Hacker News Reader

I have a hard-won TDS meter that lets me measure the total dissolved solids in a cup of coffee.

Like music, coffee is an entirely subjective experience. Like music, distilling it down to data does _not_ implicitly destroy the experience. Judging a cup of coffee good or bad solely based on data is impossible. Maybe this cup is intentionally overextracted and that song is intentionally overmastered. That doesn’t guarantee that the coffee or song are good or bad, it simply helps you understand why you do or don’t like it.

After making a thousand cups of coffee with a TDS meter, I can predict what would make a cup of coffee better without needing the meter anymore, and I’ve learned that I care more about enjoying coffee than I do about perfecting it.

If you sat through a blind test of a thousand songs, and at the end discovered that you _can_ distinguish 16/44 from 32/384, you might still _choose_ not to care. Most people don’t want every cup of coffee to be competition grade because it’s really expensive (density sorting), really difficult (dual-wielding flow-restricted kettles), and the payoff isn’t worth it every day to them. Maybe that’s how you’d feel after A/B testing 1,000 songs, as I did after pouring 1,000 cups. Maybe not.

For most people, knowing that 16/48 is indistinguishable or better than 32/384 will save them thousands of dollars and hundreds of hours of audio setup, tuning, design, repurchasing, etc.

For a few people, it’s worth it to them to go competition, either in coffee or in music. That’s certainly their right, but it’s not at all guaranteed to make them any happier than they would be with 16/48.

So you think our senses transcend what tools have the ability to measure? Maybe that was the case in 1970, but in 2000+, hearing (and vision) is completely understood scientifically and far surpassed by measuring apparatus at every frequency range. Saying otherwise is an appeal to what is called audio mysticism and is caused by placebo and confirmation bias, which was mentioned in the article.

Yes, that's what I think. I don't believe that hearing is completely understood scientifically. You're welcome to disagree.

Yeah, this is what I can't fathom. We can detect exoplanets with telescopes with 0.01 arcsecond accuracy, hear whale calls 100 miles away with an array of low frequency transducers, and model the shape of structures by forcing seismic vibration and measuring the response to 1 part in 100 million. We can image the HIV virus and detect DNA changes due to cancer. It's unlikely that we don't understand everything we need to know (or even 1000x more than we need to know regarding this debate) about cochlea, a macroscope object that undergraduates can study with a primitive microscope.

I don't mean to shift the goalposts, but neuroscience isn't completely understood, and that's what I mean by hearing and how music makes you feel.

Completely true, but the auditory nerve only "knows" what the mechanical part can respond to. If we don't understand how sound information is interpreted by the nervous system, we can still answer the question "is any neural/electrical/chemical signal whatsoever produced by a stimulation of a frequency or amplitude?"

Actually that's a defensible position when you're mixing and mastering audio. A spectrogram won't tell you which settings sound better, but your ears will.

Maybe not relevant in the context of GP's post though. In the context of digital tools telling you two signals are identical, then I suspect they are, and if you want to prove to me your ears hear better then you're going to need a blind A/B test.

Right, this debate is not about determining whether two different audio clips A and B sound better. It's about determining whether two clips can be distinguished by our ears at all.

But I don't care about proving anything to you. I don't need you to believe me for me to enjoy my music. You can tell me you prefer coffee from Brazil to coffee from Honduras and we can talk about what characteristics you notice, not whether or not you've done a blind A/B test.

Certainly, enjoy what you want to enjoy. In response to your original statement here:

>How does it affect your life that I listen to music encoded at 24/196?

It doesn't, but Monty contributes to the Xiph project, which constantly has to deal with thousands of people saying "why doesn't Opus/Vorbis support 192 kHz??" This is his answer to that general question. There is actual money on the line for companies that use audio codec technology, and supporting useless formats that have no psychoacoustic effect is a waste of time, money, and effort.

