Marantz Reference SA-KI-Pearl SACD/CD player Measurements
To measure the Marantz Reference SA-KI-Pearl, I used the magazine's loaner Audio Precision SYS2722 system. (See www.ap.com and "As We See It" in the January 2008 issue.) For some tests, I also used my Audio Precision System One Dual Domain and the Miller Jitter Analyzer.
Measured with the "provisional" Sony test SACD, the SA-KI-Pearl offered a lower maximum output level for SACD than for CD: 1.9V RMS vs 2.52V, a difference of 2.5dB. For CD playback, the maximum output level is a very audible 2dB higher than the CD standard's 2V RMS. The player's output preserved absolute polarity (ie, was non-inverting), and the output impedance was a usefully low 47 ohms across most of the audioband, rising slightly and inconsequentially to 52 ohms at 20Hz. Error correction for CD playback was superb and one of the best I have encountered, the Marantz player offering no glitches in its output signal until the gaps in the data spiral on the Pierre Verany Test CD reached 4mm in length.
Fig.1 shows the KI-Pearl's frequency response for SACD playback (green and gray traces), for external 96kHz-sampled data (blue and red), and for CD playback (cyan and magenta). All traces conform to the same basic shapeflat within the audioband, with then a fairly gentle rolloff above the audiobandwith the PCM responses cutting off sharply at half the sample rate. With DSD data (SACD playback), the output continued to roll off at ultrasonic frequencies, reaching 20dB at 120kHz. These traces were taken with the Filter 1 setting; there was no difference in the audioband with Filter 2, but the 96kHz playback was very slightly more extended than it had been with Filter 1 (not shown). Channel separation (not shown) was to specification at 100dB RL and 104dB LR across the audioband.
Fig.1 Marantz SA-KI-Pearl, Filter 1, frequency response at 12dBFS into 100k ohms with data sampled at 44.1kHz data (left channel blue, right red) and 96kHz (left cyan, right magenta), and with DSD data (left green, right gray, 0.25dB/vertical div.).
Analyzing the player's resolution by sweeping a 1/3-octave bandpass filter from 20kHz to 20Hz while it played dithered 16-bit CD data representing a 1kHz tone at 90dBFS gave the top pair of traces in fig.2. Above 300Hz, all that is shown is the spectrum of the dither noise, and the peak representing the 1kHz tone just touches the 90dBFS line, implying low linearity error. However, below 300Hz, peaks can be seen at the power-supplyrelated frequencies of 240, 180, 120, and 60Hz, though it is fair to note that these are all at a very low level. Feeding the KI-Pearl's data input with 24-bit data representing the same signal gave the bottom pair of traces in fig.2. The noise floor in the treble has dropped by 15dB or so, suggesting that the Marantz's DAC has a resolution of around 18.5 bits. At low frequencies, the supply-related spuriae are more clearly seen, and are a little higher in the left channel than in the right. SACD playback of the same signal gave the middle traces in fig.2; the rise in the noise floor above 2kHz is most likely due to ultrasonic noise from the DSD encoding leaking past the bandpass filter's skirts.
Fig.2 Marantz SA-KI-Pearl, 1/3-octave spectrum with noise and spuriae of dithered 1kHz tone at 90dBFS with 16-bit data (top), DSD data (middle at 2kHz), and 24-bit data (bottom). (Right channel dashed.)
Repeating the analysis with 16- and 24-bit data but using a narrowband FFT technique gave the traces in fig.3. Again, the increase in bit depth drops the noise floor by around 15dB, unmasking higher-frequency power-supply components, but these are all at or below 134dBFS, which in terms of quietness approaches the roots of the universe. Fig.4 shows the supply-related spuriae below 1kHz while the player decoded a 1kHz tone from SACD. The noise components at 60Hz and its odd multiples are most likely due to magnetic interference from the player's AC transformer; those at 120Hz and its multiples will be due to grounding issues. Again the left channel is slightly worse than the right, but again it is fair to note that none of this behavior will be audible. I suspect that the presence of the 120Hz-related spuriae may be due to Marantz's decision to use a two-pronged AC cord without a ground.
Fig.3 Marantz SA-KI-Pearl, FFT-derived spectrum with noise and spuriae of dithered 1kHz tone at 90dBFS with 16-bit data (left channel cyan, right magenta) and 24-bit data (left blue, right red).
