Head-Fi.org › Forums › Equipment Forums › Sound Science › Perfect pitch may not be so 'perfect'
New Posts  All Forums:Forum Nav:

Perfect pitch may not be so 'perfect'

post #1 of 100
Thread Starter 
Found an interesting BBC article today...
Quote:
Originally Posted by BBC 

Perfect pitch may not be so 'perfect'
By Melissa Hogenboom
Science reporter, BBC News


People classified with perfect pitch may not actually be as in tune with the notes they hear as they think.

Played a long piece of music, a study group failed to notice when scientists turned the tones ever so slightly flat. They then misidentified in-tune sounds as being sharp.

Researchers say it demonstrates the adaptability of the mind even for those skills thought to be fixed at birth.

They have published the work in the journal Psychological Science.

Only around one in 10,000 people has the ability to correctly classify a note simply by hearing it. This phenomenon is called perfect, or absolute, pitch, and has been made famous by the well-known composers who are believed to have possessed such talents, such as Mozart and Beethoven.

Graduate student Stephen Hedger, from the University of Chicago, US, had perfect pitch identified by objective tests. He explained what it meant.

"I'm able to name any musical note in isolation without the aid of a reference note. Someone with perfect pitch would be able to tell you a car alarm is honking in F sharp, for example. Generally it enables people to identify notes across a wide variety of octaves."

Mr Hedger was tricked by his colleague who secretly adjusted the pitch on an electronic keyboard as he was playing a tune. The notes were made flat by 33 cents - which is one-third of the distance between adjacent keys on a piano.

When the note was shifted back to its original correct key, it sounded drastically sharp to Hedger, who explained he found it "shocking" that he had not noticed the change.

A similar model was tested on 27 students with perfect pitch. They were played a piece of music for 45 minutes which was gradually changed over time to become flatter.

The subjects then perceived the flattened music as in tune, while the in-tune notes were perceived as slightly sharp.

"In the literature, perfect pitch is talked about as a fixed ability, so it was quite surprising to find that as little as 45 minutes of de-tuned music could temporarily shift note categories," Mr Hedger told BBC News.

Source / rest of the article : http://www.bbc.co.uk/news/science-environment-22871651

Let the implications roll out!
post #2 of 100

Implication: perfect pitch is a misnomer.

post #3 of 100
Quote:
Originally Posted by xnor View Post

Implication: perfect pitch is a misnomer.


That absolute pitch is only ever really relative pitch. If you can take a listener with verifiable "perfect pitch" and recondition them over a short term I wonder whether their skills will always rebound. Given enough sessions of messing with them (for science!) it may be possible to significantly shift the scale.

post #4 of 100

Not likely to be permanent, considering a lifetime of "perfect" pitch can easily be altered in a short listening session.  This should work both ways.

post #5 of 100

I've known this for a long time from personal experience. I personally don't notice when pitch drifts a little, like during a performance: either in the audience or on stage.

 

Also, I'm not so convinced of a strong delineation between people with perfect (absolute) pitch and those without. Some people have good enough pitch memory or whatever is required to be able to pick out notes without reference on a range of frequencies on certain instruments but not as well on others. Some people categorized with perfect pitch probably actually can hear the lowering of the tuning as reported there, but it seems like it's at least much more common to be susceptible to being "tricked" as described. I mean, some people are sensitive enough to be bothered by equal-temperament tuning.

 

edit: actually, I think most people described as having "perfect" pitch don't really have anywhere close to perfect pitch accuracy in general. Without context, if you asked them to identify a pitch, they'd get the nearest half tone, but they might not be able to reliably say if it's 20 cents sharp, 20 cents flat, or in tune with typical A440 tuning.


Edited by mikeaj - 6/15/13 at 2:33pm
post #6 of 100

Imo, it also makes the statement "I know how the headphone sounded when I bought it and X hours later after break-in it sounded so and so different" even less believable.

post #7 of 100

These tests remind me of the reversing glasses experiments on perceptual adaptation, where reversing goggles were worn for several days until an image was flipped upside down.  Perhaps this could support the notion that removing bias with ABX listening tests simply allows the brain to adjust to what is being heard so that small differences are indistinguishable from one another?  Maybe most people's brains compensate for these small, nearly imperceptible differences, where they could exist.  The end result would still be the same.


