Audiophile Snake Oil, Biases and Burn-In: Why Both Camps Are Only Half Right

I am biased. You are biased. Every single person watching is biased. I said it about a minute into my February 2026 film on audiophile snake oil, before I had opened a single bottle and moments after admitting I had spent over $35,000 on cables alone and bought more than a hundred of them. This was not modesty. It was the only honest place to start.

One cable costs $3,000, another 12 bucks on Amazon: same copper, same electrons. One camp calls the difference a revelation, the other a con. My position: both camps are only half right. The short version is my film on snake oil in audio. This piece puts my own components on the stand and adds what the film could not: a wine study, a theatre-ticket study, a blind loudspeaker test inside Harman, one maker's run-in hours and a decibel figure.

I am biased, you are biased: why the confession comes first

The question is not whether we have biases but whether we are honest enough to acknowledge them. Bias is a stage of the audiophile journey: some of us mature through it, some see it and cannot overcome it, and some are still deep in the bias tunnel with no light at the end.

A second reason is in the channel trailer: everyone in audio has an angle. My promise there was no monetization and no sponsorships; the snake-oil film signs off with no sponsors, no free gear, just my money. That money is exactly what makes me vulnerable. A reviewer sends the loan sample back. I paid. Over $150,000 into this hobby buys many things, including a strong wish to have chosen well.

Five biases, each with a moment from my own system

The film named five, in this order. Here each is set against a moment from my own system.

Expectation bias: the price tag as a prophecy

Drop $5,000 on a new DAC and your brain wants it to sound better; the price tag becomes a self-fulfilling prophecy. In a 2008 paper in the Proceedings of the National Academy of Sciences, Hilke Plassmann, Antonio Rangel and co-authors gave volunteers the same wine under different price labels. The pricier label was rated more pleasant, and the scanner showed more activity in the medial orbitofrontal cortex, a region linked to experienced pleasantness. The label did not change the wine. It changed the drinker.

My own case is more embarrassing. When I bought my Klipsch floorstanders, the dealer, who also sold QED, told me branded cable would sound better than Amazon copper. I bought QED XT25 and never checked: I heard nothing. I assumed.

Confirmation bias: hunting for refinement

You read that Transparent sounds more refined than AudioQuest. Now every time you listen you hunt for refinement, and you find it, because it is all you are looking for. Both brands have lived in my system: AudioQuest from Rocket 88 speaker cables to Black Beauty interconnects, and today Transparent Super and Reference in every critical position.

I still believe the Transparent beat the Rocket 88 on my Wilsons, audibly even to untrained ears. But a belief formed with the eyes open has a favourite. Confirmation bias feeds on what others tell you to look for, so starve it. In the new listening room I showed in February 2026, I did the first critical session deliberately alone, with no partners from The Sound Lab, so my first impressions would be mine and uninfluenced: the helicopter on Pink Floyd's The Wall came in high from the ceiling, for the first time in my system.

Sunk cost: $40,000 in my own head

I have $40,000 in McIntosh separates. If someone tells me a $15,000 integrated sounds identical, my brain physically resists that information. I need my purchases to be justified; I called it $40,000 of sunk cost in my own head. Far from hi-fi, in "The Psychology of Sunk Cost" (1985), Arkes and Blumer reported that theatre-goers who paid full price for a season subscription went to more plays over the next six months than those given a discount.

I moved from the MA9500 integrated to the C12000 and MC611 separates expecting a marginal 5 to 10%, and heard far more from the first session. I believe that verdict, and I know sunk cost would have written exactly the same one for me. The check: keep a way to return, and sell what does not work. I have sold cables over the years, each sale a small defeat for this bias.

Recency bias: give it six months

The equipment you just bought always sounds better than what you had: it is new, it is exciting, your attention is heightened. Give it six months and it is just your system. My most recency-loaded session was the first with the McIntosh MA9500, streaming from an Eversolo A6 Master Edition into my Bowers & Wilkins Diamonds: "Giorgio by Moroder", "What a Wonderful World", then "Birds" by Dominique Fils-Aimé. I hugged my dealer and broke in tears.

Tears are not evidence. Time is. That amplifier stayed my partner in crime for years, through new speakers, cables and sources, and every single time it unfolded everything I needed to hear. That is the recency test, passed over years, not in one session.

Vocabulary bias: the words that create what you hear

This one is insidious. Once you learn words like soundstage, depth or micro-dynamics, you start hearing them, or you think you do. The language creates the experience. When my dealer threw in a pair of Vincent cables with my B&W Diamonds, fat, heavy and sturdy like two anacondas, I guessed they were cheap because they were free. I still heard a noticeable difference against my generic QEDs. Were they better? That is subjective. Were they different? Absolutely.

I learned to keep those two words apart. "Different" is an observation; "better" is a verdict, and the vocabulary is always in a hurry to deliver it.

Golden ears go wrong when they refuse to admit bias

The golden ears say: trust your ears; if you hear a difference, it is real. They are right that differences can exist beyond what today's measurements capture, and that the final test happens in your own system. Their weakness is no methodology, no controls, pure subjectivity dressed as expertise.

The hardest evidence against them came from inside a loudspeaker company. In 1994 Floyd Toole and Sean Olive presented "Hearing is Believing vs. Believing is Hearing" at the Audio Engineering Society's 97th Convention: 40 Harman employees rated four loudspeakers, sighted and blind. Sighted, ratings rose significantly for the larger, more expensive models, and loyalty to Harman products crept in. Blind, the small, cheap speaker edged slightly ahead of a rival it had trailed, and the effect of where each speaker stood in the room became clearly visible. Sighted, that position effect almost completely disappeared.

