A computer hard drive that makes your music sound better?

I'm sitting here writing this week's newsletter with music playing in the background - I've just listened to tracks by (and I'm name-dropping here) Malcolm Middleton, the Flashbulb, Madvillain, PJ Harvey, Mogwai and 65daysofstatic. I'm enjoying this music being played from Plex on my Chromecast, through my TV, via a Sony home theatre amp, to my in-wall 7.1 surround sound speakers. The entire setup might have cost me $1,200, if we exclude the cost of the TV (another $1,500). But could I be enjoying my music more if I'd spent more money buying reference equipment from high-end specialist companies?

It seems unlikely. Like many things in life, there are likely to be diminishing returns as you pay more for something. But there's something particular about high end audio equipment that really annoys me. The top end of the market seems ridiculously expensive - is that money really being converted to an improved listening experience?

For those not in the know about audiophiles, there are some people who choose to spend boatloads of cash on specialist audio setups that are meant to improve the sound of their music. This might be a basement setup with a bare concrete floor, with Hi Fi separates ($20,000 valve amplifier, $7,000 CD player and $56,000 record deck) each sitting on a separate marble slab for reduction of unwanted vibrations. These components will be powered by a dedicated three-phase power supply delivered to the house independently of the power for the lights, power sockets, etc, via a $55,000 power cable. The components will be cabled together with $54,000 gold plated RCA interconnect cables, and the speakers will be wired up with “Oxygen Free” cable that costs $57,000 per metre. The amplifier may even have had its volume knob replaced by a $500 wooden one.

When it comes to people justifying these high prices for audio equipment, they will talk about qualities such as the noise floor, background, signature, PRAT (pace, rhythm and timing), depth and positioning. So, are these differences in how the music sounds real, or just wishful thinking?

So much of what is sold as high end audio, when put through blind testing, turns out to be no better than reasonably priced hardware. One of my favourites is the testing of straightened coat hangers vs expensive cables, which is a test that has been repeated recently using professional measuring equipment rather than human ears, and one which shows there's no discernible difference between the two.

As far as I can tell with these blind tests of high-end audio devices, even if a difference could be measured on professional audio testing equipment (and often it can't be), the sound quality difference from a properly configured reasonably priced setup cannot be discerned by the human ear. The difference is likely all in the listener's head, not their ears. Many audiophiles refuse to take part in blind tests, using hand-waving excuses to avoid finding out they've wasted their money.

The idea that spending lots of money won't get you a better sound is even more the case for digital equipment. With digital signals, a data stream cannot be made to be “warm”, “smooth” or “neutral”. A digital signal contains zeros and ones - there's no room for nuance of sound there. There is the possibility of errors being introduced to the signal, such as a bit being flipped from a 0 to a 1, but most digital data transfer protocols have error correction algorithms to protect against this. This might use parity/error-correction bits to allow the detection of errors, which would trigger a request for the data to be re-sent (or even the in-place correction of a single-bit error).

So for any digital audio hardware that purports to offer an improved listening experience, I become immediately skeptical. There's an argument to be made that ineffective shielding of digital components can cause unwanted signal leakage into analogue output audio after it's been converted from digital. Some of you may have noticed this when using a pair of wired headphones plugged into an older PC, or one with a cheaper motherboard. It might be that noise from some of the components will induce an audible hum into your audio. But when we're talking about even a half-decent audio setup, rather than a cheap PC, this will not be a problem. And for audiophiles with their isolated power feeds, remote DACs and optical cables, this is really not going to be an issue at all.

This brings me to a piece of hardware I found online recently - a custom-built audiophile data storage device, specifically an M.2 NVMe SSD stick. This piece of PC hardware is where digital files are stored in a computer before they're transferred into memory, pulled into the CPU for processing, and so on.

At first I thought the idea was to store your music files on this drive, but it turns out that - because it's only 333GB when used “properly”, and audiophiles like to use large lossless, uncompressed, high bit-rate files - the idea is to keep your copy of Windows on this drive. You are meant to store your music on another PC (e.g. a NAS), which is kept remote from the playing hardware. Your audio files will be transferred by TCP-IP over ethernet to your audio PC with this SSD in it (and I bet there's a good chance this transfer will involve a $10,000 ethernet cable and a $2,500 network switch!).

So, given that the files placed on this storage, and their subsequent transfer through various components, is totally digital, and that this is meant for your OS files only, there's no way that any kind of audio signature, or nuance of sound, could be applied by this storage device to any audio files. But that's not what the audiophiles who've tested this hardware say in the Audiophile Style forum:

_“The bass and tonal range is expansive”

“My Ears don't lie! I can hear differences in Freq Response, PRAT and Sound stage.”

“Bass line is punchy and full but clearly delineated. Mids are not laid back but slightly forward. The highs are as clean as the PSU quality but is still sounding a bit dry”

“The Bass line was now powerful and articulate. Mids are less pronounced and blend in with the bass. The Highs have lost the "dryness' and are now well balanced with the mids and the lows.”

“Very neutral especially in the High frequencies. No digital nasties. Extended highs without any sibilance on hardness. Bass lines are clean and articulate. Impact of bass and drums is very good. The mid range is neutral, not lacking any warmth and pace. Stereo spread is now very good; you hear sounds extreme left and right just outside of the speakers and/or HPs. Depth is good. Positioning of vocals and instruments is spot on”_

The forum goes on with people talking about the sound signature of different Operating Systems (Windows 10 vs Server 2016 vs Linux), whether to power the device via the added DC power jack or the motherboard's power supply, how long to “burn in” the drive before the sound improves, and which network card to use.

Honestly, reading this kind of feedback is depressing. It's a group delusion of people who have convinced themselves that they have super hearing, that physics bends to their will, and that black is white. At least one member of the forum spoke up, calling this “Classic snake oil”. I'll leave you with his reply when he was told to go and complain somewhere else (just before having his account deleted):

_“I call out meaningless BS whenever I see it. Too many people make money off other people's ignorance. That bothers me a great deal.

I had hoped to learn new things about audio reproduction by engaging with this site. Unfortunately I have encountered a community largely composed of adherents to a fantastical proto-religion in which logic is scorned, science is ridiculed and BS is exalted.”_