Outboard DACs

Unsurprisingly, the Lina DAC X’s core performance bears very close comparison with that of the Lina Network DAC 2.0 [HFN Feb ’24] beginning with the 2.05V and 5.91V balanced outputs and 109.2dB/117.0dB A-wtd S/N ratio(s) – in 2V and 6V modes, respectively – and carrying through to the vanishingly low/cable-agnostic ~1ohm source impedance. These are a function of the analogue output stage while distortion is linked to both output and preceding digital stages (the latter becoming dominant at lower signal levels). The new M3 Ring DAC mapper runs at twice the rate and offers reduced ultrasonic noise over the M2 option, with M1 and M3 trading fractionally higher peak level THD (0.00017-0.0003% vs. 0.00012-0.00025%) for a reduced minimum of 0.00004-0.0001% over the top 20dB of the DAC X’s dynamic range [Graph 1]. Jitter is further squeezed in the DAC X at 9-12psec over all sample rates versus ~20psec in the Lina 2.0 [DXD upsampling; black, Graph 2], but these are all spectacularly low figures anyway! Graph 2 also clearly illustrates the difference in requantisation noise offered by the new DSDx2 mode [blue] versus the Lina DAC X’s standard LPCM-to-DSD upsampling [red].

The response(s), time domain behaviour and stopband rejection will depend on your choice of dCS’s six custom digital filters [see boxout]. The biggest subjective differences are likely to be experienced between F4 (slow linear phase), F5 (fast minimum phase) and F6 (apodising/fast linear phase) while the more ‘conventional’ linear phase F2 offers the flattest responses, out to ±0.01dB/20kHz, –0.50dB/45kHz and –7.5dB/90kHz with 48kHz, 96kHz and 192kHz files, respectively, albeit with a merely average 35dB image rejection. dCS’s Ring DAC includes digital headroom, so all filters aced the intersample clipping test. PM