe ground connections on the RCA jacks on the rear panel of this sample are galvanically connected to the Source:One's chassisbut, peculiarly, they are not visible in an FFT-derived spectrum of the same output signal (fig.3).
Fig.2 Sooloos Source:One, 1/2 octave spectrum with noise and spuriae of dithered 1kHz tone at -90dBFS, 16-bit data (right channel dashed).
Fig.3 Sooloos Source:One, spectrum of dithered 1kHz sinewave at 90dBFS, 16-bit data (blue left, red right; linear frequency scale).
The presence of this noise resulted in a higher-than-usual amount of positive linearity error below 100dBFS (fig.4), but was low enough not to obscure the Source:One's reproduction of an undithered tone at exactly 90.31dBFS (fig.5). The three discrete DC voltage levels described by this signal are clearly visible, with good waveform symmetry.
Fig.4 Sooloos Source:One, linearity error, 16-bit CD data.
Fig.5 Sooloos Source:One, waveform of undithered 1kHz sinewave at 90.31dBFS, 16-bit data (blue left, red right).
Harmonic distortion was very low into high impedances (fig.6), with the second harmonic the highest in level at 90dB (0.003%). Some higher-order harmonics can also be seen, though these are all below 110dBFS and should therefore be inconsequential. However, even though the Source:One has a low output impedance, it is not happy driving very low impedancesinto 600 ohms, a regular series of distortion harmonics appears (fig.7). Intermodulation distortion was very low (fig.8), but some spectral spreading can be seen around the 19kHz and 20kHz fundamentals in this graph.
Fig.6 Sooloos Source:One, spectrum of 50Hz sinewave at 0dBFS into 100k ohms, 16-bit data (blue left, red right; linear frequency scale).
Fig.7 Sooloos Source:One, spectrum of 50Hz sinewave at 0dBFS into 600 ohms, 16-bit data (blue left, red right; linear frequency scale).
Fig.8 Sooloos Source:One, HF intermodulation spectrum, 19+20kHz at 0dBFS peak into 100k ohms, 16-bit data (blue left, red right; linear frequency scale).
I assessed the Source:One's rejection of word-clock jitter by playing the Miller Audio Research test signal, developed by the late Julian Dunn, and performing narrowband FFT analysis. The result is shown in fig.9: the central peak in this graph is the 11.025kHz tone, and the spectral spikes toward the sides are data-related sidebands. These are close to the test signal's residual level, but perhaps of more subjective significance are the lower-frequency sidebands that can be seen to either side of the central peak. These are spaced at ±86Hz and its multiples, the same frequency as the spuriae seen in fig.2. Again, this behavior may have resulted from the fact that the analog output jacks of Kal's review sample were grounded to the chassis, which, Sooloos's Rob Darling subsequently informed me, was due to a manufacturing error and is not typical of overall production. Even so, the calculated jitter level was still very low, at 254 picoseconds peakpeak, which makes it unlikely that this behavior underlay KR's preference for the Source:One's digital output feeding his Meridian digital preamp over the sound of the Sooloos's analog outputs.John Atkinson
Fig.9 Sooloos Source:One, high-resolution jitter spectrum of analog output signal, 11.025kHz at 6dBFS, sampled at 44.1kHz with LSB toggled at 229Hz, 16-bit data. Center frequency of trace, 11.025kHz; frequency range, ±3.5kHz (blue left, red right).