Okay, this is a pretty amazing auditory illusion. Here's what I think is going on. In the first syllable, there's only one major spectral peak below 2.5kHz. It has a wide bandwidth, which is consistent with an F1 and F2 very close together: an /ɑ/ (for "Laurel"). /1
The higher spectral prominence dips down about halfway through the word, between the two syllables. If the lower spectral prominence is F1 & F2, then the higher one must be F3. A low F3 = /ɹ/! Given the overall frequencies, the voice sounds male. /2
But what if we treat that higher spectral prominence as an F2, rather than an F3? Then we have a very high F2 in the first syllable, consistent with a front vowel or approximant, e.g. /j/. The F2 stays pretty high and the F1 gradually rises, giving a percept of /jæ/. /3
The fall of the F2 between the two syllables is then consistent with an /n/, although we don't see the general amplitude dampening that we normally associate with nasals. The F2 rises and F1 falls again at the end, resulting in /jæni/ overall. /4
The frequencies are overall higher than in the "laurel" interpretation, so the "yanny" speaker doesn't sound male, but also not unambiguously female either. /5
The big question is why do some people hear one and not the other, and why is it so hard to flip? Part of this is likely due to audio hardware - laptop speakers, fancy headphones, cheap headphones, background noise - but I think there's a cognitive dimension too. /6
Suppose you hear a sound with spectral prominences at 1 kHz, 1.5 kHz, 2 Hz, and so on. The even spacing of the harmonics will lead you to resolve a pitch of 500 Hz - even though that frequency isn't physically present in the signal. This is the "missing fundamental". /7
One of the first research projects I was involved in looked at individual differences in how these kinds of sounds are resolved. In short, some people tend towards hearing the "missing" f0, and others tend towards attending to the spectral structure. /8
My hunch is that "spectral listeners" are more likely to have the "yanny" percept, while "f0 listeners" are more likely to have the "laurel" percept. (You can see our missing fundamentals paper here: www2.hawaii.edu/~rory9/papers/… ) /9
If these kinds of individual differences are non-randomly distributed throughout populations, there are obvious consequences for our understanding of sound change via misperception (see Alan Yu's work on /s/~/ʃ/ perception, for example). /10
Anyway, it's possible to transform the "Laurel/Yanny" sound so that some people perceive it more in one direction than the other, but the effects seem to be stronger for some people than others, suggesting that it's a continuous rather than binary attention preference. /11
One of the participants in our pitch perception experiments actually told me that he was able to hear both pitches simultaneously. (He was a musician, interestingly enough.) So it's possible to overcome these illusions! /12
Anyway, that's all I have to say on that. Signal processing and acoustic phonetics are cool, but to get the whole story we have to think about psychoacoustics. Btw I totally hear "Laurel" fwiw. /fin
Update: Help me test my hunch about Yanny/Laurel and pitch perception! Have a listen and respond to the poll!