Edited by sonitus mirus - 6/16/13 at 6:00am
post #8 of 100

It is quite common in experimental psychology to find that the brain doesn't react to slow changes, adapts to its environment to stay in a "comfortable" zone, and reacts strongly to contrast. Which is roughly what happened here.

 

It should come only as a big surprise to those who think that some people somehow have a direct acces to a tuning fork in their head.

post #9 of 100
Quote:
Originally Posted by sonitus mirus View Post

These tests remind me of the reversing glasses experiments on perceptual adaptation, where reversing goggles were worn for several days until an image was flipped upside down.  Perhaps this could support the notion that removing bias with ABX listening tests simply allows the brain to adjust to what is being heard so that small differences are indistinguishable from one another?  Maybe most people's brains compensate for these small, nearly imperceptible differences, where they could exist.  The end result would still be the same.

Could you explain this?

 

You can take your time during an ABX test to get used to A or B and then play X for contrast or lack thereof. You could also rapidly switch between A and X or B and X.

 

I use both "techniques" although I dislike the rapid back-and-forth switching.

post #10 of 100

I don't know if the speed and frequency of switching from A and B is important.  If our brains can immediately adjust to subtle differences in sound, then it might be more likely to identify the differences if we knew they existed and expected to hear the changes.  A small difference might not be possible for most people to detect if our brains are unaware and try to make things right, so to speak.  This only seems to justify those claims I have read where an audio "expert" claims to know that they can hear a real difference, while they refuse to believe in a properly controlled ABX or the results they often demonstrate.

post #11 of 100

I don't think the brain immediately adjusts to a different sound. In the article they played the music for 45 minutes with a gradual change. That is what made the change in pitch "undetectable".

Quickly switching back to the initial pitch the difference was immediately noticed.

 

Everyone should be able to hear a 33 cent change in an ABX test. It's that ability to switch back and forth that really highlights differences. Outside an ABX test I wouldn't be so sure that people would detect the change in pitch, except for the unconditioned perfect pitch guys.

 

Also, I always go into an ABX test with the mindset that the files are different. Whether they really are is shown in the final result.


Edited by xnor - 6/16/13 at 9:09am
post #12 of 100
Quote:
Originally Posted by xnor View Post

Implication: perfect pitch is a misnomer.

I've always thought that, but didn't have real research like this to back it up. Further, what is the reference? In the US we use A=440 Hz, but in some countries A is slightly higher, and in the past A was lower. So the idea that someone could be "born" with an ability to identify [whatever Hz] as "correct" seems inherently flawed.

--Ethan
post #13 of 100
Quote:
Originally Posted by BBC 
...
Graduate student Stephen Hedger, from the University of Chicago, US, had perfect pitch identified by objective tests. He explained what it meant.

"I'm able to name any musical note in isolation without the aid of a reference note. Someone with perfect pitch would be able to tell you a car alarm is honking in F sharp, for example. Generally it enables people to identify notes across a wide variety of octaves."
...

That's only one aspect of perfect pitch. Another aspect is the ability to sing a specific pitch without reference or cue from another voice or instrument. This aspect was discussed in a BBC radio documentary on Auto-Tune entitled "Creating Pitch Perfect" (presented by classical singer Catherine Bott).

http://www.bbc.co.uk/iplayer/episode/b01shwkq/Creating_PitchPerfect/

The programme is available indefinitely to listen online for UK listeners (or anyone with use of a UK IP address).
post #14 of 100

It might not be the 'dictionary correct' usage of the terms. In fact they are just used interchangeably, but from my experience I have heard musicians use the terms absolute pitch and perfect pitch differently. 

 

One can acquire perfect pitch with a reasonable degree of accuracy. A lot of people are just born with it but often they are oblivious to small fluctuations in pitch. I think I belong in this category.

 

Absolute pitch is innate from birth. A lot rarer than perfect pitch. These individuals often struggle to perform in an ensemble as they are acutely aware of even the tiniest of pitch alterations. I have met several singers who fall in this category. They often consider it a curse as they need all other accompanying instruments to be tuned to their pitch of preference painstakingly. 


Edited by uchihaitachi - 6/17/13 at 8:56pm
post #15 of 100

I've known a few people with "perfect" pitch, and this doesn't really surprise me. It should actually just be called absolute pitch, since the possessors of this trait can't tell you the precise frequency or anything like that. They could tell you "an F a few cents flat" or something like that.

New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Sound Science
Head-Fi.org › Forums › Equipment Forums › Sound Science › Perfect pitch may not be so 'perfect'