As We Hear It: An audiophile comments on the Taiko Audio computer. The Mark Jenkins / Antipodes Audio Darko interview. On MoFi's One-Step DSD. And optical HDMI.

German "ear horns" circa 1917.

It's great to look inside my mail bag and post comments or questions here from readers once awhile. I must say that the clarity of thought and insight from readers have been inspiring over the years.

With permission, here's a comment from an audiophile in Europe regarding his direct experience with the Taiko Audio SGM Extreme "audiophile" computer server/streamer.

Since this reader has quite a unique high end system which can be easily traced, this message has been anonymized in a number of ways. Without further ado, a comment from SonicDeepThroat (SonicDT):

Received: July 2022 

Hi Archimago,

I thought I'd drop you a line as I found your blog post about the Taiko SGM Extreme very interesting. I auditioned this machine a few days ago and thought I would share my conclusions with you. (I tried to leave a comment on the relevant blog post, but Google wouldn't let me sign in to do so.)

The test was conducted using my <unique, high end> speakers (which are popular with objectivists such as the Audio Science Review denizens, as they measure extremely well and contain a lot of bang for the buck, whereas the Taiko is popular with the subjectivists on the What's Best Forum and not at all popular in the Audio Science Review). The <speakers> allow multiple sources to be connected simultaneously, with switching being undertaken on the fly, so I connected the Taiko via USB-Audio and my present streaming transport – a modest Auralic Aries Femto – via S/PDIF and then played the same song on both, frequently switching between the two to eliminate the need for audio memory.

The Taiko delivered a clear and consistent improvement on everything that was played through it: more presence and impact, especially on the vocals.

However, the improvement was pretty small. It certainly wasn't worth anywhere near the £30,000 it would have cost to obtain: the Taiko was given every advantage, as the dealer had brought Stillpoints, a £2,000 power cable and an "audiophile" USB cable costing about the same, yet still the improvement was pretty modest.

Thus, whilst I don't think the SGM Extreme is a "scam" or "snake oil", I agree that's completely over-spec'd for its purpose and that all its expensive parts don't really translate into a significant jump in performance. The dealer agreed that the Taiko wasn't doing much, but stated that it must be my speakers as every other Taiko demonstration had produced a step change in the sound, citing the case of one customer who started sobbing as he was so overwhelmed by the audio improvement.

That makes me wonder whether the answer to the Taiko's acclaim – including the breathless 600-page thread on What's Best Forum – results from some sort of mass hysteria: it would be very difficult to tell a sobbing man that the astonishing improvement he thinks he's hearing is actually quite minimal, particularly when you're a dealer wanting to make a sale.

My perception of the Taiko is that it combines incredible looks, both internally and externally (it's one of the most beautifully finished pieces of audio equipment I have ever seen) with great marketing and customer service that makes people feel really, really special (this was obvious from the few hours the unit was here, with constant calls and support from the Taiko team), causing people to overestimate its performance.

The irony is that the SGM Extreme's main competitor, the Pink Faun 2.16 Ultra [Ed. price starting at £23,500, or ~US$28,500], which I've also auditioned, is decidedly the better sounding transport, but there are no 600-page threads about that and anyone entering the Taiko thread and making that claim would probably be crucified. Why? Less marketing, for a start. More fundamentally, the 2.16 Ultra is ugly (or at best utilitarian) both inside and out.

That's not to say I consider the 2.16 worth the best part of £30,000 either. In my home test – which was wired up the other way, with the Ultra using S/PDIF and the Auralic Aries USB-Audio – the 2.16 excelled on audiophile stuff (piano decays were vastly more realistic, acoustic instruments sounded much more convincing) but was pretty indifferent on rock music, and I mostly listen to rock. Worse, the Euphony software was a pain in the backside, which didn't endear the Ultra to me.

If somebody produced a streaming transport combining the best of both units (the Taiko's looks and consistent small sound upgrade, the 2.16's larger upgrade on acoustic stuff, backed by reliable software) I just might consider buying it. As it is, I'm not proceeding with either, though if you put a gun to my head I'd buy the Ultra over the Taiko. Other units I've tested produced mixed results: a definite upgrade with an Auralic Aries G2.1 and no discernible difference whatever with an Aurender N20 or a dCS Network Bridge.

With all best wishes, SonicDeepThroat.

Thanks for the detailed comment SonicDT.

Clearly, you have spent much time, and other resources into extracting the best sounding audio possible. While the identity of the speakers have been anonymized, I have heard them at a local high-end dealer and indeed I believe these are one of the few speakers that both objectivists and subjectivists can agree will produce amazing sound quality using powerful modern technologies and design.

