Pitch Perception Skewed By Modern Tuning

The feed deliverers us news of research suggesting that the use of A as the universal tuning frequency has made our ears less discerning of the notes immediately around it. Here's the abstract from PNAS describing research with people possessing the rare quality of "absolute pitch."

Frist Psot?

[mcrbids]

It's interesting that pitches can be amalgamated by experience. Which is a basic part of human nature - the mind adapts to fit circumstances, and if the key of A is what we tune in to, why wouldn't our minds adapt to fit this reality?

It's all how it works. The article is weak on details, but this post is probably bigger. If every time you heard a sound like a jet engine, you got smacked upside the back of your head, wouldn't you get jumpy when you heard anything that sounded like a jet engine, even if it wasn't *exactly* the same?

Sometimes it's funny how Science has to prove the stuff that "Everybody Knows". (TM)

Re:Frist Psot?

[Incoherent07]

Adding to the shrug factor, the twelve-tone pitch system as a whole is a human invention. This makes perfect pitch that much stranger, because it means people have an innate ability to attune themselves to an artificial note naming scheme.

So since that scheme can vary somewhat, it would make sense that depending on "which" A your perfect pitch is tuned to, you may have trouble distinguishing G# or A# in a different tuning.

Re:Frist Psot?

[plams]

No, no, no! Twelve-tone pitch is derrived from perfect intervals, such as perfect thirds [wikipedia.org], fourths [wikipedia.org] and fifths [wikipedia.org]. These can be defined very cleanly as the integer ratio between two frequencies (look up just intonation [wikipedia.org]). The ratios are mathematically beautiful and simple, and also sound particularly good. The temperated (12 note) scale used by nearly all instruments today is an attempt to fit these intervals into a common scale. You may say that this approximation is a human invention (even though it's cleanly defined as freq = 440hz * 2^(n / 12), where n is the semi-note distance from A4), but as a whole? No.

In other words, it proabbly wouldn't make any sense to use a 16 note scale or something like that. The 12 note scale has roots in something very mathematical, not something random or "human".

Our Ears?

[DynaSoar]

I don't see anything in TwholeFA that says anything about "modern" tuning as opposed to non-modern. The choice of A is arbitrary. Until this is replicated using different notes as the target, they've got too much confluence of musical memory and their theoretical genetics to do more than use the conclusions as the basis for more work. (And what good grant-using researcher doesn't; a PNAS publication makes that very easy for them).

The use of "none were musically naive" is a poor operational definition because it's too vague. Better to use "professionally trained performers with X years performance experience". Those with a lot of listening exposure and only enough performance experience (even if just by themselves) makes it likely that those with true AP and those with relative pitch (RP; being able to tell a pitch compared to another) are mixed together. The latter can have an extensive musical memory and be able to compare a presented tone with a song in memory that they know is in a certain key. They may well have done so, because they included at least one subject with skewed scores that were very consistent in their skewing (always one sharp off) as an AP subject.

The memory problem will probably also come out if they replicate this (as they suggest) with people from other cultures. Those who come from cultures with tonal based languages are going to have a very good tonal memory and discrimination from any given starting note and so good RP.

I'm highly suspect of a 44% sample of AP. I used the more rigorous definition of musical experience in brain imaging experiments and had about 15% true AP among them. Many of those claiming AP had good RP, and their EEG showed more memory than auditory activation, just as those claiming and having only RP. I'm also suspect of getting the same results from sinusoidal tones vs. piano tones. The latter has multiple overtones, providing multiple cues for the pitch. I used only sinusoidal for that reason.

Having the tones presented via web transmission gives no control over the actual output. Despite having as little as 0.01% total harmonic distortion in the amplifiers, output devices such as speakers and headphone or ear buds have around 1% to 3% THD, all of the different kinds having different harmonic distortion profiles.

Their description of aging causing "sharping" due to hair cell stiffening with age is very good. But the possibility remains that the documented time distortion due to perceptual slowing with age can be involved. That needs prying apart with other perceptual testing for time distortion per subject. A longitudinal study with the same "true" AP subjects decades later would be wonderful for the aging/sharpening problem, but figure the odds.

All that aside, good AP and RP probably have the same genetic source for auditory perception (minus auditory memory). I think they're on to something.