Ultrasonics are sounds that we do not hear.
Our ear (theoretically) can hear from 20H to 20KHz. Ultrasonics are about 30KHz or higher. So we just cannot hear them.
Perhaps more properly, ultrasonics are frequencies above the normal range of human hearing.
When we say "normal range of human hearing" we are referring to a steady state sound. If you can hear a 16,000 Hz tone, but not a 19,000 Hz tone, then it is said your range of hearing ends somewhere between 16K and 19K.
However, frequency is not simply a case of a steady tone. Let us take the example of a steady tone this individual can hear, at 16,000 Hz.
What is the situation if this is not a test tone, but musical information?
What is the representation of a 16,000 Hz tone with a leading edge transient, as is the case in music?
The rise from no sound (we'll call it 0dB SPL) to a very loud sound that does not damage human hearing (say, 100 dB SPL) at a frequency of 16KHz is not the same as a steady tone of 16KHz. Frequency is critically time-dependant ... all frequencies are defined by an amplitude and a time. And we can represent this leading edge transient in terms of time.
In the case of an audio signal, the leading edge transient might be at a level of 100 dB SPL followed very quickly by a steady signal of perhaps 40 dB SPL and falling slowly. That is a crude representation of a musical transient of a 16 KHz fundamental note.
When we do consider such a signal, we come up with a frequency much higher than 16KHz, followed by a steady state frequency of 16KHz. For those who are familiar with an oscilloscope, you know that a high rise time can also be described as a high frequency resolution of the scope. A 'scope that can only handle a rise time equal to 20 KHz is not useful for testing audio signals ... it is not fast enough to capture all aspects of every potential signal in the audio band, and will not be able to handle a 19 KHz square wave accurately. For audio we typically want a scope that can resolve 10 Megahertz or more.
People do not actually hear via their ears alone ... they hear by a complex interaction with nerves in the ear, and these nerves lead directly to the brain with a very short path. The brain then interprets the nerve impulses and we "hear", or become aware, of a sound. Without a doubt, the vast majority of what is going on here is done in the brain, not the ear. I can put you in an environment where there is always a 16KHz sound. After a while, your brain refuses to pay attention to it, we no longer become aware of it, and it can be said we no longer "hear" it.
So, the brain is the critical component here. It can make us aware of a sound, including things we know we can easily hear. It can also make us unaware of these same things, even though it's clear that we should be able to hear them.
There is much evidence to suggest that the brain is aware of fast, high frequency leading edge transients, and that in the absence of such transients, our brains interpret the realism differently. We do not need to be listening to reproduced music to experience a fast rising high frequency sound ... it happens in real life with naturally occurring sounds, and forms part of our fight-or-flight defence system that all mammals share. In other words, it is not something the brain is in the habit of ignoring.
In this way we can see that if your hearing is such that the upper limit is 16,500 Hz with a steady state tone it is not the same as saying there is a "wall" that prevents you from detecting a signal above 16,500 KHz under all conditions. It is quite likely your brain will interpret a sound with a fast rise time equivalent to a frequency above 16,500 Hz as different than one without this fast rise time component. It does not take much to describe a signal with a rise time equivalent to well above 20 KHz in this example.
It is this aspect of Ultrasonic reproduction that is the subject of debate, and despite what some may say, there is a body of evidence to suggest we can detect, but not necessarily "hear" frequencies above the normal range of human hearing. The obvious problem here is we test our hearing with steady state tones only, so if an audiologist tests you and tells you your upper limit is 17,500 Hz, then it is normal to assume this means you cannot detect any sound above 17,500 Hz. But, no-one has tested you for "detection", you are only tested for steady state awareness. The tools an audiologist has at his disposal is not capable of such testing, so we have no true answer to the question.
Some have demonstrated an ability to detect a difference between music that naturally contains ultrasonic information versus the same music where such information is absent. In some cases, the ultrasonic information is largely or wholly distortion components of an audio system, and some people can reliably detect such distortion as a lessened sense of realism and sometimes even complain that it does not sound natural to them, despite exemplary performance in the usual human hearing region.
If we consider the audiologist's report as the final word this is impossible. So, another answer must be considered.
The natural areas to explore here are the areas of how the brain itself interprets sound information it gathers from it's audio nerve system. There is no serious debate with regard to whether the brain is the architect of what we hear ... the audiologist even considers it a proven fact. It has been shown conclusively, for example, that if we ask a panel of people to listen for a certain aspect of sound reproduction ... say, bass response ... they cannot recall anything about other aspects of the reproduction ... the brain filters that out when you decide to concentrate on the bass sound ... but they can clearly articulate the bass experience. When we say someone has a trained "ear", what is really happening is that person has a trained "brain", that pays more attention to sonic issues than ordinary persons would.
If that is the case, then we have to consider the possibility that ultrasonic information plays a role in our perceived realism of reproduced music on sound systems. It follows, then, that as designers of audio equipment, we should consider if there are possible non-linearities in the near ultrasonic range we should address, rather than dismiss them as irrelevant or impossible to detect, and further, if choosing a loudspeaker, whether it's ultrasonic performance (which we can measure and most manufacturers provide) should be considered in our evaluation of it's suitability for audio.
Home
Forum
Projects
Experiments
Circuits
Theory
BLOG
PIC Tutorials
Time for Science