It is wonderful to hear from someone with the experience of having connected with the Taiko/dealership directly to arrange for a home visit. Often, one of the excuses from purely subjective audiophiles is that "your system isn't good enough to hear the difference" which would not apply here.

I was blown away by the almost-600-page thread on What's Best Forum. I see no reason to go through all the posts since most of it appears to be just speculations and opinions. I found the last few pages informative enough I think. Fascinating for example the comment presumably by Emile Bok (founder of Taiko Audio) about USB cable lengths, saying that the Intona 1m cable sounds significantly better than the shorter variants; denizens there also chimed in about sound quality with 1.5m USB lengths perhaps being optimal. I don't understand, what's the point of spending €24k+ on a computer that is so frail such that sound quality may become suboptimal due to slight variation in USB cable lengths if we are to take these claims at face value!?

(BTW, here's the Intona USB cable page, as a sign of obfuscation, notice how Intona doesn't bother having units or values on the "Cable length" for many of their graphs! Why is that, Intona? Is this good engineering practice, or are you guys just fishing for a certain type of customer?)

I see that the Taiko computer has grabbed the attention of YouTubers like this:

Great discussion at 9:30 on the "Marketing Material" as he considers the ideas and advertising claims around this product. I think it's easy sometimes to take for granted just how bizarre and cult-like certain ideas in audiophilia can be, so it's good to have videos like this reminding us from the perspective of a general tech blogger.

There's obviously a lot of hype with claims about audio performance improved using the dual-CPU, plus FUD to suggest that PCIe is needed to lower noise in the Taiko website. That part about "much lower system noise overall, giving you black backgrounds, huge space rendition and brings an ease to the musicians (sic) performance, only matched by the very best vinyl and tape playback systems" is plainly a laughable lie - one just has to go to an audio show and listen to vinyl playback compared to digital on the same system to recognize how much worse the LP noise floor is compared to digital on top-end gear! (Easily measurable of course.)

IMO, anyone who is willing to use a lie like this (tantamount to telling someone black is white in order to make a sale) must be treated with "Extreme" suspicion. I have no problem with people preferring vinyl over digital - that's certainly a subjectively legitimate preference. But the way Taiko uses claims about "noise" or "black backgrounds" with LP playback compared to digital is simply nonsense. As usual, they provide no verification of such benefits.

I like SonicDT's query about whether those 600 pages on the What's Best Forum represent something like "mass hysteria". Or maybe in contemporary language: "social contagion". These days with social media, we have seen the rise of things like "TicTok Tics" as an example of the psychological impact among certain members of society. Obviously, audiophiles are only a very small subgroup of on-average older individuals so hopefully our psychological maturity can be a bit more sophisticated. Nonetheless, we are all prone to the usual mechanisms of bias and the power of marketing; especially in luxury goods often presented with an aura of scarcity, or supposed scientific advancement, as being something "special".

Whether otherwise "bit perfect" devices will sound different, I will leave for each of you to determine (I have my opinion that in the 21st Century, better digital devices should act as "Bits Are Bits"). Regardless, I agree that even if there are audible differences (this could be related to specific equipment since digital inputs themselves could impart their own noise or timing characteristics), as SonicDT observed, "the improvement was pretty small". And from there, each of us can then gauge the asking price and the value one gets from such a product. The idea that switching an otherwise bit-perfect computer could result in such qualitative differences as to induce an adult to tears and sobs sounds apocryphal to me.

Many times on chat forums, I hear people talk about "millionaire audiophiles" throwing money around to buy all kinds of exotic stuff like this computer as if basically no big deal. In my experience with "normal" multimillionaires and even hundreds-of-millionaires (don't know any billionaire audiophiles yet), they're nothing like the rich, clownish celebrities we see on TV and on tabloids. Real millionaires know the value of money - that's how they get rich - and I suspect most multi-millionaire audiophiles who have even the vaguest notion of computer technology would find the idea of hearing a "big" difference using this computer to be rather unlikely, and fawning over this, rather foolish.

As I've discussed in the past, I think it's important to find respectability as audiophiles (as opposed to ending up as audiophools). The only way to "rehabilitate" this hobby is to speak truth against questionable claims and just be honest with other hobbyists when it's quite obvious that the value proposition is extremely poor - even if we don't fully use words like "scam" or "snake oil". [I would at best consider the Taiko computer a type of Class C "Audio Jewelry".]

--------------------

In other related news, here's an example of an interview in Summer 2022 that's trying really hard to spread FUD (Fear, Uncertainty, and Doubt) - this plodding discussion between Mark Jenkins of Antipodes Audio and John Darko (notice also the Stereophile links from the mid-1990's[!!!] on that page used as supportive references). I suppose it comes across as an awkward interview because it sounds like Mr. Jenkins is searching for words to spin the non-issues he's trying to hype about his "audiophile" computers.

