The Fifth Element #28 John Atkinson May 2005
This Follow-Up concerns the X-01 Super Audio CD player ($13,000) from Teac subsidiary Esoteric (footnote 1), which John Marks wrote about in his February "The Fifth Element" column. Not only was JM impressed by this massive single-box player's prowess in playing SACDs, he also enthused about its CD performance: "This is the most musically satisfying CD playback I have ever heard," he wrote.
I spent a month auditioning the X-01. (Though this is a multichannel player, I used it only as a two-channel source.) While the Esoteric is undoubtedly a superb player, I did ultimately feel its balance to be somewhat on the cool side of neutrality with both CDs and SACDs. The Esoteric will take a system already leaning in that direction—for example, the NHT Evolution T6 loudspeaker, which I reviewed last month—too far down that road, I feel. It didn't quite scale the heights of resolution and sonic ease achieved by the +$40,000 dCS SACD system that I wrote about in March, but there is the small matter of the latter's price to be taken into consideration.
The Esoteric X-01's maximum output level from its unbalanced jacks was 2dB higher than standard, at 2.54V, sourced from a moderately high source impedance of 990 ohms. The left and right outputs are duplicated on balanced XLRs, these apparently wired with pin 2 hot. The maximum output from these jacks, sourced from 500 ohms, was not quite twice the unbalanced level, at 4.24V. Neither set of outputs inverted signal polarity.
The X-01's correction of CD errors was one of the best I have ever encountered. Playing the Pierre Verany Test CD, there were no audible glitches in the X-01's analog output when traversing gaps in the disc's data spiral up to and including 2mm in length, even though, monitoring the Esoteric's digital output with RME's DigiCheck program, I could see the data error flag illuminating. Only when the gap in the data reached an extraordinary 3mm did the X-01 mute its output.
The Esoteric's CD frequency response, with or without pre-emphasis, was flat. However, looking at its SACD response (fig.1), the player rolled off rapidly above 30kHz, the output being 12.5dB down at 40kHz. Esoteric has apparently chosen to more aggressively filter the DSD format's ultrasonic noise than have Sony and dCS. The channel separation (not shown) was superb, at almost 130dB in the midrange and bass and still more than 115dB at the top of the audioband.
Fig.1 Esoteric X-01, balanced frequency response at –3dBFS into 100k ohms (DSD data, right channel dashed, 0.5dB/vertical div.).
In common with other high-end players these days, spectral analysis of the X-01's output while it played CD data representing a 1kHz tone at –90dBFS really shows only the spectrum of the dither (fig.2, top pair of traces). A similar analysis with DSD data (fig.2, bottom traces) reveals an increase in resolution of 12dB or so, though some AC supply artifacts are unmasked by the higher-resolution format. These are higher in level in the left channel, but are still so low that they will be inaudible. Repeating the analysis with the "Analog Mute" track on the provisional Sony Test SACD (not shown) confirmed the presence of more aggressive ultrasonic filtering than is usual.
Fig.2 Esoteric X-01, 1/3-octave spectrum of dithered 1kHz tone at –90dBFS, with noise and spuriae, 16-bit data (top), DSD data (bottom). (Right channel dashed.)
The plot of the X0-1's linearity error with CD data (fig.3, top trace below –90dBFS) was a little noisier at very low levels than the best I have measured (see the behavior of the Musical Fidelity X-DACV3 elsewhere in this issue), but is still excellent. With DSD data (fig.3, straight line), the error was below 0.25dB down to the –120dBFS limit of this graph. Even at –130dBFS, which is at the limit of my test gear's performance, the level error was less than 5dB. Reproduction of an undithered CD sinewave at –90.31dBFS was perfect, the three voltage levels symmetrical and clearly defined (fig.4). With dithered DSD data, a sinewave at –90dBFS had a good waveform (fig.5), confirming the excellent resolution shown in fig.2.
Fig.3 Esoteric X-01, left-channel departure from linearity, 16-bit/DSD data (2dB/vertical div.).
Fig.4 Esoteric X-01, waveform of undithered 1kHz sinewave at –90.31dBFS, 16-bit CD data.
Fig.5 Esoteric X-01, waveform of undithered 1kHz sinewave at –90.31dBFS, DSD data.
As revealed by figs.6 and 7, the Esoteric had extraordinarily low harmonic and intermodulation distortion from both its balanced and unbalanced outputs. With a full-scale 1kHz tone into a fairly low impedance (fig.6), the THD (actual sum of the harmonics) lay at 0.012% left and 0.0006% right. The higher figure in the left channel is due to an increased amount of third harmonic, though at –99dB, this is still negligible. The 1kHz difference component resulting from the X-01 playing back a full-scale mix of 19kHz and 20kHz tones lay at just –108dB (0.0004%), with the higher-order components at 18kHz and 21kHz just reaching –100dB!
Fig.6 Esoteric X-01, spectrum of 1kHz sinewave, DC–1kHz, at 0dBFS into 4k ohms (linear frequency scale).
Fig.7 Esoteric X-01, HF intermodulation spectrum, DC–25kHz, 19+20kHz at 0dBFS into 4k ohms (linear frequency scale).
However, note the shaping of the noise floor around the tones in this graph. I noticed this behavior with all high-level, high-frequency signals I tried with both CD and SACD playback. In essence, the background noise in the top audio octave will be modulated by the signal. Does this matter? On the one hand, such modulation noise can be more audible than its absolute level might suggest. On the other, the X-01's misbehavior appears to affect only the high end of the audioband, and might well be masked both by the musical signal and by the human's ear's lack of sensitivity and discrimination in this region. And real music rarely has high levels of the high frequencies that stimulate this behavior.
This noise modulation can also be seen in the graph showing the X-01's rejection of word-clock jitter (fig.8). The grayed-out trace was taken with a high-level 11.025kHz tone on SACD, the black trace with a similar tone on CD over which has been laid an LSB-level low-frequency squarewave. There are no jitter-related sidebands present with the SACD spectrum, but there are with the CD data, at ±15.6Hz, ±108Hz, and ±229Hz, as well as at higher frequencies. The 229Hz sidebands are data-related; the others are unknown in origin.
Fig.8 Esoteric X-01, high-resolution jitter spectrum of analog output signal with CD 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 with DSD data representing an 11.025kHz tone at 0dBFS.
The weighted sum of these spuriae, calculated by the Miller Audio Research Jitter Analyzer, came in at 520 picoseconds. While this is still good performance, the jitter is significantly higher than the best modern components I have measured, and the measured figure doesn't take into account the modulation of the noise floor. (Note that the 6dB increase in the signal level on the SACD results in a higher noise floor than with CD, despite the latter's reduction in absolute resolution.)
On all the traditional measurements, the Esoteric X-01 performed in an exemplary manner. I wish all high-end audio components measured so well. However, the two aspects that bothered me were the restricted ultrasonic response on SACD playback and the modulation of the high-frequency noise floor. While the jury is still out on the subjective effect of the DSD format's extended high frequencies, the good sound of a player such as the McCormack UDP-1, which offers no better than CD resolution on SACD playback while preserving the ultrasonic information, is anecdotal evidence for its importance. I'm more bothered by the noise-floor modulation in the top octave. While I have given above the reasons this might not be subjectively significant, the engineer in me wishes it weren't present at all.—John Atkinson
Footnote 1: The X-01 costs $13,000. Manufacturer: Esoteric, TEAC America, Inc., 7733 Telegraph Road, Montebello, CA 90640. Tel: (323) 726-0303. Fax: (323) 727-7656. Web: www.teac.com/esoteric/NewEsoteric/index.html.