What to trust ear or measurement?

[richard12511]

I think you also meant to say 8k-12khz.

Indeed.

 

[thefsb]

My recollection is that in the 1950s and 60s, audio markets were overwhelmingly domestic. Then at the start of the 1970s we all raised our heads and looked around. I remember the impression that e.g. the U.S., UK, Japanese and German markets had evolved toward different sounds. E.g. the Brits were mellow and midrange-y, the Americans were bassy, the Japanese were trebly, the Germans liked what was called boom-and-tizz, which now we would call a smiley EQ.

To some extent these impressions were absolutely true and absolutely measurable. LPs were overwhelming the majority source; LPs required RIAA de-emphasis and re-emphasis; the RIAA curve, despite the best of intentions, was never properly established as an international standard. At first, each record company used its own best guess; soon, de facto national standards emerged; but always each market was measurably different.

Thus a U.S.-cut LP played through a Japanese phono stage sounded different than if played through a Fisher. No illusion. Was the "national sound" demanded by local customers and hence supplied, or was it merely supplied and hence unthinkingly accepted? Impossible to say. But it wasn't until 1983 that the whole audio world was listening to the same thing.

A lot of gear back then had tone controls. Useful.

I'm not sure. I think neither musical appreciation nor certainly musical ability is sufficiently widespread or baked-in to represent an evolved survival behavior. Not like detection of motion in the peripheral vision, for instance. I think attraction to music is a common but fairly take-it-or-leave-it mutation.

On the other hand, I wonder if a certain 20th century cultural construct still influences us today. People first got home radios in the 1920s and 30s. It's easy to forget what a huge thing that was. Unlimited news, information and entertainment in the home! As always, the leap from nothing to something is the biggest and most important. People loved their radios. Their emotional involvement was immense.

Due to contemporary technology, radios sounded warm, chesty, and midrange-dominated. I think people instinctively internalized the feeling that radios should sound like that. My grandma was proud of hers. "A lovely tone," she would say. A generation later, my mother set up home, and despite (what we would call) better alternatives, she went for the same sound. "A lovely tone!"

It's as if most people think along two parallel tracks. Talking to a friend across the kitchen table sounds one way; listening to a guy on the radio sounds completely different. Most people think that's entirely natural and correct, because of some kind of longstanding cultural inheritance. I think it explains Bose's success, for instance. Bose tapped into that strand of warm, chesty familiarity. A lovely tone!

Music is a human universal. But almost all the details are cultural. In that respect it's a lot like language. I'm personally attached to the idea that music and language machines of the brain share a lot of components.

A friend of mine years ago explained that the cello is the most compelling instrument because it has the register and tone of a patriarch. A lovely tone!

 

[thefsb]

Matt, that simply isn't true. If your preference is to go birdwatching with rose-tinted binoculars, why would the ornithological community accept your reports of bird colors as meaningful? Your viewing might be pleasant to you, absolutely, but it represents no scientific currency. It would have to be accepted as an empirically observable "fact" that Mr. Hooper personally prefers to view birds through a chromatically distorted lens, but such an observation doesn't bless Mr. Hooper with the same scientific credibility as Mr. Audubon or a thousand others. Actually it rules it out completely. "Don't listen to old Hooper," the scientists would say. "He thinks everything is pink."

Matt's right on this one. The science in question is the science of preference. Preference is its subject matter.

For example, Amir uses criteria to evaluate measurements of a loudspeaker. The criteria were derived from scientific measurements of what people prefer. The science is statistical, looking for "what the average person" prefers. It was motivated by the needs of engineers working on audio products who need a statistical view of market preference.

The science does not tell us what is the more accurate acoustic presentation of a master.

 

[thefsb]

That is a very nice question! Could be a very attractive and interesting theme for another Ph.D. theses...

Some bird species have a language of song that evolves culturally and others have a language of dance. Have you seen the natural history docos that show how it's the boy bird with the best dance moves that gets the girl?

