Ayre D-1x DVD-Video/CD player Measurements

Sidebar 3: Measurements

Because the review sample of the Ayre D-1x didn't have a video output, I couldn't access the analytic signals on the Chesky test DAD (CHDVD171) to examine its behavior with 96kHz-sampled data. However, as the sample's primary role was as a CD player, I don't feel this to be a problem.

Like other machines based on DVD transports, the Ayre offered excellent error correction, coping with gaps in the data of up to 1.25mm in length without audible glitches. The maximum output at 1kHz was 3.04V RMS from the balanced XLRs, and half that from the unbalanced jacks. This is 2.34dB lower than the CD Standard's 2V, which will work against the player in A/B comparisons unless this is allowed for. The balanced source impedance was 560 ohms across the audioband; the unbalanced impedance half that at 280 ohms. Neither output inverted absolute polarity (the XLRs are wired with pin 2 "hot").

The Measure/Listen switch on the D-1x's rear panel affects the nature of the reconstruction filter. With this switch set to Measure, the Ayre's frequency response is perfectly flat from 10Hz up to 10kHz, with then a slight top-octave rolloff apparent of -0.4dB at 20kHz (fig.1, top pair of traces). With the switch set to Listen, there is no change below 14kHz, but the output above that frequency drops. While this slight in-band rolloff will have no significant subjective downside, it allows the reconstruction filter to be better behaved in the time domain, which some commentators believe to be associated with better sound. (Wadia, Pioneer, and Sony all offer players with filters optimized in this manner.)

The response was identical from the balanced outputs (not shown), but does not feature the 0.2dB channel imbalance seen in fig.1, which was peculiar to the unbalanced outputs. Pre-emphasized CDs were played back correctly, with no de-emphasis error. Channel separation, also not shown, was excellent from both sets of outputs, any crosstalk remaining below -100dB over most of the audioband.

Fig.1 Ayre D-1x, unbalanced, frequency response at -12dBFS into 100k ohms with rear-panel switch set to Measure (top above 10kHz) and Listen (bottom above 10kHz). (Right channel dashed, 0.5dB/vertical div.)

Fig.2 shows the audioband spectrum of the Ayre's balanced output while the player decoded dithered 16-bit data representing a 1kHz tone at -90dBFS. The traces are free from harmonic spuriae or idle tones, and the smooth noise floor is basically that of the dither used on the data. Extending the measurement bandwidth to 200kHz and repeating the analysis with "digital black" (not shown) revealed a very clean analog noise floor, with power-supply spuriae around the -130dBFS level.

Fig.2 Ayre D-1x, balanced, 1/3-octave spectrum of dithered 1kHz tone at -90dBFS, with noise and spuriae (16-bit data, right channel dashed).

The D-1x's DACs offer excellent resolution, any linearity error (fig.3) remaining below 2dB to below -110dBFS. As a result, the player's reproduction of an undithered 1kHz tone at -90.31dBFS, which is described by just three voltage levels, was superb (fig.4).

Fig.3 Ayre D-1x, right-channel departure from linearity, 16-bit data (2dB/vertical div.).

Fig.4 Ayre D-1x, waveform of undithered 1kHz sinewave at -90.31dBFS, 16-bit data.

Though the Ayre's distortion performance was never less than excellent, its behavior was different from the unbalanced and balanced jacks. From the balanced outputs into a high 100k ohm load, the dominant distortion harmonic was the third, at -84dB or 0.006%, with other harmonics hovering around the -100dB level (fig.5). This picture changed only slightly when the load was drastically lowered, to 600 ohms, suggesting the D-1x has good drive capability. But from the unbalanced output into 100k ohms (fig.6), the second harmonic rises from -100dB to -73dB (0.02%), and a power-supply artifact at 120Hz makes an appearance at -90dB (0.003%).

Fig.5 Ayre D-1x, balanced, spectrum of 50Hz sinewave, DC-1kHz, at 0dBFS into 100k ohms (linear frequency scale).

Fig.6 Ayre D-1x, unbalanced, spectrum of 50Hz sinewave, DC-1kHz, at 0dBFS into 100k ohms (linear frequency scale).

The Ayre offered basically superb performance when asked to decode data representing 19kHz and 20kHz tones, each at -6dBFS. (These sum to give a combined waveform that crests at 0dBFS.) The 1kHz difference tone from the balanced outputs remained below -96dB (0.0015%), even with a 600 ohm load. However, this did rise to a still low -82dB from the unbalanced output. This was with the left channel; the right channel was slightly worse at -74dB.

The picture was also different when the rear-panel switch was changed. With it set to Measure (fig.7), the only components that can be seen above the noise floor are the difference tone at 1kHz and the higher-order products at 18kHz and 21kHz. Set to Listen (fig.8), the slower ultrasonic rolloff associated with the reconstruction filter's better time-domain performance results in aliasing products appearing in the audioband, even though the true intermodulation products are slightly lower in level. These will probably be at too low a level to have any subjective effect, however.

Fig.7 Ayre D-1x, unbalanced, with rear-panel switch set to Measure, HF intermodulation spectrum, DC-25kHz, 19+20kHz at 0dBFS into 8k ohms (linear frequency scale).

Fig.8 Ayre D-1x, unbalanced, with rear-panel switch set to Listen, HF intermodulation spectrum, DC-25kHz, 19+20kHz at 0dBFS into 8k ohms (linear frequency scale).

Finally, the D-1x produced a fairly low level of word-clock jitter: 415 picoseconds with the analytical signal used by the Miller Audio Research analyzer. This was almost entirely due to two pairs of sidebands (fig.9), the first power-supply-related at ±120Hz (blue "2" markers), the other data-related (red "3"). The former was not affected by the various grounding arrangements I tried with the test setup (this test was performed using the unbalanced outputs) and could well have an effect on the perceived nature of the Ayre's low frequencies. Note the higher-than-usual noise floor in this graph, which to some extent is due to the D-1x's lower-than-usual output level from its unbalanced output. However, the noise is smoother, with far fewer spectral spikes, than I have found with other DVD-V players I have assessed with this test.

Fig.9 Ayre D-1x, unbalanced, high-resolution jitter spectrum of analog output signal (11.025kHz at -6dBFS sampled at 44.1kHz with LSB toggled at 229Hz). Center frequency of trace, 11.025kHz; frequency range, ±3.5kHz.

The Ayre D-1x's measured performance was superb by any standard, though it does suggest a different in character between the unbalanced and balanced outputs.—John Atkinson

Ayre Acoustics
2300-B Central Ave.
Boulder, CO 80301
(303) 442-7300