Musical Fidelity Digilog D/A processor Measurements
Sidebar 3: Measurements
Measured using the Audio Precision System One with a swept bandpass filter, both units exhibited a slight negative deviation from linearity, the amplitude of a dithered 1kHz tone at 90dBFS not quite reaching the 90dB level. While the Black Box 2's small 120Hz peak at 110dB is indicative of a full-wave rectified power supply whose ripple frequency is double the line frequency, with harmonics evident at 2kHz, 3kHz, 5kHz, and 7kHz (fig.1 in that review's measurements), attempting to measure the same thing with the Digilog resulted in a severe ground loop, the reason for which is still not clear. The spectrum for the Digilog shown in fig.1, therefore, was measured using the Audio Control SA3050A analyzer and can't be compared directly with the Arcam's graph. However, it is broadly similar, with harmonics noticeable at 3kHz and 5kHz and the main low-frequency noise component, the 60Hz mains frequency, apparent at approximately 108dB.
The Digilog's deemphasis curve (fig.2) is essentially perfect, while the Black Box 2 exhibited a slight error. In the linearity department, the Black Box 2 outperformed the Digilog, the difference at 90dB being 2dB and 3dB respectively. Linearity error is the difference between the actual level recorded on the test disc and the measured output of the player. As the signal level decreases, DACs become less linear. Fig.3 shows fade to noise with dither track on the CDS test disc as decoded by the Digilog. This is a signal whose amplitude varies continuously from 60dB to 120dB over a period of 30 seconds. A perfect DAC (which doesn't exist) would produce a straight line. Another way of looking at the Digilog's linearity is shown in fig.4. The vertical axis is error, and the horizontal axis is signal level. Since both the Digilog and Black Box 2 use the same grade of the TDA1541 DAC, these differences are probably random. Testing different samples may yield different results.Robert Harley