More digital cable fun

Audiostream has another digital cable promotional article.

Within it is a little aside of a type that I often see:

<some company> is choosing to keep mum on the specifics relating to how these cable’s differing materials and construction impart different sonic flavors.

In fact no one has ever revealed the secret of how an ethernet cable (or any cable for that matter) imparts different sonic flavours – for the simple reason that it is impossible. We know it is impossible because this is a man-made, engineered system, one of whose primary purposes is to make it impossible. If, in fact, an ethernet cable physically changes the sound (as opposed to placebo effects), then this is not a digital audio system at all. It would probably be best to repair it, re-design it, or throw it away sell it to the nearest person who believes that manufacturers of audio cables “keep mum” only because they have valuable secrets to safeguard.

Digital cables: a lack of understanding

There’s a new article about digital cables which is directly relevant to my post a couple of weeks back – it even features the “eye diagram” I referred to.

It purports to ‘debunk’ the idea that “bits are bits”, but in reality it does nothing of the sort. It starts off with a question that it fails to answer and doesn’t come back to again:

Asynchronous USB audio to the rescue?

The remainder of the article is about isochronous USB and seeks to suggest that it is not just about getting all the data through intact, but that there are issues with timing, jitter and noise. Judging by the comments that follow, the readership is entirely unaware that it is a non sequitur they have been reading.

There is a strange little “Note” placed halfway down in tiny lettering that says

Do not confuse ‘asynchronous USB’ with ‘Isochronous,’ an asynchronous USB system still uses Isochronous mode to transfer audio.

I am not quite sure what that is supposed to mean, but again it seems to be an appeal to the false idea that it is beyond the wit of man to deliver digital audio data without it being corrupted, or affected by timing. As I have said before, if this were true, then systems such as TIDAL could not work: the data has come through thousands of miles of cheap cable not even made of silver or raised from the floor on little ceramic pots, through hundreds of utilitarian digital switches, and eventually into your home via the cheapest connector that your telecoms company could provide. And yet it works perfectly; all the electrical noise of the world’s internet, and all the jitter that comes from being re-routed dynamically hundreds of times every second, is completely absent.

As I said in my earlier post, we need to understand the system at a higher level. Asynchronous USB and other on-demand packet-based systems are immune to cable quality once it is beyond some minimum standard, and electrical isolation removes the possibility of noise being injected into your audio system – even enabling you to listen to music off the WWW without all the electrical noise of the global internet spoiling your listening pleasure. Really, getting data from a file/CD/stream into your DAC and reading it out at a precise rate is a simple engineering problem, rendered trivial and cheap for the consumer by some very clever people who actually understand the problem, and have solved it.

HD Vinyl

Apparently there is talk of developing “HD Vinyl”…

Imagine a vinyl record that has 30% more capacity, 30% greater volume, and double the audio fidelity of a typical LP sold today.

The layers of irony inherent to this concept are many:

The HD Vinyl process… involves… perfecting the topographic, computer-generated, 3D modeling imprint before any physical manufacturing takes place.  “We adjust the distance of the grooves, we correct the radial/tangential errors, and we optimize the frequencies,” Loibl continued.  “You could say we ‘master’ the topographical data, which is a totally different approach.”

After that, a ‘pulsed high-energy Femto-laser’ burns the audio directly onto the stamper.

Over a year ago, I was pondering on the idea that the way hi-fi was going, we would end up with 3D-printed phonograph cylinders. I am glad that the process is well under way.

Understanding vs. Knowledge vs. Expertise

Reading Archimago’s latest article, I am struck by two things:

(a) he’s probably a much nicer person than me

(b) the high end audio business may not be so much a scam as a <word that begins with “cluster”>.

What I’m thinking is that millions of words (not Archimago’s I hasten to add) and thousands of product developments can all stem from an incomplete understanding of something, plus the obscuring of central ideas with mere facts and ‘expertise’. Case in point is the subject of digital cables, discussed in the article. Archimago has endless patience and ‘plays along’ with the manufacturers and their disciples, giving them the benefit of the doubt by testing their products, even though he knows that they will always give more-or-less identical results. I am more inclined towards the idea that we should just ignore them. Unfortunately, neither approach can rid us of the cloud of confusion that surrounds the subject, and which must ultimately divert attention away from real progress (and with the Kii Three and Beolab 90, for example, we at last have some real progress to evaluate, and hopefully to make cheaper).

Which of the the following propositions are false?

1. Digital audio is degraded by jitter.

2. Cables carrying digital audio signals impart jitter onto those signals.

3. High quality cables can be designed to reduce jitter compared to others.

4. Digital audio can be improved by high quality cables.

I would say that none of them is completely false in all circumstances, and herein lies the problem. An entire industry is sustained on these propositions; manufacturers talk knowledgeably (and who knows, maybe honestly) about them. But while the propositions may all be true individually and with qualifications, they are an incomplete representation of the role of cables in digital audio. An understanding of digital audio at a system level would show that many implementations are certain to be immune to cable quality (and the remainder of implementations are highly insensitive to it). Those bi-directional implementations that deliver the audio data in packets on demand, slaved to the DAC’s sample rate, are immune to cable-induced jitter by design. Case closed. We could simply move on.

But if you have the misfortune to enter into a discussion about the subject on a forum, you will quickly encounter many vociferous people with detailed factual knowledge and expertise regarding cables, and that that is all they want to talk about; it won’t be long before someone brings up “eye patterns”, transmission line theory, the “skin effect”, “six nines” copper, monocrystalline metal and so on. Or they want to talk about listening tests and statistical ‘p-values’. This lower level knowledge, even if correct within its own little sub-fields, is not really relevant, but unfortunately forms a smokescreen that obscures all understanding.