Robert A. Heinlein's hifi? I could clock-sync our upstairs Soundaware D300Ref to Terra's 22.5792MHz clock signal, downstairs to Gaia at 45.1584MHz. Upstairs would for the first time introduce an external masterclock. That's because its now legacy Terminator DAC lacks the clock outputs of the II and Plus and has no clock input either. Downstairs would replace the Plus as already uprated clock generator. That suggested less improvement headroom if any. As always, sonic feedback is allowed only if we've tried it. All else is theorizing, trolling or trading in fake news. In my 2nd scenario, even if Terra's OCXO were equal to the Terminator Plus, it has a far bigger power supply dedicated to just clocks, naught else. Shouldn't that make a difference? If I heard one, how could I know whether it was the clocks, power supply or both? Here's what Alvin Chee, global Denafrips sales manager, had to say: "I personally think that the Gaia/TPlus clock sync will be a class above." Regardless, Terra still answers the question of all Aries, Pontus and Venus owners on how to get in on the Terminator II/Plus reclocking: with it plus a €435 Iris or one of Denafrips' bigger DDC. But the DAC needn't be a Denafrips. It can be anyone's. Only the DDC is limited to those which accept Terra's clock-rate options. In Apple speak that's called controlling your infrastructure. It's virtually closed to outsiders.

Thinking readers made another connection. Directly, Terra won't work with the top two Terminators. This kit only flaunts clock outs not ins so isn't made for each other. To play today's game with a Terminator, we'd still need a Denafrips, Soundaware or copasetic D/D converter. Slaved to the Terra clock generator, their digital outputs now supply our DAC with actual audio signal whilst their digital inputs turn them into a digital hub for alternate digital sources. In our system, Gaia gets music signal via USB from the iMac's Audirvana player, via AES/EBU from the Avatar CD transport and via coax from a Soundaware D100Pro SD transport. The single connection from Gaia to Terminator Plus is via I²S over HDMI. As an aside, I would try adding Terra to our Plus via Gaia. My primary focus would simply be on standard DACs via Soundaware D300Ref below; then reclocking Avatar with Terra. After all, Gaia + T.Plus do a swell reclocking and buffering job with just two boxes. Where's the fun in complicating things with three? Unless it made at least a virtual meter of sonic difference, I'd not bother with a few millimeters. The Avatar/Terra experiment would stand in for all components which take a 44.1kHz clock input.

Finally, if you just do RedBook and derivative sample rates, one 75Ω BNC-terminated clock cable suffices. You won't do video-based rates. If however you spin/stream 24-bit 48/96/192kHz tracks, they'll distort unless you also run the second clock cable in their support. A 1-meter such cable from Sommer will set you back €16.40 at Thomann [right]. You needn't part with $3'299/m for a Nordost Valhalla 2 (or, yikes, twice that for two). When cable becomes thrice or more costly than the component it connects, eyeballs and heads tend to roll. For our high-speed clock-signal application, the most important spec will be constant 75Ω impedance to avoid signal reflections and the temporal blur they can create. At least that's what Chris Sommovigo of Black Cat Cable told me who should know. And now we'll leave behind our familiar terra firma to explore the strange terra incognita of home-audio masterclocks. Houston, commence your countdown.

Terraforming. In SciFi novels, it preps an alien planet for human arrival. For the Denafrips Terra, I just needed a spare 75Ω digital cable from my stack of Sommovigo options.

This signal path was .flac/.aiff files on SD card in a Soundaware D300Ref ⇒ Denafrips Terminator DAC ⇒ icOn 4Pro  ⇒ 80Hz high pass ⇒ Crayon Audio CFA-1.2 ⇒ sound|kaos Vox3awf, 80Hz low-pass ⇒ Dynaudio 18S subwoofer ⇐ passive power delivery by Furutech NCF, resonance attenuation by HifiStay.

Terra's MCKL1 set to 22.5792MHz ⇒ Soundaware's equivalent input. To compare with/without Terra, I only had to press the D300Ref's rear-panel switch. That enables or disables its clock inputs. What could be easier?

In case you're not hip to SD card playback, in our household it beats CD spinning by just a whisker but standard USB streaming off an iMac with Audirvana by a full lock of hair. To have USB pull up equal must first run it through this Soundaware or equivalent. That becomes a so-called USB bridge. It isolates DAC from computer and reclocks/dejitters the music data. This particular deck exploits an ultra-cap power supply as virtual battery so the AC could disconnect and sound continue as long as the high-farad capacitors maintain sufficient charge to power the circuit. Whether I could hear Terra's influence on this setup would depend on whether its clocks were more precise than Soundaware's. Or would femto this and nano that meet the limits of my hearing? Just because a scope can measure it doesn't always mean we can hear it especially with many zeroes behind the decimal before anything happens. With femto clocks, we deal with 10 to the 15th power—from the Danish word femten for fifteen—where a femtosecond is a quadrillionth of a second so fifteen zeroes. Meanwhile clock accuracy as metronomic deviation is measured in parts per billion. Remember 1997's romantic comedy My Best Friend's Wedding? Julianne evaluates herself as pond scum – "well, lower actually. I'm like the fungus that feeds on the pond scum". To which Michael replied, "lower. The pus that infects the mucus that cruds up the fungus that feeds on the pond scum."

Unlike platonic love, does planktonic hifi love matter? How low can we really hear? How small still factors before imagination takes over and we listen to SciFi? Here's an interesting tidbit from a pending Bel Canto review. Their head designer John Stronczer favors multiple stages of jitter reduction so his flagship €37'900 Black DAC/processor runs three cascoded reclockers, his €13'500 EX version two. In the far costlier high-end than today, that makes for some evidence that serialized clocking stations can have measurable/audible benefits. It speaks to the obvious question that since all DACs already reclock the incoming signal, why should we repeat the same step? I'm too technically ignorant to know whether it's allowable but do wonder whether in layman's terms, one might view this like multi-stage water purification where not a single filter but multiple filters in series get the incoming water to ever higher purity? That simile goes for USB/Ethernet bridges which transfer digital audio signal. Today, our masterclock only takes control over the crystal oscillator inside a DAC or digital transport. The masterclock does not conduct any actual audio signal. It just becomes the new metronome which clocks the digital samples. TikTok.

Just then Mario CanEver's ZeroUno PureDAC arrived. That runs a 'HyperStream' Sabre 9018 with ESS' patented on-chip jitter-reduction circuit. It'd become another Terra test station. But first the standard Terminator. For the triangulators among us, there's this from former contributor Michael Lavorgna now helming Twittering Machines: "If you want the best sound you can get from the Pontus II, I'd say that means you'll want to add a Denafrips DDC and go I²S all the way. With a combined price of roughly $2'200, I know of no other digital-to-analog converter solution that comes anywhere near the Iris II/Pontus II combo in terms of conveying music with all of its attendant power and beauty intact." It's another fine user endorsement for including a good DDC aka bridge in the digital signal path. Would masterclocking be its 'i' dot? The Lavorgna quote also confirms my repeat findings that at least in the Denafrips universe, I²S over HDMI is the preferred pipeline for sending digital audio signal to their DACs followed by AES/EBU, then BNC, then RCA.