Interesting. I understood the first one without ever hearing anything like it. Maybe I cheated, I had to play it twice because the volume was down too low the first time for me to make it out. Anybody else able to understand these easily though?

Ok, the next one I didn't get. That's pretty trippy.

That's so cool!

I was able to understand the first one only after listening to it about 5 times. I could also understand most parts of the others without hearing the clear speech, but it's amazing how much they jumped out when you knew what they were supposed to say.

This is not entirely dissimilar to the principle that allows cochlea implants to work, and is the auditory equivalent of pareidolia.

If it were the auditory equivalent of pareidolia, people would hear words in the sine wave speech that were not said. I think it's more like an optical illusion - where you can't see the image at first, but then once you've got it, you can't stop seeing it.

That was neat! I could only understand one of them right off the bat; I had to listen to the others a few times to pick out words. I had to imagine the whistley sound had an English accent and then I could hear it just fine.

Ha, C3PO, you've got nothing on me now - I can understand R2D2 now!

That is amazing. I really have to listen to the normal speech first - but once I do, it manipulated version is clear.

What was interesting to me was that at the first listen only certain words would jump out like "museum" or "lunchtime," but after a few more then I could actually hear the whole thing.

This is an example of "top-down" processing in which your brain makes certain assumptions about the world and adjusts your perceptual experience to conform to those assumptions.

In the case of sine wave speech, it is usually not intelligible until you have been primed to expect a certain content. This happens in everyday conversation to a different extent: you are primed according to the content and context of the conversation.

This is also proof that human speech sounds are not inherently different than any other sound in the world, it is simply your brain that has adapted to interpret speech as a meaningful set of sounds, and everything else as rather arbitrary. Once you realize that sine wave speech is actually a form of speech, the other examples become easier to understand.

I hope you enjoyed the post!

If you want an example of audio pareidolia, check out this and there's some other similar things on youtube too. If you don't feel like watching a video, the link is the Indian version of thriller "translated" by captioning it with the closest sounding English words. If you close your eyes and don't speak Indian, you won't hear it; but with the words there, you hear it.

Wow... that's really cool.

Thanks. I discovered something: Salad is a cargo I could do.

I showed these to my Aunt and she got all of them straight away. The opnly explaination we can come up with is that she works with dementia patients and has to decipher what they are trying to say when they speak.

Often with dementia patients they have the words set up inside their head but they don't come out in the correct form that we would recognise as words.

I think the best example of pareidolia is this. Seriously.

I don't know, that includes a lot of making fun of how they pronounce English words (like "G-R-double E-N Leaves"). And was that made by a speed-reader? I couldn't read anything that filled up the screen because it went by too quick, I'm tempted to watch it again but it hurts my eyes.

I listened and couldn't make out any words the first time. I had my son listen..he has perfect pitch...and he got every one right before listening to the translation.

Okay, I just tested the daughter, and she missed parts, too.

R2D2 noises were made with an ARP 2006 Synthesizer.