I find it curious that so many separate components are used (that presumably aren't cheap, as they target a low-volume, pretty niche industry).<p>It seems to me that most of these can be replaced by any computer with an analog audio output (and since this is Jeff Geerling we're talking about, it can be a Raspberry PI)?<p>Maybe put two and use the "silence detector" box to automatically switch between them in case one goes down?<p>In fact, I wonder if a custom board can be designed with 2 separate stages (each powered by a compute module and DAC) and some circuitry to do the silence detection and switching, with some GPIO lines going back to each to inform them of which one is being the chosen one for the final audio output. This would effectively condense this whole thing into a single box. But not being a broadcast engineer, I'm just talking out of my ass here, maybe there's a good reason they're doing things they way they do.