the queen of prog?

Message Topic Search Topic Options Post Reply Create New Topic Printable Version Translate Topic

 Dean wrote:

 Star_Song_Age_Less wrote:

I do tend to go on and on.... basically the point of all this blather is that 1% does not mean 1 Hz.  It means 1% of the difference between one note and the next.

I know that 1% does not equate to 1Hz, but as you said this Just-noticeable Difference (JND) varies with pitch and this variation is not a linear relationship that can expressed as a generalised 'rule of thumb' percentage. 

The reason I am sceptical is that 10% equates to 165-cents or just over 1½ semi-tones. What this means is that the average Joe cannot identify a C followed by a C# as two different notes. At low frequencies this is probably true, but not at higher frequencies so it would make more sense to me if 10% was worse-case rather than average.

There's definitely a mis-communication here between what you're thinking of as 10% and what I am.  By my definition, 10% would always mean 10 cents  because the Hz value of 1 cent varies depending on which pitch you're starting from.  (But this does indeed vary non-linearly).  The 10% should never ever be anywhere close to even a half step between the notes on the western scale.

An average Joe definitely can tell the difference between a C and C#.    It's extremely rare for a person to be so tone deaf that they couldn't tell those two things were different.

And yes, near the ends of the frequency range of human hearing, this pretty much falls apart - but singers don't sing in those frequencies anyway.

I'll be on a semi-break for the winter soon, so I'll be able to use some time to look at this stuff in some detail again.

 Dean wrote:

 Star_Song_Age_Less wrote:

While we are on this topic, though, this is kind of interesting and fun in a weird way:  http://www.musicianbrain.com/pitchtest/
If
 you go to the link above, wear headphones.  My pitch discrimination 
compared to their base frequency was 1.1%-ish, if I remember right.  
Took it a long time ago.

That was pretty interesting and fun. 

And very curious...

Here are my results:

 Thank you for taking our Pitch Discrimination Test at http://musicianbrain.com/pitchtest.At 500 Hz you can reliably hear pitch differences of
3.375 Hz, which means you did better than approximately 66.5% of people who took
our test!

Being in the top 33.5% of those who took the test suggest that I am much better than average... 
...and
 that surprised me a little. I am a non-musician musician (I play and 
compose music badly and can just about manage to play a scale on a 
violin). Then we don't know the musical ability of those who took the 
test, the variance or the total population size. There could be a wide 
margin of error in these results, and I suspect that there is.
However,
 3.375Hz on 500Hz equates to 0.675% accuracy or 11-cents on the semitone
 and this suggests that the 50% average is probably a lot better than 
10% accuracy you cited for non-musicians. It would be cool to see other 
peoples' results.
I took the test at 9:27 am, it would be interesting to see how different the results are at various times of the day.

My guess would be that the people who take this test tend to be interested in music or at least in sound (some perhaps just in neuroscience), so I would guess there's a bias toward accuracy.

I might actually post this one as its own thread - then again that would bias their test toward how proggers hear pitch!  Ha!

Being a non-musician musician is still more attuned than a complete non-musician though, which most people are.