dCS P8i SACD player Measurements
The dCS P8i put out a maximum of 6V from both CD and SACD with its Level set to "6V." Set to "2V" with the Menu button, this dropped appropriately, to 2V RMS. The source impedance was less than 1 ohm from both the unbalanced and balanced outputs. The signal polarity, with the Phase set to Normal, was noninverting from the single-ended outputs but inverting from the balanced outputs, implying that the XLRs are wired with pin 3 rather than pin 2 hot.
Tested using the Pierre Verany test CD, which has calibrated gaps in its data spiral, the P8i had the best error correction I have encountered. Monitoring its data output with RME's DIGIcheck program, there were no flagged errors or audible glitches until the gap in the data was a yawning 4mm! The dCS also coped with double gaps in the data, each up to 3mm in length, without audible dropouts.
Frequency response for CD playback with Filter 1 was flat (fig.1, top pair of traces). The responses with filters 2 and 3 didn't seem any different, while filter 4 gave a rolloff of 0.75dB at 20kHz (fig.1, middle traces). However, some ripples developed in the treble with pre-emphasized CD data (fig.1, lower traces). Fortunately, pre-emphasized CDs are rare these days. With SACD playback, the response extended up to 40kHz before starting to roll off gently, reaching –1dB at 60kHz and –3dB at 75kHz (fig.2). Channel separation (not shown) was better than 110dB in the audioband.
Fig.1 dCS P8i, CD frequency response at –12dBFS into 100k ohms with Filter 1 and without de-emphasis (top), with Filter 4 and without de-emphasis, and with Filter 1 and with de-emphasis (bottom). (Right channel dashed, 0.5dB/vertical div.)
Fig.2 dCS P8i, SACD frequency response at –3dBFS into 100k ohms (right channel dashed, 1dB/vertical div.).
Playing back upsampled CD data representing a dithered 1kHz tone at –90dBFS gave the top pair of traces in fig.3. The 1kHz component peaks at exactly –90dBFS and below 7kHz or so; the noise floor is due to the dither recorded on the test CD. The noise increases above that frequency, however, presumably due to the upsampling process. With DSD data, the noise floor drops significantly in the low treble and below, revealing that the DSD encoding has up to 20dB more dynamic range at these frequencies. The P8i easily resolves a 1kHz tone recorded at –120dBFS (bottom pair of traces). Above 7kHz, however, DSD offers no more resolution than 16-bit PCM. Some very-low-level AC-supply spuriae are apparent at 60Hz, 120Hz, and 180Hz with the DSD data, though I have a suspicion that these are actually encoded on the test SACD I use (Sony's "provisional" disc). Extending the measurement bandwidth to 200kHz and playing back upsampled CD data representing a low-level DC offset reveals the massive noiseshaping used by DSD (fig.4). Significant energy exists at 88kHz, approximately 43dB below peak level.
Fig.3 dCS P8i, 1/3-octave spectrum with noise and spuriae of (from top to bottom): dithered 1kHz tone at –90dBFS, upsampled 16-bit and DSD data, dithered 1kHz tone at –120dBFS, DSD data (right channel dashed).
Fig.4 dCS P8i, 1/3-octave spectrum with noise and spuriae of –1LSB, upsampled 16-bit data (right channel dashed).
The P8i's linearity error during CD playback was negligible down to below –115dBFS, the plot of error revealing merely the effect of the recorded dither noise (fig.5). The dCS player's reproduction of an undithered 16-bit sinewave at exactly –90.31dBFS was essentially perfect, the three DC voltage levels being clearly visible, with excellent waveform symmetry (fig.6). A DSD-encoded sinewave at the same level reproduced as a good sinewave, though with some HF noise apparent (fig.7).
Fig.5 dCS P8i, left-channel departure from linearity, upsampled 16-bit CD data (2dB/vertical div.).
Fig.6 dCS P8i, waveform of undithered 1kHz sinewave at –90.31dBFS, 16-bit data.
Fig.7 dCS P8i, waveform of undithered 1kHz sinewave at –90.31dBFS, DSD data.
Harmonic distortion components in the P8i's output were very low with SACD playback, with just 0.00045% THD (fig.8, true sum of the harmonics; note that the noise floor in this graph is that of the test analyzer, not the P8i). The innocuous second harmonic was the highest in level, at just –107dB (left) and –112dB (right). Intermodulation distortion was similarly low, even with CD data (fig.9).
Fig.8 dCS P8i, spectrum of 1kHz sinewave at 0dBFS into 8k ohms, volume control at –6dB, DC–10kHz, DSD data (linear frequency scale).
Fig.9 dCS P8i, HF intermodulation spectrum, 19+20kHz at 0dBFS peak into 8k ohms, volume control at –6dB, DC–24kHz, upsampled CD data (linear frequency scale).
I assessed the dCS's rejection of word-clock jitter using the Miller Audio Research Analyzer. Playing back a CD-R with the diagnostic signal (a high-level 11.025kHz sinewave overlaid with a 229.5Hz LSB-level squarewave), the P8i raised just 198 picoseconds peak–peak of jitter sidebands. These sidebands can be seen in the narrowband spectral analysis of the player's analog output (fig.10); data-related sidebands (red numeric markers) were at the residual level, and most of the measured jitter came from sidebands at ±15.6Hz (purple "1"), ±47.4Hz (purple "3"), and ±1355Hz (purple "12"). These components were also present at the same levels with SACD playback of an 11.025kHz tone (not shown). However, the noise floor with both CD and SACD playback was a little dirtier than I have found with other high-end disc players.—John Atkinson
Fig.10 dCS P8i, high-resolution jitter spectrum of analog output signal (11.025kHz at –6dBFS sampled at 44.1kHz with LSB toggled at 229Hz), 16-bit upsampled CD data. Center frequency of trace, 11.025kHz; frequency range, ±3.5kHz.