Audio Research CD2 CD player Measurements

Sidebar 3: Measurements

The Audio Research CD2 proved well able to cope with damaged CDs. The Pierre Verany test disc has deliberate data dropouts; the CD2 could play up to track 33, a 1.5mm gap in the data, without the sound muting. The player's output level at 0dBFS from its single-ended jacks was 0.7dB higher than the CD-standard 2V, at 2.16V. The channels matched to within 0.05dB at 1kHz, which is excellent. From the balanced outputs, the output offered the expected doubling, to 4.3V. The source impedance was similar to specification at 350 ohms (balanced) and 195 ohms (unbalanced), these figures varying slightly with frequency and channel.

The top traces in fig.1 show the CD2's frequency response at 0dBFS—absolutely flat! The lower traces, offset by -1dB for clarity, show the response with de-emphasis. I was puzzled by this, as the technical description implies that the de-emphasis is carried out in the digital domain, meaning that it should be flat—yet there is a -0.67dB error apparent in the mid-treble. With the few CDs that are pre-emphasized, the CD2 will sound just a little more laid-back than is strictly accurate. Channel separation (not shown) was almost 100dB in the midrange, with a slight rise in the crosstalk apparent above 1kHz, reaching a still-superb 87dB at 20kHz.

Fig.1 Audio Research CD2, frequency response (right channel dashed, 0.5dB/vertical div.).

Fig.2 shows a spectrum of the CD2's output while it decodes dithered data representing a 1kHz tone at -90dB. There is a slight blip apparent in both channels at 200Hz—this could be a measurement artifact—and the traces are superbly clean. Performing a similar spectral analysis for a "digital black" signal and extending the measurement bandwidth to 200kHz gave the trace in fig.3. The 10dB drop in the noise floor compared with fig.2 suggests that the DAC actually mutes when it detects this signal.

Fig.2 Audio Research CD2, spectrum of dithered 1kHz tone at -90.31dBFS, with noise and spuriae (16-bit data, 1/3-octave analysis, right channel dashed).

Fig.3 Audio Research CD2, spectrum of digital silence, with noise and spuriae (16-bit data, 1/3-octave analysis, right channel dashed).

The Crystal delta-sigma DAC used by the CD2 offers both a very low noise floor and excellent linearity, as shown by fig.4, which shows the left channel's level error for a dithered 500Hz tone. (The right channel was effectively identical.) The error is negligible down to -100dBFS, and still better than ±2dB until below -110dBFS! Accordingly, the shape of an undithered 1kHz sinewave at -90.31dBFS (fig.5) is accurately reproduced, with the three distinct voltage levels easily discernible.

Fig.4 Audio Research CD2, left channel, departure from linearity (2dB/vertical div.).

Fig.5 Audio Research CD2, waveform of undithered 1kHz sinewave at -90.31dBFS (16-bit data).

The Audio Research's analog stage has good linearity; hitting it with a full-scale mix of 19kHz and 20kHz tones generated very little intermodulation (fig.6).

Fig.6 Audio Research CD2, HF intermodulation spectrum, DC-22kHz, 19+20kHz at 0dBFS (linear frequency scale, 20dB/vertical div.).

Finally, I used the Miller Audio Research analyzer (footnote 1) to look at the CD2's jitter performance. This drives the CD player under test with data representing an 11.025kHz tone at -10dBFS and a 229Hz tone toggling the LSBs on and off while it performs a high-resolution FFT analysis—32,768 points, with 64 FFTs averaged—to examine the noise floor in the analog domain. Stereophile used to examine jitter at the DAC's wordclock pin, using the Meitner LIM analyzer; while this produced consistent results, it is the effect of jitter in the analog domain that really matters. We have therefore switched to using the Miller analyzer.

Fig.7 shows an analysis of the CD2's noise floor around the central tone. The grayed-out spectrum showed for comparison is that of the Myryad MC 100, reviewed in the January '98 issue of Stereophile. (The Myryad's noise floor is 3dB high overall because the FFT length used for its measurement was only 16,384 points.) The total jitter level is 168.6 picoseconds (ps), compared with the Myryad's 166.7ps—both are among the lowest I have measured, though the CD2 does not have the rise in noise around the base of the fundamental that the Myryad has. The sidebands marked in red are 229Hz-related jitter components—the highest in level, marked "1," contribute 93.2ps worth of jitter. The bins marked in blue are spurious noise components; those marked in purple are jitter-related sidebands, but, at 1550Hz spacing, are of unknown origin. Overall, this is very low jitter.—John Atkinson

Fig.7 Audio Research CD2, high-resolution jitter spectrum of analog output signal (11.025kHz at -10dBFS with LSB toggled at 229Hz). Center frequency of trace, 11.025kHz; frequency range, ±3.5kHz. Grayed-out spectrum is that of the Myryad MC 100.

Footnote 1: Miller Audio Research is the company run by Stereophile's new Test & Measurement Consultant, English reviewer Paul Miller. Paul gave three presentations on what CD-player measurements mean at HI-FI '98 in Los Angeles. He can be contacted via e-mail at .—John Akinson
Audio Research
3900 Annapolis Lane North
Plymouth, MN 55447-5447
(763) 577-9700