 

[Martin]

I trust measurements to tell me how a particular component will perform, if it adds distortion or coloration. I trust my ears to tell me if I like the sound of a component.

I have yet to find an electronic component that measures well and sounds bad. There are a few that measure poorly and sound good to me. (My 300b monoblocks are likely one such component.) And there are many that measure poorly and sound like crap. I thought my PS Audio GCPH phono preamp sounded good until I played a 1 kHz tone though it and surprisingly heard a loud spray of harmonics (easily measured on the dB Meter app on my iPhone). My Pro-Ject Phono Box RS has no such problem and sounds very good to me.

Now electroacoustic transducers (headphones and speakers) are an entirely different matter. I have heard transducers that measure well but sound meh, and transducers that measure well and sound great, and vice versa. Speaker and headphone sound is a matter of personal taste. I strive for flat response and use room correction in both my home theater and will soon in my stereo system. I do not EQ my headphones.

There is no accounting for personal taste.

Martin

 

[thefsb]

I don't like to speak for other people either, so let's postulate a theoretical Mr. X, who prefers a colored and inaccurate playback chain because overall he kinda likes the mellifluous and euphonic sound it makes. That's humanly valid, societally valid, possibly musically valid, possibly valid in many, many other different ways, but a deliberate and whimsical departure from accuracy can never be scientifically valid.

You need to elaborate this proposition in more formal language in order for us to understand what it means.

For example, we can apply scientific methods to propositions such as "Among a, b, c, ..., subject X prefers a," or "Given choices a, b, c, ..., subject X will prefer a."

If we establish some definition of accuracy in the acoustic presentation of a recording in a controlled test setting and label that a, together with a number of different presentations labeled b, c, ... then we don't get to say that the preference for a among the others is the scientifically valid one. That's either circular or a misuse of the word scientific or both. In any case, the proposition would make no sense.

 

[MattHooper]

Your first para - no. It's the same as saying that someone who prefers a colored hi-fi system is abandoning scientific rigor and exiting the project, which is the unbiased interrogation of the master file presented. Which is fine, but a strange message for a science forum.

Your second para - sure, illogical choices can be studied by scientists, and reasons for them can be discovered, but understanding their derivations doesn't make the illogical choices themselves scientific.

You claimed that a preference for a colored or non-neutral sound can never be scientifically valid.

That is nonsense, as scientifically valid experiments can and are used to document and quantify preferences. Again: are you actually unaware that the research continually cited on this forum, from Floyd Toole, Sean Olive and many others, scientifically documented listener PREFERENCE?
The results are that a majority prefer speaker Type A, a minority prefer speaker type B (and C) etc. How in the world is one preference "scientific" while the other is not? It's just preferences, scientifically documented!

And what in the world makes one of those preferences "illogical?" Is it "illogical" to prefer chocolate ice cream over vanilla? Is it "illogical" to prefer a cross country trip on a motorcycle instead of flying? If someone who would choose to fly said "using a motorcycle to ride across the country is illogical because flying gets you there much faster" that person would clearly be missing something, right? Sure, if you PREFER a faster method - your goal is to get from coast to coast fastest - then the jet is the more reasonable choice. But if you PREFER many of the things you get from a motorcycle ride that you don't get from a jet flight - the feeling of riding the bike, the open air, the taking in of scenery, stopping off wherever you want and when you want, and on and on..." then riding the motorcycle is the logical choice over the jet. You don't get to just call someone else's preference "illogical" because it isn't YOUR preference and doesn't fulfill YOUR PREFERRED goal.

If someone prefers, say, a bit more lower or upper bass impact than is strictly neutral, that isn't being "unscientific." That's just another fact - a fact about what someone prefers. And given a preference for a bit more impact, introducing a bit more bass is logical to suit that preference.

