Interview with Peter Haughton Ph.D., Author: Acoustics for Audiologists
Share:
AO/Beck: Hi Peter. Thanks for meeting with me today. I'd like to start by learning a little about you. Can you please tell me about your education?
PMH: We lived in north London. I had an ordinary grammar school education followed by a BSc in physics at the Regent Street Polytechnic. I stayed on at the polytechnic for about three years doing work on the thermal conductivity of waxes that were being considered for use in heat exchangers. In 1968 I joined the Astbury Department of Biophysics at the University of Leeds where I worked on the mechanical properties of the cell walls of some simple algae. Plant cells are enclosed in a tough polysaccharide membrane (the cell wall) and it is something of a mystery how the cell manages to enlarge as the plant grows. It was believed, and probably still is, that the mechanical properties of the cell wall are important in controlling plant growth.
I completed my PhD in Leeds in 1968. I worked for two years at Leeds, and between those years I spent a year in Seattle at the University of Washington. At Leeds, I worked on the viscoelastic behaviour of artificially regenerated cellulose; in Seattle I was in the botany department, again looking at plant cell growth. That was a good year. I'd just married. We managed to get in a camping tour of the west coast as far as LA, returning to Seattle via Lake Powell, Death Valley & the Grand Canyon.
AO/Beck: That does sound like a good year! Please tell me your position in the Department of Medical Physics? How long have you been there and what are your primary responsibilities?
PMH: I joined the Medical Physics Department at Hull Royal Infirmary in 1975. The post was that of a hospital physicist but we're now called clinical scientists. The Infirmary is a district general hospital serving a population of about 500,000 in the north east of England. Medical Physics provides general technical support for the clinical work, traditionally in radiotherapy and imaging, but we are quite diverse and take on support roles in other areas including physiological measurement. In 1975 the ENT surgeons had heard of tympanometry but didn't really know what it was. They thought they needed help from Medical Physics, and that was my good fortune. I came to work here with no knowledge of hearing whatsoever. One could do that in those days. We were expected to get on with the job and learn new subjects for ourselves. It was definitely fun.
My primary responsibilities are to manage the Audiology Department and to participate in the technical work of the Medical Physics Service.
AO/Beck: Let's talk about your new book, titled ACOUSTICS FOR AUDIOLOGISTS, published by Academic Press (ISBN # 0-12-332922-1). One of the striking features to me is that your book is more scientific and more in-depth than many of the books currently used in graduate education. In other words, this book is more quantitative than previous texts. Is that correct?
PMH: Yes, you are quite right. My work with the examinations board for the national audiologists' training (BAAT part 1) has brought home to me the lack of any real physics in the present curriculum. However, when you look at textbooks on audiological subjects, you will find that an understanding of basic physics is essential. The same goes for many of the publications that appear in the audiological journals. In the UK things are about to change because audiology is going to be taught at degree level. I hope this will lead to a more thorough teaching of basic acoustics.
What I have tried to do in the book is to steer a course between a text that might be appropriate for someone studying physical science and the very elementary accounts that one often finds in the opening chapters of audiology books. Undergraduate acoustics texts are well outside the experience & physics-capability of most clinical audiologists. They are not exactly a push-over for physics graduates either! Obviously great simplification is needed. I have, however, tried throughout to make statements that are scientifically true, that is, not over-simplified. I have also tried to show, how acoustics is related to other parts of science such as optics, electricity and mechanics. The idea is to give readers a wider picture, or at least hint at its existence.
You used the word quantitative. Science is not much use unless you can apply it to something and in acoustics that often means being able to measure or calculate the values of variables such as sound pressure. We therefore need as a starting point quite precise statements about the relations between variables. Students will find that success in doing physics calculations' greatly enhances their confidence as well as their understanding of the subject. To put it simply, you can't get the right answers unless you know what you're doing.
AO/Beck: Can you please address the issue of mathematics, as presented in the book, and also regarding the audiologist's understanding of sound and acoustics as they relate to mathematics?
PMH: Mathematics is frequently a problem. The trouble is that the subject is inherently difficult. It requires learning, memory, discipline and practice. You can't get away with simply expressing an opinion or making a vague generalization. But in the end it is really a matter of moving symbols about according to agreed rules. So far as the book is concerned, I have used mathematics mainly for its value in showing relations concisely and unambiguously. All I ask is that students understand the symbolism. It won't matter at all if they can't derive one expression from another. Some introductory lessons in simple algebra and a couple of introductory lectures on understanding mathematical notation would probably help. What is essential is to be able to look at a mathematical expression, to know what it means and to understand the role of each of the variables within it.
I have kept the mathematics simple. There is no calculus. Vibrations are almost always given in terms of a sine function rather than the complex exponential form to be found in more advanced texts.
