A proper analysis of RF energy effects in a digital audio system requires a measurement device to convert time-domain signal into frequency-domain. That is, to display digital waveforms and pulses as a graph of signal energy (dB) versus frequency. I use a Signal Hound BB60C to scan a range of frequencies from a cable (direct input) or transmitted (with a directional antenna). In the chart below, a set of direct signals are measured: the time-domain is taken from my oscilloscope and the frequency-domain from my BB60C.
The time-domain images are an instantaneous snapshot of the logical levels of the digital signal. The frequency-domain images are averaged over a time-interval to indicate the energy at a given frequency. You'll notice that even though the time-domain waveforms are very similar, the frequency-domain graphs are not. This difference is due to the high frequency ripple on the time-domain waveform and the phase shift (jitter) over time - only visible in the frequency-domain.
Also recorded were measurements of radiated RF energy from various digital components - sources, upscalers, power supplies. This involved connecting my BB60C to a wideband directional antenna and 'pointing' it at the device under test. Tests with varying separation distance (up to 10 meters) were performed inside a RF isolation tent or at a remote location to avoid excess impact from environmental RF noise.
All measurements were combined with listening tests and while those are somewhat subjective I assured myself that the test conditions were non varying and the path from DAC to my ear was consistent and transparent as possible. I have 'golden ears' but I also repeated tests over time and used associates to confirm my impressions.
Over several months and dozens of tests, I have collected the evidence for my bold assertion that RF spectrum energy reaching a DAC* has a causal impact on sound quality. Its not the only causal impact for an audiophile to address but certainly significant. For certain: a DAC* always sounds better when in the presence of less RF energy: via signal, power or radiated through free-air.
I am working to provide solutions for this phenomenon - known in the industry as 'EM/RF Coupling'. The coupling mechanisms are: common-mode impedance, capacitive, inductive and radiated. Some aspects are well understood, for example: power and optical signal isolation removes common-mode induced currents. However, some aspects are still shrouded in mystery that I am working to better understand. This is part 1 of several posts on this topic.
*Only Chord Electronics DACs were tested so far. More tests on more DACs is coming.