Disc Players

Six years on from our review, and lab test, of the CD9 [HFN May '13], and despite Audio Research playing down any substantive changes inside its new 'SE' model, there are still measurable differences between the old and new units. Running production changes and a new manufacturing facility make it virtually impossible to maintain precisely the same spec. over this period.

Sure enough, the tube complement is unchanged, as is the Burr-Brown (now TI) PCM1792 DAC. The Fast and Slow (linear phase) digital filter options still offer the same ~122dB and 8dB stopband rejection(s) with –0.3dB/20kHz and –3.4dB/20kHz treble roll-offs, all traded against greater (Fast) and lesser (Slow) ringing in the time domain. And differences? Well, this latest SE version has a slightly lower 4.5V (vs. 4.8V) peak output although the 296ohm source impedance, increasing to 509ohm at 20Hz, is the same. And yet the A-wtd S/N ratio has been improved from ~106dB to ~108dB (via CD and all digital inputs), honing its low-level linearity from ±1dB closer to ±0.3dB with CD.

Switching between native [black trace, Graph 1] and 'Upsample' [green trace, Graph 1] modes reveals a similar, slight increase in distortion over the top 25dB of its dynamic range (0.015 to 0.025% at 1kHz) but the frequency responses are slightly flatter in the SE: –1.3dB and –5.0dB/45kHz vs. –2.0dB and –5.8dB/45kHz (96kHz files with Fast and Slow filters). The biggest difference lies in the jitter spectra – relatively unchanged at 150psec vs. 120psec with CD but increased from 25psec via USB to >900psec in the SE (S/PDIF and USB) thanks, largely, to an added PSU modulation. The effect? Typically, some added bloom and warmth to the bass. PM