Fig.4 Marantz SA-KI-Pearl, FFT-derived spectrum, DC1kHz of DSD-encoded 1kHz tone at 0dBFS (left channel blue, right red).
Linearity error with CD data (fig.5) was negligible to below 105dBFS, and with its low noise floor, the Marantz's reproduction of an undithered 1kHz tone at exactly 90.31dBFS was well-nigh perfect, with the three DC voltage levels clearly and symmetrically resolved (fig.6). A very slight degree of DC offset is evident, at +25µV in the right channel and 25µV in the left. Again, these will be inconsequential. With DSD data at the same level, the result is a pretty good sinewave (fig.7).
Fig.5 Marantz SA-KI-Pearl, left channel linearity error, dBr vs dBFS (2dB/vertical div.).
Fig.6 Marantz SA-KI-Pearl, waveform of undithered 1kHz sinewave at 90.31dBFS, 16-bit data (left channel blue, right red).
Fig.7 Marantz SA-KI-Pearl, waveform of dithered 1kHz sinewave at 90dBFS, DSD data (left channel blue, right red).
The Marantz's output stage offered low levels of harmonic distortion, even into the punishing 600 ohm load (fig.8), where it didn't behave differently than with the high 100k ohm load (not shown). The two channels did differ slightly, in that while the subjectively innocuous second harmonic was predominant, it lay at just 112dB (0.00025%) in the left channel (blue trace), but at 94dB (0.002%) in the right. Both channels also had some fifth- and seventh-harmonic content apparent, though at 110dB (0.0003%) and 118dB (0.00012%), this will not affect the Pearl's sound quality. While the Marantz offered very low levels of intermodulation distortion, even into 600 ohmsthe difference tone at 1kHz resulting from a signal consisting of high-level tones at 19 and 20kHz lay at just 99dB (0.0011%)its rejection of ultrasonic images for CD playback depended on the filter chosen. Filter 1, the default, allowed fairly strong images of the twin HF tones to leak past the filter's stopband (fig.9), while the sharper Filter 2 almost eliminated these images (fig.10).
Fig.8 Marantz SA-KI-Pearl, spectrum of 50Hz sinewave at 0dBFS into 600 ohms, 24-bit data (left channel blue, right red; linear frequency scale).
Fig.9 Marantz SA-KI-Pearl, HF intermodulation spectrum, Filter 1, 19+20kHz at 0dBFS peak into 100k ohms, 24-bit data (left channel blue, right red; linear frequency scale).
Fig.10 Marantz SA-KI-Pearl, HF intermodulation spectrum, Filter 2, 19+20kHz at 0dBFS peak into 100k ohms, 24-bit data (left channel blue, right red; linear frequency scale).
Finally, playing back the analytic J-Test signal from CD, the Marantz offered excellent rejection of jitter, though its noise floor looked more granular than the norm (fig.11). The jitter level, measured with the Miller Analyzer, was a very low 266 picoseconds peakpeak. The optical S/PDIF data input offered only mild rejection of word-clock jitter, however; when sourced from my PC via 15' of plastic TosLink, the J-Test data gave rise to a jitter figure four times higher, 1.13 nanoseconds pp, this almost entirely data-related (fig.12).
Fig.11 Marantz SA-KI-Pearl, high-resolution jitter spectrum of analog output signal, 11.025kHz at 6dBFS, sampled at 44.1kHz with LSB toggled at 229Hz, 16-bit CD data. Center frequency of trace, 11.025kHz; frequency range, ±3.5kHz (left channel blue, right red).
Fig.12 Marantz SA-KI-Pearl, high-resolution jitter spectrum of analog output signal, 11.025kHz at 6dBFS, sampled at 44.1kHz with LSB toggled at 229Hz, external 16-bit data via Toslink connection. Center frequency of trace, 11.025kHz; frequency range, ±3.5kHz (left channel blue, right red).
The Marantz Reference SA-KI-Pearl's measured performance reveals it to be a well-engineered SACD player, and better in some ways than the Marantz SA-11S2 Michael reviewed in February 2009. However, for best sound its Toslink input should not be used with jittery data sources.John Atkinson