While writing the book I was aware that the mathematics and physics content might discourage readers. The comments about this in the preface and in the preamble to chapter 1 should be noted!
AO/Beck: I noticed throughout the text, at the end of the chapters, there are questions which can serve to self-test the reader to be sure they pick-up the key concepts. I suppose this means that the text was written with the graduate student as the target audience?
PMH: Not necessarily. Before the book was written the publishers sought the opinions of several reviewers who looked at a small sample of the text and the general plan. They all recommended that exercises should be included. I wonder how many readers will use them. It was quite a job checking all the answers and I'm especially grateful to my colleague Mike Spicer for his part in this. The questions after each chapter are quite varied in their difficulty. Some are easy but they do require an understanding of the text. Others are perhaps a serious challenge at this level.
The exercises are there for anyone who wants them. Yes, if the book is used as part of a taught course, lecturers may want to draw on the exercises as examples for classwork and home study. There are also a few suggestions for practical exercises which would, in most cases, require access to a laboratory.
AO/Beck: When I was studying acoustics, it occurred to me that virtually everything we studied was based on sinusoids and other pure forms of acoustic energy. However, with rare exception, we almost never listen to pure tones, and I wonder if you can comment on that, with particular regard to how the book approaches this issue?
PMH: Science succeeds because it reduces complex problems to simple ones. If you want to understand the acoustics of a complex waveform you have to start with a simple one - a sinusoid. It should be noted that many of the acoustic variables are definable only for sinusoidal vibration. Angular frequency is an obvious example. The transition from pure tones to complex sounds is conceptually fairly painless. We are usually quite happy to talk about spectra and frequency components' and we have Fourier's mathematics to back us. This is dealt with in Chapter 5.
While we're on the subject of sinusoids, allow me a small indulgence. I've noticed that in recent years the word amplitude has come in for some serious ill-treatment. Physiologists use it freely to describe, for example, the voltage in the ABR waveform but in physics and mathematics it means only the peak value in a sinusoid. I'll concede that some authorities do allow an extension to other vibrations, such as a modulated sinusoid, but all agree that amplitude denotes the peak value. What is unpardonable is the use of this word to denote the general value of a time-varying quantity. I have seen instantaneous amplitude used in more than one text. In my view this is a contradiction in terms that will lead to utter confusion in any subsequent analysis.
AO/Beck: For many of us in clinical practice, basic issues and answers are something we studied a long time ago -- and we rarely refer back to! Therefore, I was hoping you might address a few basic issues. For example, can you offer a pragmatic discussion on the decibel?
PMH: I've dealt exhaustively with decibels because of their great abundance in audiology. The subject is treated in the main text and again, with numerous examples in an appendix. The thing to bear in mind is that at heart, a decibel is a way of expressing sound power on a logarithmic scale. This conveniently leads to a scale in which equal scale intervals correspond to equal power ratios. To use the scale in an absolute sense we need to agree on a zero-value. That's all there is to it except to say that a Bel is a scale interval corresponding to a 10-fold change in power and a decibel is one-tenth of this interval. It corresponds to the sound power ratio 101/10 = 1·258. Adding decibels could mean adding intervals on a logarithmic scale. In effect this is multiplication, as done on a slide-rule.
AO/Beck: Could you please discuss directional microphones as they apply to hearing aids?
PMH: The subject is dealt with briefly on pages 166-172. So far as I know, directional microphones used in hearing aids are responsive to the gradient of the sound pressure, that is, to the rate at which pressure changes with position in the field. Sound pressure itself is omnidirectional but pressure gradient depends on the direction in which the wave is travelling. I think it's usually the case (and certainly true for plane waves) that the pressure gradient is greatest in the direction of travel. A pressure-gradient response can be obtained by supplying the microphone from two ports separated by a small distance. One port communicates with the front of the diaphragm and the other with the rear. The pressure difference - which drives the diaphragm - is then proportional to the change in pressure over the distance between the ports. The cunning trick is to delay the arrival of sound from the rear port relative to that from the front port. This is done, in effect, by inserting an acoustic resistance and compliance (analogous to an electrical resistance-capacitance element) in the line to the rear port. The polar response of the microphone depends on the delay. It can, if desired, be cardoid, supercardoid or hypercardoid.
AO/Beck: Would you please discuss the difference between admittance and impedance, as they relate to acoustics?
