Optical has about twice jitter of coaxial but in case of Benchmark suppression is so high that it doesn't matter. Jitter (noise in time domain) creates sidebands at very low level (-80dB typ) but still audible since not harmonically related to root frequency. With complex signal whole bunch of sidebands becomes low level noise (hard to detect since present only when signal is present). Transitions of digital waveform become uncertain (timewise) when in slew rate is low and noise is present (transport, DAC or cable) or when transitions are fast (high slew rate) and signal reflects on impedance boundaries or combination of both.
To minimize it I would use very well shielded digital cable (double or even triple shielded) that is about 1.5m long very well characteristic impedance matched. Shorter cable might bring reflection back on the same transition that caused it affecting threshold point (since limited slewrate) but the longer one will make reflections stronger.
I would try to get good transport and power supply conditioner/filter (I use it in spite of Benchmark DAC1).
Benchmark solves the problem while Bel Canto DAC3 is a little better (according to Stereophile review).
To minimize it I would use very well shielded digital cable (double or even triple shielded) that is about 1.5m long very well characteristic impedance matched. Shorter cable might bring reflection back on the same transition that caused it affecting threshold point (since limited slewrate) but the longer one will make reflections stronger.
I would try to get good transport and power supply conditioner/filter (I use it in spite of Benchmark DAC1).
Benchmark solves the problem while Bel Canto DAC3 is a little better (according to Stereophile review).