Again, it's the same old thing about how hard it is to get precise timing out of a computer, they even seem to be avoiding the dreaded "j"-word (jitter) boogeyman until about 55:30 (maybe they're recognizing that jitter in DACs isn't all that audible?). All those words in this conversation, but not a shred of evidence provided; not a single example of what kind of improved timing they're talking about compared to a cheap Raspberry Pi streamer or standard computer. No quantification that "noise" is an issue even though they spend quite a bit of time on this as well. Then there's all this silliness about how the digital transport is able to vaguely make listeners "get up and dance" (around 1:14:00) using anecdotes (something about "slowness" affected by power supplies, really?). IMO, talking like this with zero attempt at correlating any of these claims with reality testing reflects poorly of Antipodes and their marketing of expensive computers that cost >US$2k and even up to $17k+.

The only tech-like stuff I've seen from Antipodes consists basically of unsubstantiated cartoons. I can't even find basic concrete information on these computers - for example, what CPU does the Antipodes K41 Music Server use given it's priced at $10+k? Is it fast? Is it slow? Does it produce much heat?

Darko still believes just because thousands of people claim something (59:00), that it must be true and "all of them" cannot be affected by the placebo effect. I bet Darko himself is fooled by the placebo effect in many of his "impressions" ;-). Folks, there are thousands if not millions of people in aggregate who believe they can hear the voice of some deity or claim that some pseudoscience medical product has miraculous effects, so what?

If a million people believe in the power of healing crystals, is that proof that I also should start purchasing magic crystals as well? As for Mr. Jenkins - yes, of course you can run a business based on the placebo effect - just open your eyes and critically examine the examples all around.

It's good that Darko asked about measurements at 1:19:00. Notice how, like a politician, Jenkins doesn't answer the question directly and doesn't address the fact that these digital audiophile computer companies do not publish actual data on what they're "improving" whether it be jitter or noise. In 1:26:40 or so, Jenkins claims that "if there's something gross going on that shows up in measurements, you've probably already heard that". This is not true as demonstrated by numerous measurements. We can see gross anomalies in jitter yet it's not very audible, THD can be easily demonstrably high, yet listeners may only perceive subtle variations. We've arrived at the point in history where measurements can tell us information far beyond the resolution of what sounds "good enough" for human hearing.

Again, let me reiterate that there's nothing wrong with buying a nice looking computer; so too there's nothing wrong with buying an expensive handbag, wristwatch or car (as discussed recently). Just be careful with putting too much faith in the unsubstantiated, sometimes even near-spiritual tenets (Jenkins in fact conjures up the word "magic" at around 1:27:20) promoting stuff like this to audiophiles. Beliefs in false claims provide financial gains to those who make questionable products and "journalists" who sell this stuff rather than rewarding companies with truly great R&D that can push performance and value forward for the consumer.

I personally agree with this observation from Teale back in the 1950's considering the importance of speaking truth as a moral imperative not just in our little audio microcosm, but also broadly of injustice, politics and ideologies in society:

"It is morally as bad not to care whether a thing is true or not, so long as it makes you feel good, as it is not to care how you got your money as long as you have got it."
                    --- Edwin Way Teale, 1953, Circle of the Seasons

--------------------

The irony of the MoFi One-Step...

A tangential musing came to mind as I was writing the comments above about the importance of the audiophile hobby being respectable and likely needing some rehabilitation. I was automatically drawn to the recent MoFi controversy.

To me, the whole MoFi OneStep-using-DSD Affair is simply a symptom of the falsehoods in audiophile culture and hype-driven commercialization of luxury products. It looks like this debacle is big enough to have made it into general news.

Let's be honest guys & gals, analogue media - tapes, LPs - obviously have significantly restricted accuracy, and incapable of "perfect" reproduction compared to the potential of digital. However, among the cultural myths in audiophilia, the belief of an "analogue ideal" has been perpetuated to the point where companies did whatever they could to achieve this because of the demand created by a romantic idea. The MoFi One-Step process basically tapped into this "Holy Grail" of "AAA" as a selling point, hence the betrayal to find out that this in essence was a fantasy for so many years. It's no surprise we're now hearing of a class action lawsuit (also see here).

Recently, in this TAS interview, Jim Davis (MoFi president) said:

"When we tried DSD, it was immediately clear this was a vastly superior method for maximizing information retrieval. Developed as an archival format, DSD is sonically transparent, with a very low noise floor. Combined with the painstaking transfer process described below, the capture is a virtual snapshot of the master, revealing detail and nuance at a level that conventional methods could not. Counterintuitively, this capture yields, in our evaluation, superior sonics compared to a cut that is direct from the analog tape to the lathe."

