Hacker News Reader

Being able to enjoy something even though you know something better is out there is something that I've managed to achieve in many areas of my life.

For example, I love a great beer. I'll go out of my way to try something new. But when I'm sitting in the sun watching baseball with my wife, I'll get a Budweiser with my hot dog and it will be glorious.

There was a time in my life where I would complain about the lack of craft beers at the ballpark. I was an idiot and enjoyed life less.

Audio is subjective but differences in quality can still be objectively measured. You play the same music through different equipment and have the listener rate them without knowing what they are. Consistent improvements will show up across different listeners, and differences that come down to personal taste will still show up across different sessions with the same listener.

The most telling thing about the audiophile world is how allergic they are to anything that resembles this sort of blind testing.

Right but, people often match volume/level when testing so things sound different, because +3db tends to sound better to the human ear.

Some things are very hard to measure objectively, though, despite being real. (Anything subject to chaos theory, for sure.)

I mix records. I use two different sets of monitors for mixing - a pair of Tannoy System 12 DMT monitors (equivalent would cost about $4000; they were very popular in hip-hop studios back in the '90s), and a more modern pair of Focal Alpha 50 powered monitors (about $800/pair new). Both monitors are quality professional gear. In terms of frequency response, they're both quite flat, although the big Tannoys have more bass extension.

They don't sound anything alike.

I find the Focals much more useful and predictable for the critical listening of mixing. They're also more tiring to listen to for extended periods. They're brutal, especially for transients and the very leading edge of sounds. I don't trust mixing on the Tannoys. But if I want to listen to music, rather than mix, I'll take the Tannoys any day. They treat transients more gently, and make for a much more pleasant, euphonic experience.

Internet audiophile wannabes will then tell me I'm hearing distortion. Which is nonsense. Two speakers say the same thing in different ways, and the differences are very, very difficult to measure - but they are nonetheless quite real and plainly audible.

What you're talking about absolutely is measurable. There has been tons and tons of research into various aspects of audio reproduction and how to measure different things and to what degree they're audible. Techniques for measuring transient response have been around since the 1940s.

Unfortunately most places writing reviews of audio equipment don't bother to go much beyond frequency response and distortion. Audioxpress ran a really nice 2 part article a decade ago describing a bunch of different types of speaker measurements and summarizing some research into how audible each aspect is.

https://www.audioxpress.com/article/testing-loudspeakers-whi...

I did transient response measurement while studying recording in college (onto recording media). But when you start getting into things like frequency-dependent phase response, radiation patterns, reflections... there's a tremendous world of logically audible speaker behaviors that could maybe be measured in isolation, but you get so many of them that you can't assemble a meaningful picture of how a speaker is going to sound.

And that's the goal, isn't it? Measurements provide expectations. And that's where the conversation breaks down - the "objective" mind gets so preoccupied with the measurement that they assume what isn't measured cannot exist. The map replaces the territory, and there are no hills because the map is obviously flat.

> But when you start getting into things like frequency-dependent phase response, radiation patterns, reflections... there's a tremendous world of logically audible speaker behaviors that could maybe be measured in isolation, but you get so many of them that you can't assemble a meaningful picture of how a speaker is going to sound.

These things aren't particularly difficult to measure, we even have standard plots for them. Polar response, energy time curves, cumulative spectral decay, electrical phase and group delay are all standard measurements for a loudspeaker.

Yes, but can consumers look at those and derive meaningful expectations for their subjective experience of using the speakers? And how many of those things are sensitive to dynamics, volume, signal complexity, etc?

I think this stuff says a lot more about psychology than it says about technology.

When I'm mixing music, my "final arbiter" for mix quality is the crappy stock speakers in my car. A mix may sound great on my studio monitors, but if it's not happening in that extremely inaccurate car system, it's not happening. A lot of professional mixers use "bad" speakers like Auracubes or Minimus-7s just to emulate that experience.

And of course, this kind of experience is also inseparable from context. It's not just that they're car speakers... it's that they're in a car. They're in the door, not aimed at my ears. Their bass-reflex behavior is from the volume of the door and whatever venting it has, and reflections are completely crazy.

You can measure all the differences in TV picture you described (and you didn't even mention colour gamut & dynamic range). You can measure the differences in optical glass. You can't measure the difference between a $10/m speaker cable and $10000/m speaker cable (usually... some of the expensive speaker cable is so crazy it may mess up the sound), or an amplifier sitting on a $5000 stand. And $1000 power cables, plugging into your usual wall outlet...