PMH: One is merely the reciprocal (the inverse) of the other. You might say the basic concept is that impedance is the ratio of a force or force-like quantity such as sound pressure to a velocity or its equivalent. So, for example, the acoustic impedance of the ear is the ratio of the sound pressure in the ear canal to what is called the volume velocity. The latter is the rate of flow of air in the canal as a consequence of the presence of sound. It is particle velocity multiplied by the cross-sectional area of the canal. In terms of their amplitudes or rms values, the ratio of these quantities is equal to the magnitude (modulus) of the impedance. In general, sound pressure and volume velocity are not in phase. They would be in the same phase in a plane progressive wave, but in the ear canal we find that the pressure lags the velocity when the system is dominated by its elastic rather than its mass properties. This is the usual situation at low frequencies. The full specification of impedance supplies both the magnitude and the phase information.
At the low frequencies usually used in tympanometry, the wavelength of sound is very large compared to the length of the ear canal. For this reason the sound pressure is virtually the same everywhere in the canal and it is therefore equal to the pressure at the tympanic membrane. The volume velocity (the airflow) is the sum of the flow of air in and out of the tympanometer probe and the equivalent flow at the tympanum as it moves in response to pressure. Now think of sound pressure as electrical voltage and volume velocity as electric current; they are exact analogues. The ratio of voltage to current is electrical impedance. So now, in the analogue, you have two electrical impedances, one representing the ear canal and the other the eardrum. They are connected in parallel because the same voltage (sound pressure) exists across each and the total current (volume velocity) is the sum of the current in each one. Now forget the phase. If voltage and current were in phase impedance would be the same as resistance. Remember from the theory of elementary electrical circuits that when resistors are in parallel you find the resistance of the combination R by the formula: 1/R = 1/R1 + 1/R2. This is where admittance comes in. Admittance equals 1/(impedance). It's obviously easier to work with admittance than with impedance because we can add directly instead of having to find the inverse of each term. 1/R = Y = Y1 + Y2.
AO/Beck: Regarding the state of the art of electroacoustic characteristics of hearing aids, is there some major change you would personally like to see in the near future? Is there something that we're missing or something that needs to be changed or updated as soon as possible?
PMH: As a general answer I'd have to say that I don't know enough about it. One thing I notice, though, is that a great deal of hard work goes into test box and real-ear measurement for fitting hearing aids. Part of the problem is matching the fitting, based on coupler/ear measurements, to the audiogram. Some hearing aid manufacturers produce hearing aids in which the audiogram is more or less obtained directly in situ by having the hearing aid generate test tones. It's true that this can't give accurate hearing levels (because there isn't a normal reference) but it does permit a good estimate of sensation level which should be sufficient for fitting purposes. Perhaps more can be done in this area. My rather pessimistic view is that the work done in the 1940s has not really been improved on very much. It may be a fact of physiology that no amount of tinkering with the incoming signal can compensate for the sensory defect. Despite all the wonderful technology we still find the aided performance is generally about where it was decades ago. That's a provocative view to upset your readers. Those sending me hate-mail won't receive a reply!
AO/Beck: Thanks very much Peter, I appreciate your knowledge and time, and I want to thank you for speaking with me today.
PMH: Thanks very much indeed for giving me the opportunity to take part in this discussion.
PMH: We lived in north London. I had an ordinary grammar school education followed by a BSc in physics at the Regent Street Polytechnic. I stayed on at the polytechnic for about three years doing work on the thermal conductivity of waxes that were being considered for use in heat exchangers. In 1968 I joined the Astbury Department of Biophysics at the University of Leeds where I worked on the mechanical properties of the cell walls of some simple algae. Plant cells are enclosed in a tough polysaccharide membrane (the cell wall) and it is something of a mystery how the cell manages to enlarge as the plant grows. It was believed, and probably still is, that the mechanical properties of the cell wall are important in controlling plant growth.
I completed my PhD in Leeds in 1968. I worked for two years at Leeds, and between those years I spent a year in Seattle at the University of Washington. At Leeds, I worked on the viscoelastic behaviour of artificially regenerated cellulose; in Seattle I was in the botany department, again looking at plant cell growth. That was a good year. I'd just married. We managed to get in a camping tour of the west coast as far as LA, returning to Seattle via Lake Powell, Death Valley & the Grand Canyon.
AO/Beck: That does sound like a good year! Please tell me your position in the Department of Medical Physics? How long have you been there and what are your primary responsibilities?
PMH: I joined the Medical Physics Department at Hull Royal Infirmary in 1975. The post was that of a hospital physicist but we're now called clinical scientists. The Infirmary is a district general hospital serving a population of about 500,000 in the north east of England. Medical Physics provides general technical support for the clinical work, traditionally in radiotherapy and imaging, but we are quite diverse and take on support roles in other areas including physiological measurement. In 1975 the ENT surgeons had heard of tympanometry but didn't really know what it was. They thought they needed help from Medical Physics, and that was my good fortune. I came to work here with no knowledge of hearing whatsoever. One could do that in those days. We were expected to get on with the job and learn new subjects for ourselves. It was definitely fun.