That finding hurts me more than any cable argument. I have said on camera that room acoustics matter more than almost any component upgrade, yet with the eyes open the brain stops hearing the room and starts hearing the badge. The golden ears are also wrong that a $100,000 DAC always beats a $10,000 one, and I count myself here: I have bought things I later realised made no audible difference.

The measurement camp goes wrong when it claims to capture everything

The measurement gurus answer: if it does not measure differently, it does not sound differently. Period. They are right that a well-designed $1,000 DAC measures perfectly, and right that sighted bias is real; the Harman study is their weapon, not mine.

Their weakness is assuming we have identified every measurement that matters to human perception. We have not, and science is about humility. The skeptic's favourite move, bias exists therefore the difference does not, is not a conclusion; it is another bias. Biases are person-specific: they tell you to distrust a listener, not that what he heard is missing.

There are real differences: not everything that is claimed, but not nothing either. After opening the bottles with my own money, my own systems and my own ears, this is where I stand. Cables make a real audible difference, not always worth the price, and the more resolving the system, the more they matter: on a $2,000 setup, save your money. DACs hit hard diminishing returns after about $5,000. Power conditioning (IsoTek Aquarius and Corvus, in my case) gave me mixed and subtle gains, though I have heard conditioners transform systems on noisy power. Is it snake oil? No. Is it universally necessary? Also no.

Slide across for readable labelsStartEnd

Read every label as live text

• Golden ears
• Measurement camp
• Right
• Wrong
• Differences beyond today's measurements
• Final test: your own system
• Refuse to admit bias
• No methodology, no controls
• $100,000 DAC always beats $10,000
• Sighted bias is real (Toole & Olive, 1994)
• Good $1,000 DAC measures perfectly
• Every perceptual measurement already identified
• Bias proves no difference
• Where both camps meet
• 1. Fundamentals first
• 2. Audition in your own system
• 3. Level-match within 0.1 dB
• 4. Judge again after six months
• 5. Keep the right to return

Do speakers need to be broken in? Burn-in, case by case

Yes, speakers need a break-in period, because their suspensions are mechanical parts that loosen with use. In the film I asked whether gear needs 100 hours of playing before it sounds its best, and for speakers my answer was yes, absolutely. That is mechanical fact. The rest of the chain is where honesty gets harder.

Speaker break-in period: the maker's hours and the listener's weeks

How to break in speakers is the easy part; not fooling yourself is the hard part. The maker's number is far smaller than the 100 hours in my question. Bowers & Wilkins, which built the 800-series Diamonds I owned, prints a running-in note in its owner's manuals. The one for its CMC centre speaker asks for "15 hours of average use for the mechanical parts to attain their intended design characteristics", and up to a week for temperature effects to settle after cold storage. At an hour or two a night, 15 hours is one to two weeks of ordinary evenings: play music you know, at the level you normally listen.

Those same weeks are when your ears adjust and your attention runs high. Speaker burn-in and recency bias arrive in the same box: no verdict until the speakers have had their hours and you have had your weeks.

Electronics and cables: sometimes it is my ears adjusting

Electronics are more debatable, though capacitors do stabilise over time, and I act on that. When my system was installed in the new room, I ran 10 to 20 hours of burn-in on the new power circuits before I started positioning the speakers, to get those capacitors charged and properly warmed. With cables, I have never reliably separated burn-in from my own brain burning in to the sound. Sometimes equipment sounds better after break-in; sometimes my ears have simply adjusted. I cannot always tell the two apart, and that admission is more valuable than a firm position either way.

A method that respects both camps

None of this ends in "it depends". It ends in a method that costs almost nothing:

1. Fundamentals first: room treatment, speaker positioning and good source material, before any controversial upgrade.
2. Audition in your own system, and take the first impressions alone. Never buy because a reviewer told you to or because the price suggests quality.
3. Level-match. Stereophile's John Atkinson, in a sidebar on level matching: "It's essential in listening comparisons to match levels to at least within 0.1dB. A component that's, say, 0.2dB louder than another won't be heard as being louder. Instead, listeners will comment on greater apparent detail or better dynamics." That is vocabulary bias with a decibel figure attached.
4. Take time. Let speakers break in and your ears settle; judge again after six months.
5. Keep the right to return. If you did not hear it in your system, return it. No ego, no sunk cost, just honesty.

Audiophile snake oil, the real kind: certainty

The real snake oil in this hobby isn't expensive cables or power conditioners. It's certainty.

The certainty of the skeptic who has never listened. The certainty of the audiophile who never questioned. Both are poisons. Different bottles.

I know the taste of the second bottle. On my flagship Yamaha receiver, my new B&W Diamonds sounded hollow on every song. I rang my dealer, shouting that the speakers were defective. He reminded me he had warned me: my electronics were not a match. Yamaha had been my go-to brand for years, so I was sure he was wrong and called B&W's distributor, who asked why I had been sold these speakers and whether nobody had told me about the limits of my amplification. I had royally messed up. The only thing defective that day was my certainty. The cure was the MA9500 of that tearful first session, and it became my single best purchase.

The difference between an honest audiophile and a deluded one isn't what they hear, it's whether they are willing to question it.

Unlike the two camps fighting online, I am actually willing to change my mind.