We can measure the difference in cables though! I can put them through a scope or analyzer and I can even show you real hard differences between them.

What we can't do is prove that makes a difference to your ear vs mine and that's the real problem.

So there are two difference classes of scam here. The one where it's true there is a difference in this $10 and $10000 cable, but they're effective identical. And the scam where this is no measurable or even logical difference at all.

It's the former scam that I think allows the "audiophile" industry to thrive.

At audio frequencies (and well beyond), you won't be able to measure any meaningful differences: https://www.lifewire.com/speaker-cables-make-a-difference-31...

And the absurdity becomes much more obvious when contrasting with the differences in speakers and speaker crossovers...

That depends on impedance behaviors between components, too.

As a guitarist, I did guitar cable listening tests. With passive guitar pickups (high impedance, low output devices), different cables have easily audible fingerprints. With active pickups, there's no difference at all. The difference between cables is no longer enough to audibly affect the signal.

And within this realm, more expensive doesn't always mean better. I've had cheap cables sound better than expensive cables.

That's a point, I was assuming low impedance output stages as would be standard for any ordinary hifi equipment these days (but not necessarily for the really expensive stuff of course...).

Some of the expensive cables are over-dimensioned or crazily constructed leading to excess capacitance and/or impedance which would lead to terrible performance with a high impedance driver.

Impedance is a difficult and complex thing (which is why failure mode #1 for dudes on the internet being all "objective" and "scientific" is failing to grok the difference between impedance and resistance). So how an output stage is constructed (or an input stage, for that matter) has a lot to do with its behavior in the face of 10 octaves of frequency range and 50+ decibels of real-world dynamics.

Is it transformer-coupled? (Rare in hi-fi, common in pro audio) Capacitor coupled? Direct coupled? Differential? All of these have different and frequency-dependent capacitative, inductive, and resistive behaviors. Put in those terms, it's kind of a no-brainer that cables would matter.

But in "objective" internet-land, it's all straight wire with gain, perfect flat resistance. Which is fine on paper...

There are a lot of areas in media production and consumption where you get into seriously diminishing returns and it's not hard to spend 10x, 100x, or more for differences in quality that require expert use and a discerning consumer to really appreciate the differences or even notice them.

It probably makes sense to go first class when you have professionals with big budgets who are trying to get every detail just so. All the very high-end microphones and lighting gear probably make less sense for a shoestring budget podcast or YouTube series.

Especially because measurements like Total Harmonic Distortion (THD) use sine waves for their measurements. People pay thousands of dollars for marginally lower THD (say, .02 % instead of 0.5%) but they listen to music, not sine waves.

Well, I assume this. Perhaps people buy high end audio equipment so they can listen to sine waves.

This is the reason for measurements of intermodulation distortion and THD+N by frequency. A proper evaluation of, say, a power amplifier will look at a dozen, maybe a couple dozen different measurements; taken all together, these provide a reasonably complete picture of how the amp performs.

But - from an engineering perspective, music can in fact be decomposed into a series of sine waves - and effectively all digitally recorded music goes through this process as part of a/d conversion (sort of; it's complicated) or FFT-based digital equalization. See: https://en.wikipedia.org/wiki/Fourier_analysis

THD is a measure of non-linearity, which adds harmonics not present in the original signal. Yes, the standard tests use sine waves, but the results are applicable to other signals. Just because most people don't listen to sine waves for fun doesn't mean it's a valid measurement.

On the other hand, the highest THD in a sound reproduction system comes from the physical transducers. Worrying about the THD of an amplifier makes little sense when the speakers provide an order of magnitude more distortion.

-I fully agree with your sentiment, except I'd like to stress that a loudspeaker provides _several_ orders of magnitude more distortion than a competently designed amplifier driven within its limits.

I used to live on the Avenida Paulista in São Paulo, Brazil and one of the interesting quirks about that location is the number of high power TV and Radio antennas in the immediate vicinity. At the time I had a hi-fi system that had previously sounded great, but in this particular location was impacted by the high electromagnetic interference in the area. Most of the noise got in between the RCA cables between the pre-amp and the power amps.

I spent the better part of a year building cables and trying out different designs to isolate the noise, such as having a floating woven metal shield with a small capacitor at one end to ground the shield. I don't remember all the different things I tried, but they definitely had varying levels of material impact. That said, with RCA cables I never got the noise completely out of the system and eventually switched everything out for something with balanced XLR connectors, which eliminated the problem.

tl:dr; expensive cables don't really matter. cable design does to a point, but if you really want great sound without interference, you need to switch to a system with a balanced design. Unfortunately, this requires changing all components, not just the cables.

https://en.wikipedia.org/wiki/Balanced_audio

A microphone is just a camera for sound.

