Classé cdp-202 CD/DVD player Measurements

Sidebar 3: Measurements

The Classé cdp-202's balanced jacks put out a maximum of 3.8V from both CD and DVD, the single-ended RCA jacks 1.915V, the latter 0.4dB below the CD standard's 2V RMS level. The source impedance was a low 100 ohms from the unbalanced jacks, 596 ohms from the balanced jacks. The signal polarity was noninverting from both the single-ended and the balanced outputs, suggesting that the XLRs are wired with pin 2 hot, the AES-recommended practice.

Monitoring the cdp-202's digital output with RME's DIGIcheck program running on my PC, there were no flagged errors or audible glitches until the gaps in the data spiral on the Pierre Verany test CD were at least 2mm long. This is excellent error correction. Interestingly, DIGIcheck indicated that all 24 bits in the Classé's digital output are active, even for CD playback.

The cdp-202's frequency response was flat from 10Hz to 10kHz for CD playback, with just a small degree of rolloff at 20kHz (fig.1, middle pair of traces) and zero error for de-emphasized discs (bottom traces, offset by –1dB). Playing back test tones from a DVD-Audio disc I created using Minnetonka Software's DiscWelder program, the cdp-202's ultrasonic output (offset by +1dB) rolled off to –1dB at 32kHz and –3dB at 45kHz. This graph was taken from the balanced outputs; the right channel is 0.1dB higher in level than the left. Repeating the measurement from the unbalanced jacks (not shown), the left channel was now 0.05dB higher than the left. Channel separation was superb from the balanced jacks, any crosstalk being buried in the noise floor across the audioband (fig.2, lower traces). The crosstalk did rise with increasing frequency from the RCA outputs (fig.2, upper traces), but was still below –90dB below 10kHz, which is still excellent separation.

Fig.1 Classé cdp-202, balanced frequency response at –12dBFS into 100k ohms from CD with (middle) and without (bottom) de-emphasis, and from DVD-A (top). (Right channel dashed, 0.5dB/vertical div.)

Fig.2 Classé cdp-202, channel separation: unbalanced (top), balanced (bottom).

Fig.3 shows 1/3-octave spectral analyses of the Classé's output while it decoded dithered 16-bit (from CD) and 24-bit (from DVD) data representing a 1kHz tone at –90dBFS. All the traces peak at exactly –90dBFS, and the 16-bit traces are dominated by the recorded dither noise. (The swept-bandpass filter technique used gives a spectrum that appears to slope up with frequency.) Increasing the word length to 24 bits drops the noise floor by up to 12dB, implying a resolution of around 18 bits, which is good performance. However, it also unmasks some very-low-level power-supply components at 120Hz and 240Hz. Both sets of traces are free from any distortion harmonics.

Fig.3 Classé cdp-202, 1/3-octave spectrum with noise and spuriae of (from top to bottom): dithered 1kHz tone at –90dBFS, 16-bit CD data, 24-bit DVD data (right channel dashed).

As expected from fig.3, the cdp-202's plot of amplitude error against absolute level (fig.4) showed very good performance down to –105dB, with the error below that level due to the 16-bit data's dither content. The low level of audioband noise and excellent linearity meant that the Classé's reproduction of an undithered sinewave at exactly –90.31dBFS, which is represented in 16-bit data by just three DC voltage levels, was excellent (fig.5). Increasing the word length to 24 bits gave a good facsimile of a sinewave (fig.6), even at this very low level.

Fig.4 Classé cdp-202, right-channel departure from linearity, 16-bit CD data (2dB/vertical div.).

Fig.5 Classé cdp-202, waveform of undithered 1kHz sinewave at –90.31dBFS, 16-bit CD data.

Fig.6 Classé cdp-202, waveform of undithered 1kHz sinewave at –90.31dBFS, 24-bit DVD data.

The Classé player also had low levels of harmonic distortion. Fig.7 is an FFT-derived spectral analysis of its output while it played back a 24-bit/1kHz tone from DVD. The noise floor is flat with increasing frequency in this type of graph, and is actually that of the analyzer, not the player. However, the very small amount of AC supply spuriae appears to be characteristic of the player, as are the low-order harmonics that can be seen. The third harmonic is the highest in level in the left channel's unbalanced output, though this still lies at –100dB and will thus be subjectively benign. The second harmonic was the highest in level in the right channel, at –103.5dB! The THD percentage in this graph (true sum of the harmonics and not including noise) was 0.0014% for the left channel and 0.0008% for the right. Intermodulation distortion was also very low, the only products visible in fig.8 lying at –100dB (0.001%). The balanced outputs performed even better on these tests (not shown).

Fig.7 Classé cdp-202, unbalanced spectrum of 1kHz sinewave at 0dBFS into 4k ohms, DC–10kHz, 24-bit data (linear frequency scale).

Fig.8 Classé cdp-202, unbalanced HF intermodulation spectrum, 19+20kHz at 0dBFS peak into 4k ohms, CD data (linear frequency scale).

I assessed the cdp-202's rejection of word-clock jitter using the Miller Audio Research Analyzer. Playing back a CD-R with a high-level, 16-bit/11.025kHz sinewave overlaid with a 229.5Hz, LSB-level squarewave (both frequencies an exact integer fraction of the sample rate), the cdp-202 raised 315 picoseconds peak–peak of jitter sidebands, measured at its single-ended jacks. While this is about twice as high as the very best players I have measured on this test, it is still very low in absolute terms. Fig.9 shows a narrowband spectral analysis of the Classé's analog noise floor while it played back the diagnostic signal from CD. The data-related sidebands (red numeric markers) are close to the residual level in the signal, almost all the measured jitter coming from sideband pairs of unknown origin (purple numeric markers). The narrowness of the central peak in this graph, which represents the 11.025kHz tone, suggests that the cdp-202 offers good rejection of low-frequency random word-clock jitter. (Note that while these measurements are different from the cdp-202's specified jitter of <200ps, there is very little correlation between the results of different measurement techniques. My figures can be compared only with other measurements performed in an identical manner.)

Fig.9 Classé cdp-202, high-resolution jitter spectrum of analog output signal (11.025kHz at –6dBFS sampled at 44.1kHz with LSB toggled at 229Hz), 16-bit CD data. Center frequency of trace, 11.025kHz; frequency range, ±3.5kHz.

Repeating this test with a 24-bit version of the data recorded on a DVD-R gave a slightly higher amount of word-clock jitter: 406ps peak–peak. While the data-related sidebands have disappeared from the graph (not shown), the higher-frequency sidebands all increased slightly in level. I have no idea what this behavior indicates, but I doubt that it will have audible consequences.

Overall, the Classé cdp-202 offers superb measured performance that will get the most from CDs and DVDs.— John Atkinson

Classé Audio
5070 Franois-Cusson
Lachine, Quebec H8T 1B3
(514) 636-6384