If it was true, the entirety of hi-fi industry would not exist. I myself built a number of amplifiers and, after a certain threshold, roughly .02% THD (total harmonic distortion) at 20khz, there is a very little correlation between the THD (what is usually measured)numbers and perceived quality of the sound. Which means, while it is true that is everything could be measured, no one measures the right thing (perhaps some weird subtle phase shifts or some almost immeasurable frequency response deficiencies)

Are you suggesting that things like healing crystals work, because the market for them exists?

You are putting words in my mouth. I am telling what I actually verified. Amplifiers with lower THD and IMD often sound worse than those with less impressive meausured parameters. Ergo, the measured parameters are not relevant. We need to find the actually important paramters and measure them.

But “sounds worse” in this case is relevant only to your preferences and to the preferences of some N of listeners, where N is totally unknown. The parameters you believe amp manufacturers should measure are therefore only relevant to your idea about them.

THD, SNR, frequency response and many other metrics are easy: they define either accuracy or precision. If you want more than that, add external affects hardware or DSP. The purpose of the amplifier is to amplify.

It is relevant to a very large number of people, and that is why it is reflected in the price of the equipment. A very well known engineer, who has not ever peddle snake oil, Nelson Pass, sells his rather primitive audio equipment at high prices. Jusdging

You seem to be highly inexperienced in the subject, THD, IMD etc are parameters that are measured on highly artificial signals - singke sine waves or mix of small numbers of sine waves. Real music signals more resemble white noise than clean sine waves, and there might be hundreds of different very subtle modes of distortion, that are very difficult to measure but which have very significant influence on the precepted quality of the sound.

It’s not clear to me how one person selling boutique equipment at high prices is indicative of any flaw in measurement methodology. There are many companies doing this, of dubious real value to consumers. How is Nelson Pass different, and how is that relevant to what we’re discussing?

> the entirety of hi-fi industry would not exist.

An industry which is known to sell an enormous amount of snake oil...

You are talking about hi end. Not hi-fi. Now do you have anything of substance to say?

> [I]n 2000+, hearing (and vision) is completely understood scientifically and far surpassed by measuring apparatus at every frequency range

Yes, we have instrumentation better than the meat-based transducers in the human body. That doesn't mean we completely understand said meat-based transducers or how that meat-computer in our skull interprets signals from those transducers. The auditory and visual systems are still subjects of active research with many outstanding problems.

I mean, I'll be among the first to call bullshit on audiophile snake-oil like people pushing 192 kHz/24 bit music, but in your rush to discredit such things, you've gone too far in the other direction.

Sorry, I read your comment twice... do you actually dispute any of the points made in the article, or are you simply wishing that people didn't care?

I'll just respond by quoting the article:

Why push back against 24/192? Because it's a solution to a problem that doesn't exist, a business model based on willful ignorance and scamming people. The more that pseudoscience goes unchecked in the world at large, the harder it is for truth to overcome truthiness... even if this is a small and relatively insignificant example.

This was basically a PSA back when Neil Young was pushing the Pono music player and store. https://en.wikipedia.org/wiki/Pono_(digital_music_service)

Tidal still sells on this basis too.

If you really had any qualifications, that anti-scientific rant made the irrelevant.

Was this comment machine-generated? Every comment by your account is negative and unsubstantial.

> not "unexpected 14.5kHz falloff"

And yet just yesterday I was reading an audiophile magazine review that discussed a pair of $4,000 headphones and how the "transition from midrange into treble had something of a falloff that was confirmed by their equipment".

Music lovers listen to stereo equipment to hear their music. Audiophiles listen to music to hear their stereo equipment.

If two wines were chemically identical but someone claimed that one of them tasted “warmer” than the other, would you just go with that?

Sure. I don't need to invalidate someone else's experience.

What if they were wasting a bunch of money on it and trying to convince others to do the same?

How is it a waste if it's one person spending their money on something they enjoy? You're not compelled to also do that.

A reduction in their ignorance would make them enjoy the other choice just as much, but for less money. They’re spending more just because they don’t know the facts. I consider that a waste.

You're projecting. You think someone doesn't know the facts, but they know for a fact that they personally enjoy some certain wine or speakers or whatever. That's not ignorant.

