Network Audio Players/Servers

The technical performance of Mola-Mola's Makua DAC/preamp [HFN Aug '17] was exceptional in its own right but the Tambaqui's 'purist' implementation pushes still further at the state-of-the-art. Its balanced XLR output has the same 22ohm source impedance and although the maximum output (re. 0dBFs) is reduced here by 4.4dB from 9.8V to 5.88V, the S/N ratio is only 2.8dB lower at (a stupendous) 118.5dB. Moreover, the Makua's output is/was phase-inverting while the Tambaqui is phase positive – so beware even the most precisely level-matched listening comparisons!

Mola-Mola teases in its specification for the Tambaqui by stating 'THD not measurable – estimated –140dB' although, in practice, this rather depends on level and frequency. I was still able to discern residual 2nd-4th harmonics at higher frequencies, lower than recorded for the Makua (0.00014-0.003% re. 0dBFs, 20Hz-20kHz) but still fractionally higher than that achieved by another custom DAC – the Chord DAVE [HFN Apr '16] at 0.000020.00008%. Either way, the Tambaqui's 0.00005-0.0005% puts this exquisite DAC in the same class as those very few requiring a 'recalibration' of our distortion Y-axis from –120dB to –140dB [see Graph 1]. Subjectively, it's all rather academic, of course!

Again, the custom linear phase/apodising digital filter is unchanged, offering a 60dB attenuation of stopband images (26kHz re. 22kHz at 48kFs) with responses tailored by sample rate [see Graph 2]. Output is ruler flat to within ±0.02dB up to 20kHz with 44.1/48kHz media but rolls off slightly earlier at 36kHz and 59kHz (–3dB with 96kHz and 192kHz files). Low-level linearity is true to within ±0.1dB over a 110dB range while jitter is incredibly low: less than 10psec, all sample rates/all inputs. PM