>Its tough though... I know of a couple times in my life I've thought something was OK until I saw something else. Examples, Glass. Binoculars, camera lenses, rifle scopes, telescopes. Glass differences are very real.

-Aye. I had a rather absurd discussion with a colleague here the other day; he keeps a pair of el cheapo grande binoculars in his office, some 10x50 monsters which he proudly announced cost less than $100.

The colour cast was so bad that even I (being tritanope) immediately spotted it and found it annoying. Colour fringing like you wouldn't believe. A tiny, reasonably sharp spot in the centre of one pupil, the other was nowhere near acceptable. Light falloff and softening was quite aggressive as you approached the edges.

I then brought out my Zeiss Victory 8x32s, as they happened to be in my office that day after a field trip. His first impression? 'The colours are off, and it looks unnaturally sharp.'

Sigh. Never mind that one could read the (comparably) small print on a sign across the sound on the 8x Zeiss whereas even the headline was a smudge on the 10x cheapos.

At the risk of sounding elitist, I think that for some people, low cost makes for excellent quality in the same way as high cost equates same for others.

Apparently, at least a decade ago, a lot of people prefer the sound of poorly-encoded MP3s: http://radar.oreilly.com/2009/03/the-sizzling-sound-of-music... There's no accounting for taste.

There is certainly a range of quality in audio gear, but it's a complicated thing.

Some things that are reasonably easy to notice with a stereo system:

1) Dynamic range: how loud does it get, how quiet does it get? Does it sound good when its loud? does it sound good when it's soft? Small/inefficient speakers will struggle at high volumes, resulting in high distortion. That's measurable and often not hard to hear. Big speakers with the powerful amps needed to drive them which sound good at high volume often suffer from audible hiss at lower volume, short distance listening, due to inadequate power supply noise rejection relative to the gain. Big speakers also will generally have further-spaced drivers, and that bigger geometry can have a noticeable effect on spacial coherency, especially at short distances. (Ideally you want your speakers to be a point source, not different frequencies coming from different angles)

2) Flatness in frequency response: this is easy to measure with equipment, but also quite easy to hear by playing a sine wave that slowly sweeps the audible range. If the amplitude of the wave in the signal is consistent, there shouldn't be a ton of variation in volume of the reproduction. It can be heavily effected by the room they're in, usually in the bass & mid-bass range where the wavelengths are similar to room dimensions, but some speakers will be much more consistent than others across the rest.

3) Dispersion characteristics: when you move your ears around relative to the speakers, does the frequency response change dramatically? Some speakers don't sound very good in the high frequencies once your ears go a few degrees off axis from the source.

This stuff gets a lot more obvious when doing A/B comparisons. You're right that you can't really tell how good a system is just by listening to a random song and judging it's sound quality like you can with video on a TV, and I think that stems from an inherent difference between hearing and vision. The sound you hear is ALWAYS heavily effected by the space its in, reflections and sympathetic vibrations, the angle you hear it from, and so on. Source material varies wildly in quality and recording/mastering technique. It's much harder to perceive "wrongness" in audio because it's so varied to begin with.

Power response is a really fascinating metric, and people don't bring it up very often.

The interesting thing about power response is that it's one of those things that you can't fix with an EQ.

IE, if I have a speaker that's too "hot" or it's too rolled off, I can address that via EQ. But if your power response is bad, there is very little you can do about it.

It IS possible to improve the power response by tweaking the crossover, but that is one of those things that separates the good engineers from the great ones: nearly anyone can make a speaker that measures flat on axis. The hard part is getting it to behave on axis AND off.

If you look at brands like KEF and Genelec, they've paid a lot of attention to this.

If swaros are low-highend, what's high-highend?

The bring up a different discussion about highend and impractical highend. Swaros cheaper models are practical. They're well built, they're expesnive but not outrageous. They're REAL products.

There are high-highend products that AREN'T REAL. Things that come out of shops that sell 100 all year, things that can't be used like normal products. Etc.

In this case, Miyauchi Binos or something like that. It's a bad descriptiong I made because you have have practical highend and impractical in the same brand.

Maybe a good test is if you can buy them at typical store vs only a specialty store. My local sporting goods store has Swarovski optics, but they top out at $2,000 because nothing higher would sell - after that you are in the point of diminishing returns anyhow.

Thank you, that was a helpful answer.