My pragmatic analysis of causal effects as described here clearly shows just how far you need to go to extract maximal sound quality from a digital system. A DACs sensitivity to RF noise is considerable and Audiowise products mitigate the impact of conducted and emitted RF from digital components. But it's quite remarkable that any connected cable is also a significant conduit for RF.
While working to improve the sound from of my Focal SM9 monitors, I was having difficulty getting them to be as magically transparent as my directly driven Voxativ speakers. I am talking about that wonderful sense of 'snap into focus' where the artist and recording venue become present. I tried improved XLR interconnects and got better timbre but no magic. Was it AC noise? I moved from AC mains to my big Goal Zero batteries and this was marginally better but did not have the same transient speed and low-end resolution. Was it power related? I changed from the stock to mid-level audiophile power cords and this closed the gap. I then tried much more expensive power cords and got even closer. This was a clue ...but why?
The SM9 has an outstanding power supply with huge reserves of power: more than 2,600 uF at line-level and over 17,000 uF per rail at high-level. According to Focal, this guarantees identical signal reproduction at any SPL and irrespective of AC power quality. Yet if the SM9 is so competently engineered, how can a length of power cable affect the sound? Does the cable 'really' matter? I tried a simple experiment: I moved my batteries to an elevated position immediately behind each SM9 and built a test connection using close-coupled NEMA and IEC ends. If somehow a 'better' power cord mattered for sound quality then what about 'no' power cord.
To my surprise, this experiment found the missing magic! I now heard the same incredible sound from my SM9 amplified monitors as from my directly driven Voxativs. What? How could a DIY $20 power connection provide better sound quality than a $3500 audiophile power cord? You can pause now and reflect on an answer ...or let me tell you. This had nothing to do with 'better' power; the longer audiophile power cord is an antenna; the ultra-short power connection ...not so much. That's it. The SM9 reliably amplified the input signal regardless of power cord. It was the signal from the DAC that was being changed by the RF noise.
Here's what's happening: any length of wire acts like an antenna. DACs exist in a mess of electromagnetic (RF & EM) radiation from nearby electronics, switching power supplies and the environment. All this noise can find its way into the DAC via any tethered metal cable: power inputs (AC or DC) or any galvanic inputs or outputs and connected components. The efficiency of reception or transmission depends on numerous factors but a longer length of cable is generally a better antenna. In my test case, the 2 meter SM9 power cord absorbed RF energy sufficient to find its way to the DAC via any galvanic connection.
Incredulous? Still believe that power cords make amplifiers sound 'better'; that pure copper or thicker gauge or cryogenic treatment delivers 'better' electricity to the amplifier? I wont generalize but I will go on record to say that, at least for high end DACs driving good amplifiers, this is patently false. Power cable metallurgy, design and construction matter only in their function as antenna - emitting less or more of this RF energy to manifest inside a DAC where it all matters for sound quality. Thanks to our ear/brain's extreme sensitivity to minute variations in output sample timing and an elevated noise floor, even impossibly small levels of RF noise reduce transparency.
So, what can be done to solve this problem? How can we stop cables from being antennae? (apart from making them as short as my test). I'll have that information (and more) in my next blog.