Evidence and argument

Audio is a well-understood science#

Audio is not mystical and unknowable. It's a very well-understood science, and facts in audio aren't a matter of philosophical disagreement. So objective evidence (measurement, blind tests, null tests, statistics, maths and physics) outranks feelings, adjectives and marketing.

When a tiny speaker seems to defy physics with big bass, there's usually a trade-off hidden somewhere, because anything you do in the frequency domain has consequences in the time domain. Precise definitions and the engineering textbooks are more useful than vague vocabulary. "Legendary", "magic", "mojo" and "pedigree" are red flags to me.

Some problems are solved and some are not#

When I say something is a solved problem, I mean it used to be a problem and now it isn't. Preamps, cables, conversion, clocking, amplification, mixers, digital audio and plugins are largely solved. The engineers of the past would have loved a $200 interface.

People like to extrapolate that into "AP thinks everything is snake oil". I don't. Microphones, speakers, headphones, phono cartridges and styli, room acoustics, AI stem separation and live sound are not entirely solved problems. We've got very good at a lot of those things, but they're not done. So it isn't that nothing is worth spending your money on.

The appeal to Grammy fallacy#

The audio industry's favourite fallacy is probably appeal to authority, specifically what I call the appeal to Grammy. Someone has famous credits, so whatever he says is treated as true. A very famous mastering engineer can still get the sampling theorem wrong on camera. That doesn't make him a bad mastering engineer. It means that making records sound great and being factually right about why are two different skills, and nobody's credits make him infallible.

Expertise is domain-specific: I don't go to my mechanic for a toothache.

If 10,000 people disagree, 10,000 people can still be wrong#

Truth doesn't depend on a majority. There is one set of facts, and "personal truth" is nonsense. A loud comment section can be confidently wrong: a majority of commenters said DACs sound clearly different while the blind poll said nobody could hear them.

Popularity isn't evidence either way, and neither is correlation. Famous gear and famous engineers can correlate with hit records without causing them.

An argument needs premises and a conclusion#

An argument is a set of premises that lead logically to a conclusion. Without a conclusion there's no argument. Without premises it's just an assertion.

Valid means the structure guarantees the conclusion if the premises are true. Sound means valid with true premises. Probabilistic claims can't be valid, which is why most audio studies, being empirical, aim to be cogent: not "X is true" but "there is good evidence that X is true". That's also why evolution is a theory and not a theorem.

Trust the evidence bro#

Whoever claims it sounds better has to prove it, and I hold myself to the same standard. I publish session files, audio and studies so anyone can replicate my findings. That isn't formal academic peer review, and I don't pretend it is. But it's an upgrade from the standard level of trust me bro in the scene.

Censorship#

I don't delete critical comments. If I'm wrong, I want to be called out, with an argument rather than an insult, and anyone who comes on YouTube as an expert that people learn from, me included, should be held to a high standard.