The Emperor's Old Clothes: In Theory Is Where I Do My Best Practice

An article titled, "24/192 Music Downloads...and why they make no sense", on xiph.org is getting a lot of attention on the audio forums because it claims that 16/44.1 provides better sound quality than 24-bit/192kHz. The author, Chris Montgomery, has also cleverly titled the associated file "neil-young.html" as a wink toward Mr. Young's recent calls for a move away from MP3s toward higher resolution formats. It's also worth noting that according to Mr. Montgomery's Wikipedia page, "Christopher “Monty” Montgomery is the creator of the Ogg Free Software container format and Vorbis audio codec and others" so one can perhaps see why he's taken what Neil Young said so personally.

This is another instance where an AudioStream forum would come in handy since this is certainly worthy of discussion even though Mr. Montgomery feels he's put an end to this conversation with a tidy, "Covered, done." I suppose if he was able to dictate the format of the music I listen to, his opinion on this matter would mean more to me. And while I appreciate his brief tour of the human ear, the Nyquist–Shannon sampling theorem, amplifier design, a proposed means of better reproducing ultrasonic information, and advocacy of double-blind and ABX testing, perhaps the most tasty of all is this telling tidbit:

It's true enough that a properly encoded Ogg file (or MP3, or AAC file) will be indistinguishable from the original at a moderate bitrate.
That's that personal part I was referring to earlier and I completely understand taking pride in ones work. But let's not confuse the fact that some recorded music can sound indistinguishable at lower bit/sample rates with the notion that all music will follow suite. The original recording is one important key in this equation as is the playback equipment and the listener's level of interest in and ability to discern a difference. To suggest that the hearing ability of all people can be boiled down and fit inside one tidy container is missing not only the subjective nature of the enjoyment of music but some facts as well.
Thus, 20Hz - 20kHz is a generous range. It thoroughly covers the audible spectrum, an assertion backed by nearly a century of experimental data.
I'm probably the last person that should try to refute Mr. Montgomery's claims on a scientific basis since my background and education are not science-oriented. So let me simply point out that there have been studies which show how and why people react to ultrasonic information through bone conduction (see Peripheral perception mechanism of ultrasonic hearing, Tadashi Nishimura, et al.). Here's a quote from the highlights of that research paper, "Excitation by bone-conducted ultrasound fast spreads to the low frequency range." The researcher's also point out that people's perception of ultrasound through bone conduction varies from person to person but the more important and interesting part is I do not see them call everyone who does not agree with their findings shysters or charlatans as Monty has done in the introduction to his blog post.

And let's add that no one disputes the fact that instruments, even acoustic instruments produce frequencies outside the 20Hz - 20kHz range. So if people, some more than others, perceive ultrasonic information through bone conduction when they go to a concert, it would follow that if you lop off this information some listeners may very well notice the difference. Of course this is pure speculation based on a few things I've read and a few things I've heard so I'm not suggesting this conversation is covered or done.

And let's also add this quote from the section titled "Genetic gifts and golden ears" of Mr. Montgomery's blog post,

There was a time in the 1990s when I could identify every major mp3 encoder by sound (back when they were all pretty bad), and could demonstrate this reliably in double-blind testing.
I always wonder why double-blind testing proponents who poo-poo on the concept of "Golden Ears" always feel obliged to point out that they have a pair. In any event, it will be interesting to see if every recording engineer, record producer, hi-fi manufacturer and even Neil Young revert back to good enough old 16/44.1 once they read what Monty wrote. It sure would make things simpler.

The good news? Tyll Hertsens over at InnerFidelity will be happy to hear that Monty's answer is—better headphones,

The easiest fix isn't digital. The most dramatic possible fidelty [sic] improvement comes from a good pair of headphones. Over-ear, in ear, open or closed, it doesn't much matter. They don't even need to be expensive, though expensive headphones can be worth the money.

Keep in mind that some headphones are expensive because they're well made, durable and sound great.

Hmm. So we can base a headphone purchase decision simply on something that sounds great? While I agree (but prefer listening to speakers), that sure sounds awfully unscientific and totally subjective. And what about the associated amplifier? And source? And, dare I say it, cable? Certainly each piece of the playback puzzle matters so you could spend all the money you have on "good sounding" headphones and they'll sound beat if you connect them to crap. Especially if you only listen to lossy compressed music.

In closing Mr. Montgomery explains why he felt compelled to write this piece,

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.
I got the impression from listening to Neil Young that he feels high resolution audio just makes recorded music sound more like the version he recorded. But that's just based on Mr. Young's personal experience and we all know you can't trust a musician when it comes to talking about the sound of music. I doubt Neil Young ever even tried to write an audio compression algorithm.

I will point out that I listen to and enjoy higher resolution digital music as well as LPs and electronics that use vacuum tubes. Which clearly means that Monty and I take different approaches toward the same goal of enjoying music.

COMMENTS
deckeda's picture

 

Or perhaps it's the other way 'round. That was the best conclusion I could come up with when I first read the article a few days ago. I wish I'd had the insight to make obvious the various contradictions, as you've done.

 

I'm no expert on debating, but it strikes me as risky to promote a position so full of absolutist statements that sooner or later they'll wrap around and bite you in the ass.

 

Hey wait ... did someone say, "bone conduction?" (My confession is that I'll always remember it as the Boob Fone. Some ads in the stereo mags you just don't forget easily, especially when viewed as an impressionable youth.)

Michael Lavorgna's picture

Excellent!

& I didn't get your confession until the second viewing (OK, I'm distracted) but when I did get it, I laughed out loud. Thanks for that.

Vincent Kars's picture
deckeda's picture

First my Bone Fone reference, now this?

from the Danish-to-English word-salad translation linked to above,

One important factor first pointed out by Mr. R. Blake is a pheromone called Aliasing Intermodulation Distortion 

Hey, over the years I've called it a lot of things, but never that.

labjr's picture

The CD vs Hi-rez argument goes on forever in many forums.  Apparently(to me anyway), for whatever reason, it's easier to make 24/96 or 24/192 sound better than CD, even on inexpensive players. CD has been improving for 30 years but still doesn't sound as good( to me) as what I hear from higher resolution material. So why not embrace it?

Now when I listen to Neil Young's Harvest Album on DVD-A, I feel like he had a golden ear and the smarts to think way ahead because that sounds like it was recorded yesterday. So I'd take his word that it sounds better than CD before I listen to some blogger.

If you like CD than by all means, keep buying them. But I'm going with higher resolution whenever I can.

dalethorn's picture

I take it the double-blind test proof is based on the notion that you can consciously hear and identify everything that's important in the sounds you're given for the tests. Well, somehow that just doesn't seem logical, and it appears Mr. Young figured that out a long time ago.

firedog55's picture

Michael-

 

Unfortunately, my online research tells me that the Tadashi paper hasn't been backed up by anyone else. In fact, the opposite. Scientific results need to be repeatable to be significant. I don't think this one is.

But I don't get the constant emphasis on the 20-20k frequency thing. Maybe that has nothing to do with why hi-res sounds better to us. There are other aspects to music listening that aren't simply a result of the primary frequency we are hearing. 

One of the reasons many of us prefer hi-res is that it gives us a better sense of space - both between the instruments and of locating the music reproduction in it's space (the room, hall, etc). I'm not sure this is directly related to how high of frequencies we can hear. And I haven't seen much, if any, research on this.

X