I Can't Understand What They're Saying.

[Patrick Bateman]

Last Sunday I was at Ikea, doing a little shopping. If you've been to Ikea, you know that the store forces you to wind through the entire place before you reach the cash registers. In front of me was a couple who kept bickering. The husband kept complaining to the wife "I can't understand what you're saying!"

One component of intelligibility is sheer volume; if you shout over the noise around you, you'll be understood.

But there's another component of intelligibility that isn't considered often. This is the interaural time delay. "Interaural time delay" is just a fancy way of saying "the time delay introduced by the width of your head." The ITD is responsible for location cues, and it also contributes to speech intelligibility.

Which brings us back to our bickering couple. The reason that they couldn't understand each other was that they were walking in single file. If they were walking next to each other, or at a 45 degree angle, it wouldn't have been a problem. Here's a picture to illustrate:



If you are having trouble understanding someone, consider turning your head a few degrees, or moving a bit to the left or to the right. This will introduce and ITD, and improve your ability to understand what they are saying. It's also one of the reasons that people are easier to understand when you're sitting in a booth at a restaurant; there are reflections of their voice off the surface of the booth. (the booth cuts down on ambient noise as well.)

So what the heck does this have to do with loudspeakers?

It's one of the reasons single-driver speakers are so beloved. They don't suffer from the same ITDs as large two-ways. A loudspeaker designer can reduce the ITD present in a two-way loudspeaker by clever optimization of the crossover and the baffle. But they can only solve the problem at one point in space, which is one of the reasons that single driver speakers image so well when you walk around the room.

Just a little food for thought, something to consider before you embark on that crazy two-way loudspeaker with the huge woofer.

If you've listened to a Unity horn, the intelligibility is uncanny, due to the almost complete lack of ITDs, over a six octave bandwidth.

To explore this subject further, check out these books and articles:

#1 - "auditory perception of sound sources" - yost, popper, fay 2008

#2 - "The Role of Interaural Intensity Differences and Time Delay For Signal Detection in Noise" http://www.isa-audiology.org/periodi.../No.%202%20%20(5-206)/Feldmann,%20%20InternatAudio,%20%201965.pdf

[HK26147]

See Also: Master Handbook of Acoustics:
The Ear and the Perception of sound
particularly: Localization of Sound Sources.
It details the delay ( because of the spacing between ears ), and other related issues...

"Below 1khz the phase (time) effect dominates while above 1Khz the intensity effect dominates. There is one localization blind spot. A listener cannot tell whether sounds are coming from directly in front, or from directly behind because the intensity of sound arriving at each ear is the same and in the same phase"
( re Binaural Location )

[Patrick Bateman]

Originally posted by HK26147
See Also: Master Handbook of Acoustics:
The Ear and the Perception of sound
particularly: Localization of Sound Sources.
It details the delay ( because of the spacing between ears ), and other related issues...

"Below 1khz the phase (time) effect dominates while above 1Khz the intensity effect dominates. There is one localization blind spot. A listener cannot tell whether sounds are coming from directly in front, or from directly behind because the intensity of sound arriving at each ear is the same and in the same phase"
( re Binaural Location )

Yes!

And this frequency that you mention is actually related to the gap between our ears. For instance, the gap between my ears is about 23cm, which corresponds to a frequency of 1.5khz.

[Pano]

Hmmm.... the "blind spot" might be why my father hates listening to music thry headphones. He says it always sounds like it's behind him.


Intersting topic, thanks for bringing it up. (will go digging in the Master Handbook of Acoustics)

[HK26147]

And some distances correspond to the 1/2 wavelength distance for cancellation at other frequencies.
Of course some are thicker between the ears then others ( Arr Arr ).
The ear is definitely an adaptive design; it has ideally a 10 octave range of large and non-linear sensitivity.
A 20Hz wavelength is 56.5' long and a 20Khz is a fraction of a inch so the design features that aid in LF perception are different then those for HF where a wavelength is inches.
The ear canal has resonate peak @ 3Khz ( due to the 3cm ear canal ).
I have also see research on the differentiation of Cilia adapted, for specific function.
Some fire ( from stimuli ) once, some fire and hold and some are arranged to implement "phase locking".

Of course what about perception of sound that is not "ear based".
Skin, bone, etc

[pedroskova]

Has anyone ever noticed, while running stepped sine waves, that certain frequencies resonate in your ears while others cancel? You then look at the fr graph and it's flat.

[breez]

Patrick Bateman, can you clarify your opening post a bit? ITD is presented as having a positive effect on speech intelligibility, but as a negative trait in loudspeakers. And how do loudspeakers with non-coincident drivers introduce ITD?