Esoteric DV-50 universal player Measurements part 2
Fig.6 Esoteric DV-50, left-channel departure from linearity, 16-bit CD data (2dB/vertical div.)
Fig.7 Esoteric DV-50, waveform of undithered 1kHz sinewave at -90.31dBFS, 16-bit LPCM data.
As is increasingly the case with audio components from large corporations, the Esoteric DV-50's linearity was beyond reproach. A full-scale 1kHz sinewave was reproduced with just 0.0007% THD (true sum of the harmonics), and, as fig.8 shows, even then it was the benign second harmonic that was highest in level, at -105.3dB! This didn't change significantly with the Esoteric driving a demanding 600 ohm load.
Fig.8 Esoteric DV-50, spectrum of 1kHz sinewave, DC-10kHz, at 0dBFS into 8k ohms (linear frequency scale).
The intermodulation behavior depended on the digital filter in use. With just the FIR filter, a full-scale mix of 19 and 20kHz tones gave a difference component of around -107dB (fig.9), though the higher-order components at 18kHz and 21kHz were a little higher in level. (Note also the slight swell in the noise floor around the two primary components, which might well be jitter-related.) With the RDOT filter, the slower low-pass rolloff result in a much higher level of the aliasing component at 24.1kHz with 44.1kHz-sampled program material (fig.10). However, the inevitable in-band aliasing products all remain below -100dB, which is excellent. (For a different story, see the measurements accompanying my review in May, p.81, of the Ayre CX-7 CD player, which features a similar choice of digital filters.)
Fig.9 Esoteric DV-50, FIR filter only, HF intermodulation spectrum, DC-24kHz, 19+20kHz at 0dBFS into 8k ohms, 44.1kHz sampling (linear frequency scale).
Fig.10 Esoteric DV-50, FIR+RDOT filters, HF intermodulation spectrum, DC-24kHz, 19+20kHz at 0dBFS into 8k ohms, 44.1kHz sampling (linear frequency scale).
Finally, I examined the Esoteric's susceptibility to word-clock jitter by performing a narrow-band spectral analysis on its analog output while it played back a CD-R containing a high-level tone at exactly one quarter the sample rate, over which had been laid a low-frequency squarewave at the LSB level. (The latter signal exercises the maximum number of bit transitions.) The result is shown in fig.11. The actual jitter level was quite respectable, at 495 picoseconds peak-peak, with data-related jitter (red numeric markers) relatively low, other than the sidebands at the squarewave's fundamental frequency (red "10"s). However, not only are a large number of sidebands present at multiples of ±30Hz, but a double-humped swell in the noise floor can be seen around the central 11.025kHz tone. This latter ties in with the rise in the noise floor seen in the intermodulation spectra, and seems to be a characteristic of components that use an MPEG decoder. However, I have no idea what it results from.
Fig.11 Esoteric DV-50, high-resolution jitter spectrum of analog output signal, 16-bit LPCM data (11.025kHz at -6dBFS sampled at 44.1kHz with LSB toggled at 229Hz). Center frequency of trace, 11.025kHz; frequency range, ±3.5kHz. (Grayed-out trace is similar spectrum for DSD-encoded 11.025kHz tone.)
The grayed-out trace in fig.11 is a similar spectral analysis performed on the DV-50's output while it played back DSD data (from the "provisional" Sony test SACD) representing a 11.025kHz tone (with no low-frequency squarewave). There are no data-related sidebands present, of course, but the measured jitter level has increased to 613ps p-p, almost entirely due to a strong pair of sidebands at ±12Hz. Again, the noise floor shows the double-humped nature.
In almost all respects, and whether playing back CDs, SACDs, or DVDs, the Esoteric DV-50 gets a clean bill of health. However, its merely good jitter rejection gives rise to some minor concern on my part. Still, it's fair to note that Paul Bolin noticed nothing amiss in his auditioning.—John Atkinson