If someone has a neutral system, as many here strive for, but they put on a track that was poorly mastered or produced, sounds really thin, they can dial in a bit of tone control/eq to make it sound better, more pleasing. This is just what even people like Floyd Toole will do, and many here advocate. Is this being "illogical" or "unscientific?" Do you view listening to your system to be a strictly "scientific" experiment so that using any tone controls or EQ to taste is being "unscientific and illogical?" It's bizarre.

Anyway, it seems I can not get you to see around question-begging, so....thanks for the conversation.

 

[Inner Space]

You claimed that a preference for a colored or non-neutral sound can never be scientifically valid ... That is nonsense

Matt, whatever. I already agreed that the act of arriving at a preference can of course be studied by scientists or researchers or anyone else. And that such a preference is what it is and can't be gainsaid. My point, which clearly I'm not communicating very well, is (to continue your analogy) that if scientists studied preferences among members of a motorcycle club, and found an outlier whose preference was to stay off the bike altogether and fly instead, they might think that was kind of weird, and that such a guy doesn't really belong in a motorcycle club. I think a guy with a preference for colored and inaccurate sound doesn't really belong on an audio science forum. Simple as that.

But I'm new here, and I don't want to wear out my welcome on this thread or this site, so I'll bow out now.

 

[snapsc]

as thefsb stated: "It was motivated by the needs of engineers working on audio products who need a statistical view of market preference".

And who want to be able to take their measurement/preference correlations into account when designing future product thereby improving the chance that the product will be well received by the market...as validated by sales figures.

 

[RayDunzl]

Some bird species have a language of song that evolves culturally and others have a language of dance. Have you seen the natural history docos that show how it's the boy bird with the best dance moves that gets the girl?

 

[Wes]

Some bird species have a language of song that evolves culturally and others have a language of dance. Have you seen the natural history docos that show how it's the boy bird with the best dance moves that gets the girl?

some of you may enjoy a Gloggle search on "bowerbirds"

 

[thefsb]

Trust you to have the best youtube for that. Old low res. But that video is the balls.

 

[RayDunzl]

Trust you to have the best youtube for that. Old low res. But that video is the balls.

It's from a National Geographic episode.

The clip is just a pertinent part someone grabbed (at low res).

I like the nice lady narrating, too.

 

[RayDunzl]

Ah ha....

I don't think M.J. would have objected, himself.

https://www.nationalgeographic.com/news/2013/10/dancing-animals-spiders-birds/

 

[thefsb]

It's from a National Geographic episode.

The clip is just a pertinent part someone grabbed (at low res).

I like the nice lady narrating, too.

Right. If it weren't for her and her lovely dancing it would be just another nat his doco.

 

[richard12511]

A lot of gear back then had tone controls. Useful.

Music is a human universal. But almost all the details are cultural. In that respect it's a lot like language. I'm personally attached to the idea that music and language machines of the brain share a lot of components.

A friend of mine years ago explained that the cello is the most compelling instrument because it has the register and tone of a patriarch. A lovely tone!

Some other animals have shown to enjoy music, too, especially classical.

 

[Gregm]

I have to laugh because one of the articles sited contains this quote:

(...) I also think that we can measure things to an extent that it doesn't matter (beyond our hearing capability) and also that there are things we can't yet measure that we can hear.

Alternatively, we simply have not measured the specific characteristic we heard, or no-one gave us that specific measurement; dispersion characteristics of a speaker, for example. Regards

 

[Ceburaska]

Right. If it weren't for her and her lovely dancing it would be just another nat his doco.

The bird was awesome. Could have done without her.

 

[magicscreen]

I have yet to find an electronic component that measures well and sounds bad.

Jds Labs Atom

But I suspect that all well-measuring device sounds terrible.
The huge negative feedback demolishes sound quality.

 

[mansr]

Jds Labs Atom

But I suspect that all well-measuring device sounds terrible.
The huge negative feedback demolishes sound quality.

Uh-huh.