This review page is supported in part by the sponsors whose ad banners are displayed below

Heros. Having established already that loudspeakers most ably served by the SIT1s' handful of watts had the even simpler circuit and unique transistor of Nelson Pass show off an attractive advantage in transparency and directness, a more meaningful context meant thirstier loads. It also meant harnessing ModWright's KWA-100SE as comparator instead. With an equivalent 100-watt power spec and Mosfets in push/pull, the ModWright is a cool-running class A/B affair. In use the narrow deep and heavily ribbed Heros cases got very toasty to properly track the always-on status of their power Mosfets. At $3.995 the KWA-100SE naturally runs far cooler too than the €24.000/pr Heros twins. Would their blistering class A bias make for the decisive difference?

In again it was with the €12.000 Aries Cerat Stentor sealed 10-inch 3-way with its small Fostex midrange and Raal ribbon in a stacked Ply enclosure with internal 'anechoic chamber' walls and precision 30-part crossover. Similar to the Maximinus, my inner cynic punk again wasn't feeling very lucky with this example of unabashedly luxury-priced gear. Whilst the sound was very good, so it was with the ModWright. I didn't feel transported to a much higher plateau as the 6 times higher sticker should have. The somewhat more suave treble and minorly smoother midrange failed to span the vast fiscal gap with anything approaching a solid bridge for confident footing. Instead I saw a rickety Indiana Jones rope suspension with missing planks.

To avoid pixel waste, I pressed on straight for the €63.000 Thrax system. This first and foremost meant removing the Nagra Jazz from the equation. The Metrum Hex followed mostly out of courtesy and because the Maximinus output voltage was lower to cater to the Dionysos volume control..

The system with its designed-in dovetailing of qualities now came close to the Jazz/SIT1 pairing with one vital advantage. It offered ten times the power. Listeners with low-power speakers already know that small amps tend to have a leg or two up over their burlier brethren. There's some correlation with a smaller rev-happy engine versus a big V8 block. As long as you're not towing a horse trailer or racing up a steep mountain pass, the smaller one will be more responsive in inner-city traffic where you do most your driving. Whilst the Heros amps by themselves weren't as lucid and tonally sophisticated as the SIT monos, preceding them with the fully activated Thrax 'passive' fed them with virtual high-octane racing fuel. And that meant small-amp finesse with big-amp power.


To begin with, the Heros didn't suffer the darker slower lazier voicing Michael Fremer had initially encountered with the Greek $36.000 Ypsilon Aelius hybrid monos (unlike the Thrax those parallel their output devices six times and don't use an output transformer though an interstage transformer does handle phase splitting). Stereophile's loaners had been fitted with fatter EH 6C45PiEH instead of the originally spec'd very linear C3g Siemens pentode. Artarski's favored 5687 is very neutral. There's no 'valve drag' to overcome. Be it their P/P vs SE operation, two- versus single-stage circuit or Mosfets vs. triode-curve vertical power Jfets, the slightly mellow but very awake Heros simply weren't as transparent, illuminated and direct as the SITs. That's where the previously described denuding/accelerating action of the Dionysos narrowed the gap (and one imagines by strategic design).


Though I haven't heard Jacob George's Rethm amplifiers—one is all-tube, one hybrid—I imagine that as two candid tube lovers his and Rumen's rationale for their respective hybrids is the same. It's to give comparable if perhaps not as extreme sonics with a lot more real-world load invariance and power. That's a very practical response to the common silliness of committing to speakers which need powerful amps. None of my speakers are beyond the SITs' grasp though the Stentor did show superior bass control with the Heros. The Stentor thus was a transitional example in just that regard. For most normal living rooms, smaller speakers good to ~55Hz augmented by a superior active subwoofer for the first octave are smartest. This produces far more extended far less room-dependent bass than any reasonable passive tower speaker could produce.


But audiophile snobbism outlaws subwoofers. They've become unfit for serious fi. Thus companies into highly refined sound like Thrax who wish to appeal to a broader audience must cater to their less efficient speakers and their more bullish impedance plots. They must model the sound from their least powerful and simplest but most sophisticated amp and migrate as much as possible into higher power ratings. I reckon that's how the Heros and Spartacus monos relate.

Of the three Thrax models I thought the oldest (perhaps most mature?) the highlight. But there was a serious shadow. Think inexplicably limited step ratio of the transformer volume control. It hit its stops far too soon with 0.5V amplifier input sensitivities mated to standard 2V sources and perfectly normal 88dB speakers. That said, the sonic purity, lucidity and resultant virtual hardwiring of the listener to his or her tunes had the Dionysos play at the very top of the preamp game as I'm acquainted with it. I/o functionality and appearance were top notch too. Were Thrax to adopt Bent Audio standards and progress their transformer attenuation in no more than 1dB steps with some 50+ values below unity gain where they're vital, the Dionysos really would be a dream machine. As it was I found myself dreaming still.

Against competitors like Metrum's €3.000 Hex or AURALiC's €3.300 Vega, the €24.000 Maximinus DAC felt irrationally priced. Worse, at present it showed zero sonic advantages to redress such a garish imbalance. As a brand-new model with pending upgrades and modular options coming, final judgment ought to remain reserved. Enter the Heros monos. Here I thought of Nagra's MSA. The flawlessly billet-carved heatsink capstone eats up 40% of its raw build cost. It looks marvelous but does nada for sonics. The Heros casing is another triumph of precision machining. It too gobbles up coin. Obviously there's a market for it. I simply couldn't get excited. I found myself constantly repackaging these macho heroes to get the price in line with sonic competitors. Bad dog!


Which gets us to my final assessment effort, theoretical concept vs. audible deliverance. Some companies approach ours for sound. None equal us on silence. I tried hard to correlate that claim and its costly consequence to float all circuits between transformers on either end with what I heard. The theory really appealed. It read spot on. In practice I couldn't hear the advantage. Worse, a very obvious ground loop when my Bent Audio passive replaced the Dionysos undermined the claim that complete transformer coupling would categorically eliminate ground loops. When at the very end of my auditions one Heros went inexplicably mute though its tube filaments were on as was the green power light, my inner cynic really felt vindicated. Here I should add that Marja & Henk previously accepted Dionysos and Spartacus loaners but cancelled their review due to technical issues and noise. Had Rumen only stuck to just dispatching the preamp as solicited, my findings would have run on a lot more enthusiastic steam. Instead the assignment became mostly a demonstrator for my personal ineptitude at getting on and along with the luxury pricing program at hand. Where was Jonathan Valin when one needed him?
 
Quality of packing: An example to all competitors.
Reusability of packing: Indefinitely.
Ease of unpacking/repacking: A cinch for the source components, less so with the vertically ensconced amps.
Condition of component received: Externally flawless but my iMac couldn't see the USB input of the Maximinus. I had to default to my own SOtM USB bridge. At the end of the audition, one of the amps stopped producing sound even though its tube filaments were lit and the green power LED on.
Completeness of delivery: No owner's manual for the DAC yet.
Human interactions: Good.
Pricing: Irrationally expensive for the DAC and amps.
Final comments & suggestions: The preamp's transformer-tapped volume control is optimized for use with its stable mates, i.e. a 1V-out source and 4V-in amplifiers. With standard 2V sources (forget 4V+ balanced) and amps of 0.5V input sensitivity, even 88dB speakers will get too loud far too quickly, reaching room levels over less than 10 clicks on the dial. If Thrax mean to market the Dionysos as a universal machine as seems to be the case, they really must have many more steps below unity gain like Bent Audio offer for their OEM autoformer controls.
Thrax Audio website