The only reason he would say this is "counterintuitive" is because of all the years of indoctrination the audiophile public had been fed about expecting to hear some kind of almost mystical sound quality from a purely analogue production and playback chain!

So, basically we have an audiophile press along with some "evangelists" telling people that analogue (by extension vinyl) sounds amazing, and better than digital. The hype builds upon itself over time, but by any objective standard of fidelity, it's clearly false. The MoFi / Music Direct company then saw an opportunity to tap into the myth of this simplistic and idealistic belief - simply acting on the capitalistic "invisible hand" of self-interest.

Behind the scenes, objective reality has a way of popping delusional bubbles doesn't it? Obviously, the audio engineers found that the process worked smoother, and the results sounded understandably better done at least in part digitally. In fact, it appears that from 2011 to 2014, they even just used DSD64 (SACD resolution) which I think many hi-res digital audiophiles see as not even equivalent to 24/96 PCM (I will routinely downsample SACD rips to PCM 48kHz). Whether this was done by a Meitner ADC doesn't change the limitations of 2.8MHz 1-bit PDM.

We're left now to ponder the irony that in order for those MoFi LPs to sound excellent on the scale they were being produced, there was little option but to embrace digital audio technology. And digital audiophiles perhaps may feel a bit of schadenfreude that this debacle has laid bare to everyone the obvious, predictable fact, that analogue lovers have not been able to hear the digitization process all these years!

The only thing that would make this even more interesting is if any of those LPs might have been edited through a multi-bit PCM process at some point. I see Jim Davis taking great pains to point out the "analog EQ" used. Who knows, maybe if they had used a DXD (24-bit/352.8kHz) process with careful digital domain editing, the quality would be even better when converted to the LP. ;-) Such a final DXD master with all the EQ applied in the digital domain I think would be the true "Holy Grail" for high quality remasters as it would not have to go through the analogue losses of the final EQ and reduction in quality when cutting the music into grooves.

Nothing wrong with a good existential crisis of sorts once awhile to remind us of what is true, or meaningful. Maybe the extreme LP audiophile world can finally wake up and realize that there was nothing really all that special about the quality of "pure analogue" sound other than the added noise and distortions which some might enjoy (nothing wrong with having this preference, just be honest about it).

BTW, we can obviously see just how finicky LP playback can be if we have a peek at the $6k DS Audio ES 001 Eccentricity Stabilizer review. This obviously is not my definition of a "high fidelity" playback system; imagine trying to precisely ream out the center hole of expensive collectable LPs just to hear it sound "right". Digital jitter anomalies are nothing compared to this level of gross inaccuracy. A beautiful demonstration of vinyl's limitations and it's good to see an objective measurement device being used to improve playback!

--------------------

For those who need HDMI cables... Consider going optical?

Finally, a reader from Mexico E-mailed me a few weeks back and we got into a discussion on the importance of quality cables.

Obviously, high quality cables can last longer, may look better, have more precise-fitting connections, and might functionally reduce interference through better shielding. Also, higher quality cables are important with high bandwidth digital transmissions such as HDMI. These days, if you want multichannel lossless audio, 4K+ video with HDR and 60+fps, you'll need cables at least in the "Premium High Speed" 18Gbps class if not "Ultra High Speed" 48Gbps.

Years ago, I talked about using the Corning optical USB 3 cable which can provide some galvanic isolation for the USB data lines. Well, these days, the same concept can be applied to HDMI as well although potential for galvanic isolation is not necessarily the main issue. While I'm not going to vouch for these being necessarily "high quality", I found them on sale and gave 'em a try:

These "FIBBR" cables (capable of HDMI 2.1, 48Gbps speed) were the cheapest I found (less then US$30/10') and both the 6.5' and 10' ones above worked well with my AppleTV 4K which could be finnicky at times with lower quality HDMI wires doing 4K/60fps Dolby Vision plus lossless multichannel audio.

One more thing, notice how the cable ends are labelled:

Since light travels in one direction, this is a genuine example of cable directionality (unlike a number of High-End audiophile cables with unsubstantiated directionality arrows). Plug "1" into the source (eg, computer, HDMI player) and "2" into the destination AV receiver or TV display.

Another benefit is that the optical cable is thinner and more flexible than highly shielded wires. Look into something like this if you need longer HDMI cables like 50'. One limitation is that I have not seen the HDMI Ethernet Channel feature in these products which I suspect most of us won't need.

I hope you're all enjoying the music as we enter September (back-to-school and work, post-summer holidays), audiophiles! How time flies...

"You talkin' to me?" - Inca Tern - Lima, Peru.