The replacement unit sported, cough, UK socketry. Apparently actual (Ireland) versus virtual (US) location just proved too confusing to be accommodated in fewer than three attempts. Rather than request another substitution, I ordered up more of the Swiss Skross UK-to-universal power adapters we already had throughout the house. Whilst arguably not ideal, it was the most expedient fix. During my Kii Audio Three review—fully active speakers supplied with long UK-terminated power cords—I previewed the Guardian in both our media room and big system. These speakers take a digital signal. They only need a transport by way of PC/Mac, SD card reader, DVD player, server, DAP and such. With 6 x 250-watt mono nCore class D amps per channel plus DSP serious enough to create significant latency unless set to low latency/lip-sync mode, the Kii3's switch-mode power supplies and intense digital high-speed signal processing aka computing indicated a perfect testing ground for the Gordian.

And so it was. Plugged into its high-current outlets, these hi-tech speakers performed better via the Gordian than passive Furutech RTP-6. Improvements were in cleaner smoother treble, greater calm, superior low-level resolution during subdued sessions and fuller tone. By comparison, the Furutech power feed sounded grainier, flatter on timbres and not quite as convincing at whisper levels. With the Gordian not the focus of these audition rounds, I didn't check on read-out values, only sonics. Class D is known to respond keenly to power delivery upgrades, be it better power cords or line filtering. The Kii Three were no exception. Set to 'auto' to configure itself, the Gordian made a demonstrable difference in both systems (in the setup below, you can see it to the right of the corner palm).

Once the Kii left, the media room reverted to the German Physik HRS-120. Now I played the Gordian numbers. Display on, I watched whether adding components dirtied the AC with more RF/EMI. Here I didn't see any change. Whether one component connected or all, input EMI fluctuated from 19.5mV to 28.2mV regardless (remember, this measured the AC status of our entire household loom well beyond just one room). The value which did change a lot was total harmonic distortion for current. That shot up to 93% when the entire stack went live. Checking on the DC-coupled Goldmund Job 225, DC line offset read -0.003V with nothing powered up, 1.91% with just the Job added. And so forth. Unless you're a tech, you should be dumbfounded whilst attempting to correlate any flickering numbers with aural predictions. What do they signify sonically? A personal surprise was invariability of RF/EMI values. It seemed to combat the notion that switch-mode power supplies in televisions and DVD players aren't smooth but filthy operators. Would the most heinous criminal in their league, our music iMac, make any dent? Purists love to diss computers as noise incarnate. Plugging the Gordian into the wall outlet which powers the amps, iMac in our usual Vibex DC/AC filter fed from an outlet farther up the wall, the input EMI of 35.9-38.1mV didn't shift whether the Mac was on/off. Was the Spanish conditioning too effective? Was Apple cleaner than expected? Next I plugged the iMac directly into the wall to compare input EMI on the Gordian. Zero diff. Was that why I'd never had demonstrable returns from costly audiophile server loaners? It's when I jacked our Nord Acoustics nCore 500 monos into the Gordian that suddenly its figures danced. Katchinnnng! Now input EMI shot up to 120mV: four times higher! Firing the amps up and down kept flipping these values for clear cause and effect. This suggested a direct correlation with the earlier 3kW class D speakers.

SD card reader --> LessLoss Echo's End --> Wyred4Sound STP-SE Stage II --> Linnenberg Adagio --> Audio Physic Codex

Sonics. For starters I toggled two scenarios: Linnenberg Allegro monos powered via Gordian or our customary Vibex Three 11-R which in 2013 was €1'990 i.e. 25% more. The Spaniard made the monos stage deeper, taller and wider. It played more walk-in separated and sorted. Tonality was richer. By contrast, nothing was wrong with the Greek's presentation. Dimensionally and on small-signal transparency especially plus timbre, it was simply overshadowed. It's why I don't pursue AC conditioner reviews. Our Vibex Granada/Alhambra on the source stack, Three 11-R on amps and a Vibex Two-1R DC filter on the Zu Submission subwoofer have already sent numerous challengers packing. It's why Vibex was our final choice. When makers insist to dispatch something regardless, they're either very confident or failed to study our associated gear list. Given this outcome, I could predict that moving the Gordian into the source stack to replace Vibex's 2-piece flagship kit would fare likewise. It did. To rig up an A/B that would play to rather than against it—after all, €3'600 of Vibex should outperform €1'400 of Lab12—I replaced the amp-end Spaniards with a passive Furutech e-TP8. I then salted the battle field with the 250-watt nCore monos. Having been measured as noise leaks, wouldn't the Gordian's active compensation filtering dominate Furutech's passivity?

Incidentally, for those with less demanding systems, Lab12 just introduced the simpler Noyra filter.