Why we (often) upgrade the wrong things (first) and ignore the important things


I continue to be impressed by this person. See:

https://www.youtube.com/watch?v=8jWM0NzE_us&t=50s

He argues for applying a bottleneck principle to audio upgrades: system performance is always capped by its dominant constraint, so upgrading any non-limiting component produces only incremental refinement, not genuine improvement.

The hierarchy of constraints, roughly in descending magnitude, runs: room acoustics and speaker placement (capable of 10–20 dB swings), amplifier headroom and load stability, gain structure alignment, DAC output characteristics, and finally cables. Most audiophile upgrade behavior inverts this hierarchy — people swap DACs and cables because they’re easy, not because they’re limiting.

The gist is a distinction between **audible change** and **ceiling removal**. Any swap can produce a detectable difference; only removing the dominant constraint raises the system’s maximum capability. 

He gives a way to test it. When we remove a constraint, that feels unmistakable and immediate — dynamics expand, you stop gear-monitoring and just listen — whereas novelty-driven changes require effortful attention to detect and fade in significance.

Room treatment and gain-structure analysis are effortful; cable swaps are not. Making this worse are the ways we mis-focus on novelty, which amplifies perceived differences. We think we’ve made a structural improvement but we have not, actually.

 

hilde45

This guy's point about focusing on the constraining factor(s) certainly makes sense although he would relate this to cost if he understood basic economics.  The particular points that he makes, however, are more relevant for mid-fidelity systems than for high-end systems.  For example, his comments about amplifier headroom and speaker sensitivity are important for a mid-fi system where there might be questions about whether an amp is capable of driving a set of speakers.  If someone is constructing a high-end system, it's not hard to find amps that are able to drive a particular pair of speakers, and the focus should be on what type of sound that they produce together. 

Likewise, if someone has a mid-fi system, then the focus will be on simple basics, rather than on sound quality, and it becomes easier to justify spending  $200 or less on a pair of cables.  However, if one has a high-end system, than a pair of cables that cost less than $1,000 is likely to be an important limiting factor in a high end system.

@kerrybh 

@kerrybh

I think we agree on more than we disagree, but I'd push back gently on equating "subjective" with "preference." That flattens out something important.

Take your Magico example — some hear "detailed and transparent," others hear "bright and fatiguing." Those aren't just brute preferences. They can connect with physiological differences, with different interpretive frameworks, with different listening priorities — all genuine facts about each person. All these are genuine facts about each person. They wind up disagreeing, but we can understand why. If you add two numbers in Base 10 and I add them in Base 2, we get different answers – but this is not because they're "subjective." 

"Subjective" is often the word people reach for instead of spelling out *why* they prefer something. Once you actually unpack the preference — "I find that upper-midrange peak fatiguing after an hour" or "I love the way that treble energy reveals detail in recordings I know well" — something interesting happens. You find areas of genuine overlap (both listeners are hearing the same frequency emphasis) and areas of real difference (one finds it engaging, the other exhausting). The conversation gets traction. 

And that's my key point: the presence of disagreement doesn't make something *inherently* or *purely* subjective. It just means the evaluation is complex and multidimensional. Two people can agree on what a speaker is doing sonically and still weight those qualities differently. That's not pure subjectivity — it's a mix of shared perceptual ground and genuinely different priorities. And there are many *other* people who share those priorities. That makes it, at least, inter-subjective. Or, just objective but not universal.

Discussions go sideways not because people won't accept different preferences, but because calling something "subjective" often functions as a conversation-stopper rather than a conversation-starter. It skips right past the interesting part — the reasons.

@snilf  Thanks for your post. Really well put, and the coffee anecdote is great.

I think you're exactly right that the "objective" standard unravels quickly. The notion of reproducing the "original performance" falls apart the moment you look at how recordings actually get made. The same album remastered in different decades can sound dramatically different. The studio where the mix was finalized almost certainly used monitors nothing like what any of us play back on at home. The recording engineer made dozens of aesthetic choices — mic placement, EQ, compression, reverb — that already represent an *interpretation* of the original event. So which version is the "objective" reference? There isn't one. It's interpretations all the way down.

What I find interesting is the quiet rhetorical move where "objective" slides into "real" and then into "better." Measurements become a proxy for accuracy, accuracy becomes a proxy for fidelity to something real, and fidelity to the real becomes the standard of quality. But each of those steps is a philosophical leap, not a given. And I think that chain of identifications is often a way of dodging harder questions — namely, what makes for genuinely good *interpretive* criticism of sound reproduction? The kind that's nuanced, cumulative, built on experience and careful listening, and open to revision. That's the hard work, and slapping "objective" on a frequency response graph lets you skip it.

As you probably know, your coffee story is actually a beautiful illustration of something Hume argued about taste. He insisted that taste isn't just brute preference — it's a capacity that can be educated, refined, and improved through experience and exposure to good examples. You didn't just switch to filtered water and decide to prefer it arbitrarily. You trusted a knowledgeable friend, gave it time, and your palate genuinely developed: the cultivation of discernment. Hume would say your current preference is *better* than your old one, not because filtered water is "objectively" superior in some measurable sense, but because your taste is now more informed, more experienced, and more capable of drawing relevant distinctions. That's the kind of standard worth aspiring to — not objectivity, but educated judgment.

@emergingsoul If the videos feel too long, that might say more about your attention span than about the videos. He's working through complex topics that don't reduce to soundbites. And what might come across as condescension is probably just what it feels like when someone is being thorough about something you're not fully tracking. That's not his problem to fix.