Hacker News Reader

Comment by DiabloD3 | Hacker News Reader

I love how nobody actually answers the question in that thread.

Audiophiles neither like nor dislike tubes vs solid state amps... historically, tube amps had better corner case behavior; usually the corner case being less distortion driving difficult to drive speakers and headphones.

This hasn't been relevant, mostly, for at least 15 years. The world's best amps have been built in the past decade and are solid state. The only reason to own tubes today is for a change of pace, not for any technical reason.

View on HN

Anecdotal:

I did a kind of A/B test over several months between one of those audiophile Class-D amps that were the rage so many years ago and a simple single-ended tube amp that I had built from schematics.

Both sounded good to my ears when fired up. They were, for me, indistinguishable in sound quality.

However a funny thing happened after listening to the Class-D for weeks and then plugging the music source into the tube amp and firing it up: some indescribable "listening fatigue" just let go and was nowhere to be found with the tube output.

Like I say, I can't even put my finger on it, quantify it, but there was something just more calming about the sound reproduced by the tube amp.

Sure, not a blind test. And I have noted in the past a kind of bias I can have toward something I made myself (ha ha). But perhaps when the audiophiles start into some long diatribe about "even vs odd harmonics" there may be some subtle psychological effects they're poking around.

Regardless, I was sold from that point on. All my music listening around the house comes through my home-built tube amps and I am a happy listener.

As you say, not a blind test.

It's very easy to convince yourself that something is helping. It's very common in audio production circles to have had the experience of working on, say, a snare track, tweaking EQ or whatever, and to convince yourself that you've improved it only to realise that the EQ is bypassed or you're on the wrong channel - basically you wiggle some knobs that make absolutely no difference to the sound but still hear a difference.

I'm pleased you've found something that you enjoy, as after all, that's the point - whether you chose something on sound quality terms or not is irrelevant (as you have already realised they sound basically identical). If it means you enjoy listening to music more, you've won!

You could clear this up with a high ENOB digitizer and running stats on the error signal of different amplifiers.

Because tube amps emphasise certain frequencies/harmonics that are more pleasing to the human ear? Hence the "warmer" sound effect.

@bitwize et al answers the question much much better.

You're leaving out harmonic distortion. Tube amps produce primarily more pleasant sounding even-order harmonics. The odd harmonics of solid-state amps sound harsh.

That is an often quoted, but incorrect analysis; an amp that distorts audibly is not a good amp. I can buy both solid state and tube amps that do not distort relevantly.

The perfect amp is a wire with gain, no matter the technology it uses.

For reproduction. The topic at hand is art generation. The perfect amp doesn't exist because it changes from listener to listener and moment to moment. That said, a distorted guitar amp sounds kickass.

The perfect amp is the amp that reproduces the sounds faithfully. The rest should be handled in software. Sure if you want to distort your sound that's fine. Load this tube amp distortion profile.

A guitar amp isn’t reproducing sound - it’s making sound. The amp itself is an instrument being played by the guitarist.

The original posters general point wasn't only about guitar amps. I'm not disagreeing with their third sentence only with their first and second.

What you are saying is correct from an engineering standpoint but it completely ignores the human psychology.

Tube amps produce a lot of (not only harmonic) distortion, especially the cheap stuff. And then there are also various nonlinearities due to the inevitable transformers, tubes aging, mains hum, etc.

The problem is that musicians (and not only) were "raised" on this type of "sound". So they consider this to be the nominal, "correct" way how an amplifier is supposed to sound like - especially that one obscure box they got 40 years ago at a garage sale somewhere is the gold standard.

They don't know squat about harmonics, distortion etc, many barely understand that the blinking red light on the amp indicating clipping is not what one is supposed to see and not the signal level one is supposed to adjust the gain pot to.

But they then see that a clean solid state amplifier (or even a tube one with a lot of negative feedback and compensation) sounds different, thus it has to be the one that is "wrong" and "sounds bad".

That's where the myth about tubes sounding "better" comes from. It has little to do with actual measurable distortion value.

The engineering standpoint is the only correct one when it comes to overall audio and signal quality. The human psychology part is how people get so easily scammed with nonsense in the audiophile space

Yeah, the "human psychology" bit is how you end up buying 200$ 3.5mm cables

There's an argument to be made though that at some level it's all psychology, no? Isn't all music psychology?

I agree audiophile stuff is full of nonsense but I guess when there are differences, even if one isn't objectively better or worse, it might be good to understand the differences and history behind them because sometimes that's the nature of the thing.

Not really, no. I'm talking about signal reproduction which we can definitely measure.

https://www.audiosciencereview.com/forum/index.php

You're correct, but you're misrepresenting it as a problem. I don't have any audio hardware that doesn't distort audio in some way. I intentionally use headphones that change how my music sounds. For people who listen to music as a hobby, 1:1 sound reproduction isn't the goal and it doesn't need to be. It might be crucial for somebody editing audio, but that doesn't apply to the 99.999% of people who listen to music.

That is a patently false statement. Extensive A/B tests have shown that most people don't like a coloured sound and instead prefer a flat frequency response with low distortion. You can convince yourself that your expensive audiophile headphone with coloured sound are amazing for you, but research clearly shows that preference in relation to music reproduction is not some sort of deeply personal preference that differs from person to person. Audiophiles seem to all try and convince themselves that it is and everybody is on their own audiophile discovery journey to find the sound they like, when these audiophiles are actually subjected to a blind A/B test, they find out that they're actually much closer to the norm than they'd like to admit...

I don't know what to tell you. Given a choice between headphones with more bass and studio monitor headphones, I'll take the ones with more bass. I'll also prefer headphones that have more subdued highs, because I'm sensitive to them and they hurt my ears. I don't need a blind A/B test for that. I don't need a scientific experiment to figure out what I like. You can't always reduce human experience down to A/B tests.

Also, I wonder why you care so much about "audiophiles".

That’s why people HATE studio monitors, because they love a flat frequency response?

> amp that distorts audibly is not a good amp

Not true if you're a guitar player. (Of course, tube amp distortion can be modeled quite well digitally these days.)

There is no level of distortion that would be audible from any modern amplifier, of any design, not being deliberately overdriven. More the the point: there is no distortion behavior of any amplifier anywhere that can't be modelled beyond the ability of ears to detect by a vanilla consumer DSP. Computers are too fast and too good to make this a meaningful argument.

This is the flip side of guitar nut arguments: people swear that they have to use their analog pedals using, as you did, qualitative adjectives like "harsh". But all they really mean is that they have a specific sound out of their rig that they can't find an audio software engineer to acceptably model for them. And to be fair: that's a valid point! Software engineering is expensive. But it's got nothing to do with the capabilities or behavior of the hardware.

I can think of one other reason: Tubes can be beautifull. They glow. They literally give warmth. Tuning a radio with a cat-eye can be an irritating back-and-forth exercise, but the thing just reacts lovely to the tuning knob. Transistors and digital electronics never give that sight.

In fact, I might one day buy a tube radio where the tubes do absolutely nothing to the sound ;-)

And there's something non quantitative about sound sometimes. A LP might cracks and pops but I still feel happier with that than a 24bit high bitrate ogg/vorbis

If the room's dark enough a subtle orange glow can also be appreciated.

The only tube I have is a glow tube (VR75). I hooked it up to a variac and demonstrate it occasionally. I'd love to work with dekatrons (my initials are DEK) but they are high voltage which scares me. And learning about the history of all the various tubes is a great way to understand modern physics.

Reply on news.ycombinator.com