Canary yellow in the mines. Weng's single-mode wires were that cheerful color; and mine for the duration. This twin-lead fiber is very thin and terminates in two pins per end. Those slip into long aluminium optical converter cartridges which themselves seat in the SPF ports. This photo shows the lot including Hubert+Suhner 50Ω clock leads which were black 'n' blue without any Irish pub brawl. My first lesson in fiber-optic networking was nigh. ABC so angst or arousal, boffo or bollocks, crikey or cringe? Our communal obsession with better sound can go either way. My first task was comparing our SW-8 to the SW-6 in plain-Jane copper mode whilst exploiting the latter's isolated RJ45 port for the 1m final CAT8a stretch into our music iMac. Would that alone register? Given temperature-controlled oscillators aboard these decks, 24 hours of on-time prior to any listening was a good call. Neither machine would power down. A/B would simply involve swapping incoming and outgoing cables whilst pausing Audirvana Studio's player software and letting the iMac's network card relatch.

Here's that setup. In the underlay, note the big digital isolation transformer IC behind the solitary RJ45. If I heard any difference, would that part play prime suspect? Here is Alvin Chee's SW-6 video.

After customs inspected the SW-10 shipment, it arrived sans second Finisar optical cartridge. Weng rectified the loss by having Mouser in Czechia send me a new one. That duly arrived in an anti-static pouch. Now I worked through different permutations of single, dual and triple cascaded switches; copper or copper/fibre wiring; two switches clock-synced to the OCK-2 or the SW-10 clock-syncing the SW-6; inserting SOtM inline LAN isolators between switches; and more. Given the sheer multitude of scenarios, I'll spare you a blow-by-blow litany to skip straight to my key takeaways.

But first, a photo of one of the hookups.

Incoming Ethernet copper isolated by SOtM filter; outgoing fibre-optics to second switch; short copper going from its isolated RJ45 port to iMac's network card; both switches reclocked by OCK-2.

Kamasutra position #61?

What's essential for pleasure?

Where do exhibitionist selfie gymnastics begin?

How old are you?

If serialized galvanic isolation by inline optocouplers or micro isolation trafos is audible, we must conclude that no such solution is 100% effective on its own. Might we now see the audiophile's attack on ultrasonic LAN noise like a countertop water purifier? A really good one will combine 4-or-more different filtration types to run utility tap water through cascaded treatment stages.

It's the only notion which appeases my wondering mind as to why the above box count and cable salad should make any difference over jacking the incoming CAT8a wire from the router straight into the iMac's RJ45 port. That direct scenario is shown with two red circles. Overlay them. Hello bit-perfect signal. Bypass all the rest of it. Easy does it. But does it, really?