Pioneer DV-AX10 SACD/DVD-A/CD player Measurements part 2
Fig.6 Pioneer DV-AX10, 1/3-octave spectrum of dithered 1kHz tone at -90dBFS, with noise and spuriae (from top to bottom): 16-bit CD data, 24-bit DVD data (right channel dashed).
When I repeated the spectral analysis with "digital black" data and extended the measurement bandwidth to 200kHz, I got identical spectra with 16- and 24-bit data (fig.7, lower pairs of traces). The Pioneer appears to be one of those sneaky designs that turns off its analog output when it detects this test signal, in order to produce a better measured S/N spec. I repeated the analysis with a 1-LSB DC offset data and got the same "fake" results. However, playing a dithered 1kHz tone at -140dBFS from the Sony test SACD gave the top pair of traces shown in fig.7. At low frequencies, the Pioneer's noise floor with its mute lifted really is that low! Above the audioband, you can see the usual peak due to the aggressive noiseshaping used by the DSD encoding. However, the DV-AX10's more restricted ultrasonic playback bandwidth reduces the height of this peak compared with the Sony and Philips SACD players we have reviewed.
Fig.7 Pioneer DV-AX10, 1/3-octave spectrum of "digital black" with noise and spuriae (from top to bottom): DSD SACD, 16-bit CD data, 24-bit DVD data (right channel dashed).
(Note: When I was performing all these spectral analyses, I found that the Pioneer's analog stage picked up video scan-line interference from the TV monitor I was using. To get the best performance from DVD-Audio and benefit from the Pioneer's superb dynamic range with 24-bit recordings, turn your TV off once you have read the disc menu and selected Play.)
I test DAC linearity error by sweeping a dithered 500Hz tone downward from -60dBFS and plotting the level of a 1/3-octave bandpass filter centered on 500Hz against absolute level. Using 16-bit CD data, I got an unusual result, as can be seen from fig.8. Instead of the usual straight line typical of modern DACs, some bumps and wiggles are visible below -80dBFS, with then an increasing positive error apparent below -95dBFS. This strange behavior is at odds with the excellent performance seen in figs. 6 and 7, but is perhaps associated with the waveform of an undithered 1kHz sinewave at -90.31dBFS as produced by the Pioneer (fig.9). This should appear as a stepped waveform with three distinct voltage levels apparent. Instead, it looks like a reasonable sinewave, which implies that the Pioneer is performing some resolution enhancement.
Fig.8 Pioneer DV-AX10, left-channel departure from linearity, 16-bit CD data (2dB/vertical div.).
Fig.9 Pioneer DV-AX10, waveform of undithered 1kHz sinewave at -90.31dBFS, 16-bit CD data.
As is often the case these days with Japanese designs, the DV-AX10's analog output stage was bombproof, both the harmonic distortion plot for a full-scale low-frequency tone (fig.10) and the intermodulation plot for a full-scale mix of high-frequency tones (fig.11) appearing superbly clean, even into the nasty 600 ohm load. There are no aliasing spuriae apparent in the latter graph (see my measurements accompanying Brian Damkroger's review of the Wadia 861 CD player in the March issue), which suggests my earlier conclusion—that the square-looking waveforms in figs. 3 and 4 are due to digital filter clipping rather than to a digital filter optimized for time-domain behavior—is correct.
Fig.10 Pioneer DV-AX10, balanced, spectrum of 50Hz sinewave, DC-1kHz, at 0dBFS into 600 ohms (linear frequency scale).
Fig.11 Pioneer DV-AX10, balanced, HF intermodulation spectrum, DC-25kHz, 19+20kHz at 0dBFS into 100k ohms (linear frequency scale).
Finally, the Miller Audio Research Jitter Analyzer revealed that the Pioneer's rejection of word-clock jitter is superb. The solid trace in fig.12 is a narrow-band spectrum of the DV-AX10's unbalanced analog output while it played back a CD-R containing the Julian Dunn "J Test" analytical signal—a high-level 11.025kHz tone with the LSB toggled on and off at 229Hz. The calculated jitter level was a superbly low 142 picoseconds peak-peak, with the data-related sidebands (red numeric markers) all very low in level. In fact, the highest-level sidebands lie at ±15.6Hz—purple "1" markers and ringed in magenta—and contribute 67ps to the weighted total. A pair of sidebands can also be seen at ±577Hz (purple "5," ringed in magenta), but these are very low in level.
The grayed-out trace in fig.12 is an identical analysis performed while the Pioneer played back the 11.025kHz tone from the Sony test SACD. For reasons best known to themselves, Sony's usually meticulous engineers left off the essential 229Hz LSB toggling. As a result, no data-related jitter sidebands can be seen, but the same sidebands at ±15.6Hz make an appearance. Note that the noise floor with SACD is significantly higher than with CD. This FFT-derived spectral noise is at odds with the analog-domain spectrum seen in fig.4, for which I have no explanation.
Fig.12 Pioneer DV-AX10, high-resolution jitter spectrum of analog output signal (11.025kHz at -6dBFS with LSB toggled at 229Hz, 16-bit CD data). Center frequency of trace, 11.025kHz; frequency range, ±3.5kHz. Grayed-out trace is similar analysis for SACD playback of pure 11.025kHz tone.
Some peculiarities aside, the Pioneer DV-AX10's measured performance with LPCM data on CD and DVD is first-rate. However, on SACD the overly aggressive ultrasonic filtering likely contributes to Jonathan Scull's disappointment with the sound of the hi-rez medium.—John Atkinson