Digital Processor Reviews

I performed a full set of measurements on the Moon 891 using my Audio Precision SYS2722 system, repeating some of the testing with the magazine's higher-performance APx555 analyzer. For logistic reasons, I measured a different sample from that reviewed by JVS; mine had the serial number 310A22426340. Before I lifted the Moon onto the test bench, I unfastened the three transit bolts on its base—a good thing I read the startup guide before starting the testing! I installed Moon's MiND app on my iPad mini, which allows the 891's volume to be controlled but doesn't allow settings to be adjusted.

As JVS only auditioned the Moon 891's digital inputs in his review—I understand that another reviewer will be reporting on the sound of the 891 as an analog preamplifier in a follow-up—I looked first at the performance of the Moon's DAC. I used the AES3 and coaxial and optical S/PDIF inputs, all of which accepted data sampled at rates up to 192kHz, as well as USB data sourced from my MacBook Pro. Apple's USB Prober utility identified the Moon as "USB HD Audio" from "MOON," with the serial number string "2 (None)," and indicated that the USB port operated in the optimal isochronous asynchronous mode. The AudioMIDI utility revealed that the 891 accepts 16-bit and 24-bit integer data via USB sampled at all rates from 44.1kHz to 768kHz.

The Moon's digital inputs preserved absolute polarity from the balanced and unbalanced outputs. As set up out of the box, the volume control operated in accurate 0.5dB steps. (Smaller, 0.1dB steps can be selected with the front-panel Set-Up button and menu.) With the control set to the maximum of "80," the Moon 891's output level with a full-scale 1kHz tone was 6.52V at both output types. Switching the output levels to "Fixed," the maximum levels were 2.13V. As the maximum voltage in variable mode is well below the level where the output stages clip (see later), the Moon 891's DAC offers a well-managed gain architecture.

Fig.1 shows the Moon 891's impulse response with USB data sampled at 44.1kHz. The filter is a minimum-phase type with all the ringing following the single sample at 0dBFS. The magenta and red traces in fig.2 show the filter's ultrasonic rolloff with 44.1kHz white noise data at –4dBFS. They reach full stop-band attenuation just above half the sample rate (this indicated by the vertical green line), with the aliased image at 25kHz of a full-scale tone at 19.1kHz (cyan, blue) suppressed by >80dB. The frequency response with 44.1kHz, 96kHz, and 192kHz data (fig.3) is flat in the audioband, with a sharp rolloff just below half of each sample rate.

The red trace in fig.4 plots the error in the analog output level as a 24-bit, 1kHz digital tone stepped down from 0dBFS to –140dBFS. Even at the lowest level, the amplitude error is <0.5dB, which implies superbly high resolution. An increase in bit depth from 16 to 24, with dithered data representing a 1kHz tone at –90dBFS, dropped the Moon 891's noisefloor by 30dB (fig.5). This implies a measured resolution of 22 bits.

With undithered data representing a tone at exactly –90.31dBFS, the waveform was symmetrical, with negligible DC offset, and the three DC voltage levels described by the data were clearly defined (fig.6). With undithered 24-bit data, the Moon 891 output a superbly clean sinewave (fig.7).

The Moon's digital inputs produced very low levels of distortion. The second and third harmonics are the highest in level (fig.8), but at close to –120dB (0.0001%) are negligible. Intermodulation distortion with 24-bit data representing an equal mix of 19 and 20kHz tones, each at –6dBFS, was also very low (fig.9).

Fig.10 shows the spectrum of the Moon 891's output when it was fed 16-bit J-Test data via TosLink. All the odd-order harmonics of the undithered low-frequency, LSB-level squarewave lie at the correct levels, and the central spike that represents the high-level tone at one-quarter the sample rate is appropriately narrow. Repeating this analysis with 24-bit J-Test data via both TosLink and USB gave a superbly low noisefloor (fig.11).—John Atkinson