Digital Processor Reviews

I approached the Lens measurements from two perspectives: its effect of jitter at the word clock inside a digital processor, and how it changed low-level waveforms with its dither-generation function. First the jitter.

Fig.1 is the Theta Chroma 396's clock-jitter spectrum made by driving the Chroma with a PS Audio Lambda transport playing the CBS Test Disc. The test signal was a 1kHz, -90dB undithered sinewave. We can see the now-classic spikes of jitter energy at 1kHz and its harmonics, the result of the test signal creating interface jitter. The RMS jitter level, measured over a 400Hz-20kHz bandwidth, was 230 picoseconds.

Fig.1 Theta Chroma 396, word-clock jitter spectrum, DC-20kHz, when processing 1kHz sinewave at -90dBFS from PS Audio Lambda transport (linear frequency scale, 10dB/vertical div., 0dB=1ns).

Fig.2 represents the identical test conditions and signals, but with the Digital Lens between the Lambda transport and the Theta processor. The periodic jitter components are much lower in level (except the spike at 4kHz), and the spectrum is generally cleaner. The RMS jitter level also dropped from 230ps to 160ps.

Fig.2 Theta Chroma 396, word-clock jitter spectrum, DC-20kHz, when processing 1kHz sinewave at -90dBFS from PS Audio Lambda transport with Genesis Digital Lens in-circuit (linear frequency scale, 10dB/vertical div., 0dB=1ns).

Repeating the measurements with a Classé DAC-1 processor yielded less of an improvement in jitter performance—the DAC-1 had better jitter performance to start with. Fig.3 is the DAC-1's jitter spectrum when processing a 1kHz, -90dB sinewave without the Lens. The RMS jitter level was 135ps. Note the rather high levels of signal-correlated jitter. Fig.4 is the same measurement, but with the Lens in the signal path. The spectrum is only slightly cleaner, but the RMS level dropped to 105ps.

Fig.3 Classé DAC-1, word-clock jitter spectrum, DC-20kHz, when processing 1kHz sinewave at -90dBFS from PS Audio Lambda transport (linear frequency scale, 10dB/vertical div., 0dB=1ns).

Fig.4 Classé DAC-1, word-clock jitter spectrum, DC-20kHz, when processing 1kHz sinewave at -90dBFS from PS Audio Lambda transport with Genesis Digital Lens in-circuit (linear frequency scale, 10dB/vertical div., 0dB=1ns).

Looking next at the Lens's effect on low-level waveforms, fig.5 is the Classé DAC-1's reproduction of a 1kHz, -90.31dBFS undithered sinewave with a 16-bit input signal. We can clearly see the three quantization steps at this level: 0, +1, and -1. Fig.6 is the same waveform, but with the Lens in the signal path and set to Dither 2 (dithering the 15th bit). This is a large amount of dither—not unexpected, considering you can hear hiss from loudspeakers when the Lens is in Dither 2 mode. Note the expanded scale, needed to show the high level of dither energy overlaying the 1kHz sinewave. Whether or not fig.6 represents an increase in fidelity to the original signal is open to debate.

Fig.5 Classé DAC-1, waveform of undithered 1kHz sinewave at -90.31dBFS (16-bit data).

Fig.6 Classé DAC-1, waveform of undithered 1kHz sinewave at -90.31dBFS (16-bit data) with Genesis Digital Lens (Dither 2).

What really surprised me was fig.7, the Classé DAC-1's reproduction of this waveform with the Lens set to Dither 1. The Lens took in 16-bit data and output 20-bit data, with the lower three bits being dither. For some reason the waveform now appears to have four quantization steps, not three—as if some of the energy from the +1 level appears as the narrow spike below the -1 level. Note how narrow this new waveform component appears, seen as the negative-going spike between -50µV and -100µV. The signal's zero crossing axis also appears shifted from the zero horizontal division. Note fig.7's different scale; the waveform now traverses the range between ±100µV rather than fig.6's ±80µV. I can't think of any mechanism that would cause this behavior. Perhaps Genesis can provide some insight into this question in their Manufacturer's Comment.

Fig.7 Classé DAC-1, waveform of undithered 1kHz sinewave at -90.31dBFS (16-bit data) with Genesis Digital Lens (Dither 1).

Genesis supplied me with jitter measurements made on the Lens by Bascom H. King using his custom S/PDIF jitter analyzer. The Lens was driven by an S/PDIF signal generated by a device made by PrismSound that intentionally adds jitter to an S/PDIF signal. The signal was jittered by a 100Hz squarewave with a jitter amplitude of 10 nanoseconds (10ns). Fig.8 shows the jitter spectrum in the S/PDIF signal at the Lens input. Note that the 100Hz jitter component coincides with the 0dB horizontal division, which is calibrated at 10ns. You can see all the odd-order harmonics of the 100Hz squarewave in the jitter spectrum on the S/PDIF signal.

Fig.8 Genesis Digital Lens, word-clock jitter spectrum of S/PDIF input signal, DC-20kHz, measured with Bascom H. King jitter analyzer (0dB=10ns). Jitter signal is 100Hz squarewave with an amplitude of 10 nanoseconds (10ns).

The Lens's output jitter spectrum was plotted (fig.9) while the unit was being driven by this massively jittered signal. The Lens's total removal of interface jitter (see figs.11 and 12) is impressive. The Lens apparently does what it's designed to do: isolate the output from any incoming jitter. The residual jitter in fig.12 is the jitter analyzer's noisefloor, which is in the single-digit picosecond range. (The spike at 120Hz is probably power-supply noise).

Fig.9 Genesis Digital Lens, word-clock jitter spectrum of S/PDIF output signal, DC-20kHz, under the same conditions of fig.8, measured with Bascom H. King jitter analyzer (0dB=10ns).

Although the Lens appears from these S/PDIF measurements to eliminate jitter at its output, it doesn't eliminate jitter in digital processors. The Lens's output may be perfect or nearly so, but that doesn't mean your system will be jitter-free. Nonetheless, the cleaner word-clock jitter spectra and lower RMS jitter level measured in digital processors with the Lens connected correlated with the improvements in sound quality heard in the auditioning.—Robert Harley