Musical or Analytical? Listening Fatigue and the System You Actually Keep

Thirty minutes. That is how long I lasted in front of a system many people consider the summit of high-end audio: D'Agostino's flagship electronics driving large Wilson speakers, with a dCS Varese at the front. It was spectacular. Technically flawless. Every audiophile checkbox ticked. And after half an hour I needed a break. My brain was frosted.

That half hour taught me more about listening fatigue than any specification sheet ever has. A system can be correct in every way you can measure and still push you out of the chair, while another, with obvious character, keeps you there for hours. Here is what it is, where it comes from, how to fix it, and the one test I trust before I keep anything: do I want the next song?

Frosted: the audition that was flawless and exhausting

I had spent an extra $40,000 to move from my McIntosh MA9500 integrated to the C12000 preamplifier and MC611 monoblocks, driving my Wilson Audio Yvettes. I justified the price by expecting 5 to 10%, if I was lucky. I was wrong from the first session, and within days I was asking out loud how McIntosh does this. At that point I needed perspective, and a comparison.

So I arranged the D'Agostino audition. Nothing I heard argued with its reputation; then the clock did its work. In my film comparing the MA9500 with the C12000 and MC611 I described the result like this:

It was too much of everything. Not fatiguing, but surgical, analytical, ultra detail.

I did not accuse it of glare or harshness. Yet moments later, describing my own separates, I reached for the opposite word: "musical, pleasing, airy, non-fatiguing." One audition is not a verdict on a brand, and plenty of listeners hear surgical detail as the whole point. Both sentences can only be true if "fatigue" means two different things.

What listening fatigue is, and why it shows after thirty minutes, not three

Listening fatigue is the moment a system stops giving you energy and starts taking it. It comes in two kinds, and the hobby uses one word for both.

The first is ear fatigue, or auditory fatigue, and it is physical. What NIOSH, the US National Institute for Occupational Safety and Health, measures here is a temporary threshold shift, "a temporary increase in hearing threshold level, after exposure to noise", from which people normally recover several hours after the exposure ends. Its occupational criteria set the recommended limit at 85 dBA averaged over an eight-hour shift and halve the allowed time for every 3 dB more: 88 dBA for four hours, 91 for two, 94 for one, 97 for thirty minutes, 100 for fifteen. Those are factory-floor numbers, not living-room rules. But in the McIntosh film I described what happens at 95 decibels, when reality bends and the band materialises in the room. That was a listening level, not a measured average. If a whole session averaged 95 dBA, NIOSH's table would allow 47 minutes and 37 seconds.

Slide across for readable labelsStartEnd

Read every label as live text

• 85 dBA
• 8 h
• 88 dBA
• 4 h
• 91 dBA
• 2 h
• 94 dBA
• 1 h
• 97 dBA
• 30 min
• 100 dBA
• 15 min
• 95 dB
• Recommended exposure time (minutes)
• Source: NIOSH, Occupational Noise Exposure, Table 1-1

The second kind is listener fatigue of the mind, and it is the closest name I have for what frosted me. In the first minutes everything is new, and the brain is rewarded for noticing. In my snake-oil film I called that recency bias: new gear sounds better because "your attention is heightened. Give it six months and suddenly it's just your system." Heightened attention is expensive. Every micro-detail thrown forward is a small job for the brain, and the jobs add up. Three minutes of that is a thrill. Thirty minutes is work. That is my reading of "frosted", not a laboratory result.

A demo shows you what impresses you, not what you can live with. I did extensive listening at my dealer before I bought my Bowers & Wilkins 800 Series speakers, and they still fell apart in my room. Fatigue shows at minute thirty, and in the evenings when you find a reason not to switch the system on.

Where it comes from: reflections, mismatched speakers, bright recordings, upstream glare

The physical and mental kinds share the same suspects, and in my experience the box you just bought is rarely the first.

Reflections. My first real system cost about $10,000, and I put it in an awkwardly shaped room, 10 metres wide and 7 long. I was so proud of it that I kept inviting people over to show it off, until the problems started: comb filtering from reflections, massive bass peaks and nulls from standing waves, early reflections smearing the soundstage. I did what most of us do: the treble sounds harsh, so maybe it is time for tube amplifiers. No amount of gear was going to fix those problems. In my film on room acoustics I said the room is responsible for over 50% of what you hear.

Mismatched speakers. When I moved from Klipsch Reference Premiere to the B&W 800 Series D4, my first thought was that my dealer had cheated me. In my room, on my flagship Yamaha receiver, they were too harsh, the soundstage was gone, and they sounded more detailed than the Klipsch yet hollow on every song. They were not faulty. They were more revealing, which makes a room's problems sound bigger, and, as the same dealer had warned me, my electronics were not a match. Tweeters matter too: I find the Klipsch La Scala's horns less fatiguing than B&W's diamond tweeters, more coloured but more pleasing.

