Bel Canto PL-1A universal disc player Measurements

Sidebar 3: Measurements

I hooked up the Bel Canto PL-1A to a TV in order to reset its defaults. I then disconnected the TV for the measurements, relying on the front-panel display. The maximum output level at 1kHz was the same for CD and DVD playback, at 2.13V from both the balanced and unbalanced output jacks. It was 1.5dB lower for SACD playback, however, which will be very audible in direct comparisons. Absolute polarity was preserved for all three media, and the source impedance was a low 20.5 ohms from the unbalanced jacks and twice that from the balanced XLRs, as expected. Error correction for CD playback, assessed with the Pierre Verany test CD, was disappointing, the PL-1A playing glitch-free through data-spiral gaps of only up to 0.5mm in length. It wouldn't play at all the tracks whose gaps were 0.75mm or longer.

The PL-1A's frequency responses for CD playback, with and without de-emphasis, are shown in fig.1. Both are flat up to 10kHz, with an inconsequential 0.25dB rolloff at 20kHz. This graph was taken from the balanced L/R output jacks; the response from the unbalanced RCA jacks was identical. With hi-rez media (fig.2), the ultrasonic behavior was slightly different for SACD than for DVD-A. 96kHz-sampled DVD-A had more energy than SACD in the 20–40kHz octave, but with then a sharp rolloff above 43kHz, while SACD continued its gentle low-pass rolloff to reach –7.7dB at 90kHz. Channel separation (not shown) was superb at better than 130dB in the midrange.

Fig.1 Bel Canto PL-1A, CD frequency response at –12dBFS into 100k ohms, with de-emphasis (bottom) and without (right channel dashed, 0.5dB/vertical div.).

Fig.2 Bel Canto PL-1A, SACD frequency response at –3dBFS into 100k ohms (bottom at 30kHz) and DVD-A frequency response at –12dBFS into 100k ohms (top at 40kHz). (Right channel dashed, 0.5dB/vertical div.)

The rather complicated-looking graph in fig.3 combines four different 1/3-octave spectra, three representing a dithered 1kHz tone at –90dBFS, the fourth a dithered tone at –120dBFS. With 16-bit data (top trace below 6kHz), the noise floor is basically that of the dither in the recording, not the Bel Canto's circuitry. Extending the PCM word length to 24 bits and playing back the test signal on a DVD-A (bottom trace above 2kHz and below 300Hz) dropped the noise floor by almost 18dB in the treble, suggesting that the PL-1A's D/A circuitry offers around 19-bit performance, which is excellent.

Fig.3 Bel Canto PL-1A, 1/3-octave spectrum of dithered 1kHz tone at –90dBFS, with noise and spuriae, 16-bit CD data (top below 6kHz), 24-bit DVD-A data (bottom at 10kHz), DSD data (top above 6kHz); 1/3-octave spectrum of dithered 1kHz tone at –120dBFS, with noise and spuriae, DSD data (bottom at 1kHz). (Right channel dashed.)

DSD-encoded data on the provisional Sony test SACD gave a spectrum with almost as high a dynamic range as 24-bit PCM in the midrange, but less so in the top two audio octaves (due to the start of the format's aggressive noiseshaping) and in the bass (where it has a higher random noise component). However, DSD can be seen to readily resolve the tone at –120dBFS. One anomaly visible in this graph is that both DVD-A and SACD gave a peak at 300Hz in their spectra. I have no idea what this is due to, given the PL-1A's commendable absence of power-supply-related spuriae.

The Bel Canto's DACs offered superbly low linearity error with hi-rez media, and with CD, the error was entirely due to the recorded dither (fig.4). As a result of its low noise and excellent linearity, the player's reproduction of an undithered 16-bit sinewave at –90.31dBFS (fig.5) clearly revealed the three voltage levels that describe the signal, while extending the bit depth to 24 bits gave an excellent sinewave shape (fig.6).

Fig.4 Bel Canto PL-1A, right-channel departure from linearity, 16-bit CD data (2dB/vertical div.).

Fig.5 Bel Canto PL-1A, waveform of undithered 1kHz sinewave at –90.31dBFS, 16-bit CD data.

Fig.6 Bel Canto PL-1A, waveform of undithered 1kHz sinewave at –90.31dBFS, 24-bit DVD-A data.

Distortion at full scale was superbly low, at 0.0011% (fig.7), and didn't change significantly as the load impedance dropped to below 1k ohm. Though the second harmonic was the highest in level in the left channel and the third harmonic the highest in the right, at –100dB or lower this difference will be insignificant. Reducing the signal level to –90dBFS dropped the noise floor (fig.8), revealing that the floor in fig.7 was actually that of the analyzer—and even at this very low level, the distortion harmonics are still almost 40dB down from the fundamental, which is excellent DAC performance. Intermodulation distortion was also very low (fig.9).

Fig.7 Bel Canto PL-1A, spectrum of 24-bit, 1kHz sinewave at 0dBFS into 4k ohms, DC–10kHz (linear frequency scale).

Fig.8 Bel Canto PL-1A, spectrum of 24-bit, 1kHz sinewave at –90dBFS into 4k ohms, DC–10kHz (linear frequency scale).

Fig.9 Bel Canto PL-1A, HF intermodulation spectrum, 19+20kHz at 1V peak into 4k ohms, DC–24kHz (linear frequency scale).

Finally, I used the Miller Audio Research Jitter Analyzer to examine the PL-1A's susceptibility to word-clock jitter. The solid trace in fig.10 shows a narrowband spectrum of the PL-1A's analog output while it played back a 24-bit, 11.025kHz tone recorded on a DVD-A, with the LSB toggled on and off at 229Hz. There are no sidebands visible at 229Hz and its harmonics, but two strong sideband pairs can be seen at ±15.6Hz (purple "1" markers) and ±1350Hz (purple "6"). The actual jitter level was 232 picoseconds peak–peak, which is very low. Reducing the word length to 16 bits (grayed-out trace) introduced some data-related sidebands as expected (indicated with red circles) and increased the jitter level to 273ps, but this is still very low in absolute terms. Clearly, the Bel Canto PL-1A offers excellent rejection of data word-clock jitter.

Fig.10 Bel Canto PL-1A, high-resolution jitter spectrum of analog output signal (11.025kHz at –6dBFS sampled at 44.1kHz with LSB toggled at 229Hz, 24-bit DVD-A data). Center frequency of trace, 11.025kHz; frequency range, ±3.5kHz. Grayed-out trace is 16-bit CD data.

Overall, the PL-1A is a well-engineered player that will extract what DVD-A and SACD have to offer, and will perform impeccably when playing back CDs.—John Atkinson

Bel Canto Design
212 Third Avenue North
Minneapolis, MN 55401
(612) 317-4550