If the hypothetical person knows that the wines are identical but still buys the more expensive one and enjoys it more, then fine. Does that really happen?

Even if they’re ignorant, I don’t care how they enjoy themselves. The problem is when they start writing guides and reviews encouraging others to follow their ignorance.

The is not the only scenario. How about you made some analysis of two wines and they seem to be same, yet someone can _reliably_ tell them apart?

As long as the test is done with proper blinding, then that means the analysis is insufficiently comprehensive and there really is a difference.

Is that the case here?

Yes, that's what I and the person you responded to are arguing. See this comment [1] and the video I linked for more.

[1] https://news.ycombinator.com/item?id=19320093 [2] https://www.youtube.com/watch?v=RlCG2fK-abo

Really? I understood your comment as explicitly disavowing blind tests, and I see no mention of them in the comment you linked to.

I’m not really inclined to watch a 25 minute video on this. What’s the timestamp of the blind test?

There is the law of diminishing returns on audio quality, as in everything. Empirical data collection is not a thing to be avoided however.

Given that, your rant seems misplaced, and I can't help but think engineering is not the field for you.

Umm, Nyquist-Shannon Theory? Listening to 24Bit 192Khz can actually damage the cillia in your ear

Bingo! Exactly, Nyquist theory. But no one can make a truly good, brickwall filter at this frequency. It is much easier to make a lot less steep frequency response, whilch would start at 22 khz and would slowly go down all the way to 96 kHz. By the way, how would it damage the ears if it was thoroughly filtered?

Except that audio DACs are generally oversampling; for example, an output filter for the AD1955 DAC is only needed to reach -3 dB at 100 kHz, even if you are feeding it a 48 kHz PCM signal. This allows these DACs to produce essentially zero energy in the frequency band from Nyquist cutoff of the input signal to filter cutoff.

This resampling adds some latency, right? Ideal resampling adds infinite latency---the sinc function has infinite support.

Yes, but it's not much. To stay with the example given above, group delay is less than 1 ms at 48 kHz fs, largely indepedent of oversampling factor (2-8).

Resamplers these days take advance of our faster computers to use near-brickwall filters, like a 1024-point windowed sinc. With this, you can create a steep filter from 20-22 kHz with better than -100 dBr attenuation beyond the Nyquist frequency. This is enough to stop worrying about the aliasing caused by the bandpass filter.

I've never heard of what newaccoutnas is talking about, but I guess if a high-frequency sound is ridiculously loud, the energy could be absorbed somewhere in the cochlea. I've only heard of experiments of subsonic hearing damage, not supersonic.

Do you know what frequency is typically for the resampling? Anything like 192kHz?

Do you mean what sample rates mixing/mastering engineers work in? They're usually 44.1 or 48 times 1, 2, or 4. Software/hardware DSP effects sometimes oversample interally up to 16x. If you're asking what the ADCs use, they usually have an analog filter circuit for each input, sample at 4 or 8x to digital, and then digitally decimate to 1x, all in one $5 ADC chip.

Well, there are several things to keep in mind. 1. The sharper the frequency response the worse ringing in the phase response. 2. Human hearing is a mystery and the near ultrasonic freq may actually have some influence on the perception of the music. 3. You need to use analog filters before the ADC and after the DAC anyway, and it is so much easier to implement with a much less steep freq response, if used with higher sample rate. I mean, you sample it at 192 and the filter it down to remove anything above 96. There is no point to resample all the time up and down.

That's true for IIR filters but not FIR filters, which modern (This century) signal processing uses. A 1024-point windowed sinc will have linear phase across the entire passband, that's one of the main reasons we use FIR filters.

Listening to 16bit 44.1 kHz can also damage the cilia in your ears. Whether it actually does, in both cases, of course depends on the recorded material and the playback gain.

> "Health is an entirely subjective experience and trying to distill it down to data is both condescending and telling of how little an author cares about medicine."

Would you accept that argument from someone purporting to provide medical advice?