Bright recordings. Some fatigue is mastered in before your system ever sees it. Earl Vickers's paper for the Audio Engineering Society's 129th Convention, in November 2010, reviewed the loudness war and named listening fatigue among its possible consequences. Its own title admits "Speculation": plausible, not proven. My Yvettes reveal everything, good and bad; the warmer WiiM Ultra makes classic rock and poor recordings easier to live with. If only certain albums tire you, suspect the album first.

Upstream glare. This is an edge that follows every record, at every volume, wherever the speakers sit. Distortion is the obvious suspect, and the spec sheet the obvious tool. It is the wrong tool. In a 2003 listening study for the Audio Engineering Society, Earl Geddes and Lidia Lee had 34 listeners rate 21 stimuli. They found no significant relationship between the ratings and the standard THD and IMD figures, while a metric of their own design tracked the ratings closely. If the headline THD figure did not predict what those listeners heard, it cannot be your fatigue test. An evening with the component can.

Musical vs analytical: two words that hide real choices

Every argument about listener fatigue ends at these two words, and neither is a measurement. Underneath them sit real choices: how much colour you accept in exchange for pleasure, how much of the recording's truth you want, ugly parts included, and above all how the detail reaches you.

That last choice is the whole debate, and my separates taught it to me. They retrieve more detail than the MA9500, not less: micro-details in recordings I have heard hundreds of times. Yet in the film I said it happens "not in a fatiguing hyper-analytical way but naturally. You notice because they are simply there." So "musical" never meant less information for me. It means detail you discover rather than detail thrown at you. The MA9500 earned the same word; in the film's own words, "You can listen for hours without fatigue."

Correctness alone does not make a keeper. My own McIntosh MEN220 room correction is digital processing; placed in an analogue chain, it converts everything to digital, and the sound becomes "less organic, less natural, better balanced, perhaps more correct, but less alive." Those last words, more correct, less alive, are the analytical trap in one phrase. The La Scala gives Klipsch magic, immediate and fun; the Yvette gives truth. I bought the truth, and there is still no shame in choosing the magic.

Now the confession. I have $40,000 in those McIntosh separates, and I have said on camera that if someone tells me a $15,000 integrated sounds identical, my brain physically resists the information. That is sunk cost, and it lives in my head like everyone else's. So treat my word "musical" with the same suspicion you bring to any reviewer. The frosted brain was real. The explanation is mine.

The next song: the test I trust

My test is one line from the McIntosh film.

...the only thing I care about is the next song and the next song and the next song for hours.

In another film I said the best component is the one that gets out of the way of the music. A fatiguing system never gets out of the way; some part of you is always aware of it. A keeper disappears; of my separates I said the system is so quiet "that you forget you are listening to a system." Here is how to find out which one you have:

1. Give it an evening, not a demo. My break came at half an hour, so judge after that.
2. Use a familiar playlist, side by side. Your ears are the final decision tool.
3. Watch what you do, not what you write down: reaching for the volume, for the phone, or for the stop button halfway through a track.
4. Keep a way back. Buy controversial upgrades only with the ability to return them. No ego, no sunk cost.

If you are checking the clock, the system has failed, however well it measures.

Fixing a tiring system: placement and room, then matching, then components

Fix it in this order, cheapest and most permanent first. I learned it backwards, and I was about $10,000 poorer before I started studying room acoustics. Start from the symptom.

Placement comes first because it costs nothing: toe-in, toe-out, distance from the back and side walls, moved in very small increments. A properly positioned speaker is the difference between a $5,000 pair and a $15,000 pair. Then the room. Spend no more than $500 at first on absorption and diffusion at the first reflection points, one piece at a time; it is very likely a bigger upgrade than any cable. Corner bass traps come next, on a similar budget, for a tighter, cleaner low end. Measure with free REW software and a microphone before you buy anything else. And stop before the room dies: overdo it and "your room becomes lifeless. It becomes dead."

Then matching. My B&Ws on a receiver taught me that the amplifier has to suit the speaker. Do not bolt more resolution onto a system that cannot use it either; that is a Formula 1 engine in an ordinary car. Only then, components, and only ones you have lived with for an evening.

The system that wins the first thirty minutes is easy to buy. The one worth keeping wins the third hour, when you are no longer listening to it, only through it. As I said of my separates, only the ecstasy remains, song after song. The gear serves the music, not the other way around.

Frequently asked questions

What is listening fatigue?

Listening fatigue is the point where a system stops giving you energy and starts taking it: physical ear fatigue from level, or mental overload from detail.

Is ear fatigue the same as listener fatigue?

No. Ear fatigue, or auditory fatigue, is physical: a temporary threshold shift after loud sound, normally gone several hours after the exposure ends. Listener fatigue is attention running out, the state I called a frosted brain after thirty minutes of surgical detail.

How long should an audition last to reveal fatigue?

Longer than half an hour, with music you know. A short demo shows you what impresses you, not what you can live with.