My primary responsibilities are to manage the Audiology Department and to participate in the technical work of the Medical Physics Service.
AO/Beck: Let's talk about your new book, titled ACOUSTICS FOR AUDIOLOGISTS, published by Academic Press (ISBN # 0-12-332922-1). One of the striking features to me is that your book is more scientific and more in-depth than many of the books currently used in graduate education. In other words, this book is more quantitative than previous texts. Is that correct?
PMH: Yes, you are quite right. My work with the examinations board for the national audiologists' training (BAAT part 1) has brought home to me the lack of any real physics in the present curriculum. However, when you look at textbooks on audiological subjects, you will find that an understanding of basic physics is essential. The same goes for many of the publications that appear in the audiological journals. In the UK things are about to change because audiology is going to be taught at degree level. I hope this will lead to a more thorough teaching of basic acoustics.
What I have tried to do in the book is to steer a course between a text that might be appropriate for someone studying physical science and the very elementary accounts that one often finds in the opening chapters of audiology books. Undergraduate acoustics texts are well outside the experience & physics-capability of most clinical audiologists. They are not exactly a push-over for physics graduates either! Obviously great simplification is needed. I have, however, tried throughout to make statements that are scientifically true, that is, not over-simplified. I have also tried to show, how acoustics is related to other parts of science such as optics, electricity and mechanics. The idea is to give readers a wider picture, or at least hint at its existence.
You used the word quantitative. Science is not much use unless you can apply it to something and in acoustics that often means being able to measure or calculate the values of variables such as sound pressure. We therefore need as a starting point quite precise statements about the relations between variables. Students will find that success in doing physics calculations' greatly enhances their confidence as well as their understanding of the subject. To put it simply, you can't get the right answers unless you know what you're doing.
AO/Beck: Can you please address the issue of mathematics, as presented in the book, and also regarding the audiologist's understanding of sound and acoustics as they relate to mathematics?
PMH: Mathematics is frequently a problem. The trouble is that the subject is inherently difficult. It requires learning, memory, discipline and practice. You can't get away with simply expressing an opinion or making a vague generalization. But in the end it is really a matter of moving symbols about according to agreed rules. So far as the book is concerned, I have used mathematics mainly for its value in showing relations concisely and unambiguously. All I ask is that students understand the symbolism. It won't matter at all if they can't derive one expression from another. Some introductory lessons in simple algebra and a couple of introductory lectures on understanding mathematical notation would probably help. What is essential is to be able to look at a mathematical expression, to know what it means and to understand the role of each of the variables within it.
I have kept the mathematics simple. There is no calculus. Vibrations are almost always given in terms of a sine function rather than the complex exponential form to be found in more advanced texts.
While writing the book I was aware that the mathematics and physics content might discourage readers. The comments about this in the preface and in the preamble to chapter 1 should be noted!
AO/Beck: I noticed throughout the text, at the end of the chapters, there are questions which can serve to self-test the reader to be sure they pick-up the key concepts. I suppose this means that the text was written with the graduate student as the target audience?
PMH: Not necessarily. Before the book was written the publishers sought the opinions of several reviewers who looked at a small sample of the text and the general plan. They all recommended that exercises should be included. I wonder how many readers will use them. It was quite a job checking all the answers and I'm especially grateful to my colleague Mike Spicer for his part in this. The questions after each chapter are quite varied in their difficulty. Some are easy but they do require an understanding of the text. Others are perhaps a serious challenge at this level.
The exercises are there for anyone who wants them. Yes, if the book is used as part of a taught course, lecturers may want to draw on the exercises as examples for classwork and home study. There are also a few suggestions for practical exercises which would, in most cases, require access to a laboratory.
AO/Beck: When I was studying acoustics, it occurred to me that virtually everything we studied was based on sinusoids and other pure forms of acoustic energy. However, with rare exception, we almost never listen to pure tones, and I wonder if you can comment on that, with particular regard to how the book approaches this issue?
PMH: Science succeeds because it reduces complex problems to simple ones. If you want to understand the acoustics of a complex waveform you have to start with a simple one - a sinusoid. It should be noted that many of the acoustic variables are definable only for sinusoidal vibration. Angular frequency is an obvious example. The transition from pure tones to complex sounds is conceptually fairly painless. We are usually quite happy to talk about spectra and frequency components' and we have Fourier's mathematics to back us. This is dealt with in Chapter 5.
While we're on the subject of sinusoids, allow me a small indulgence. I've noticed that in recent years the word amplitude has come in for some serious ill-treatment. Physiologists use it freely to describe, for example, the voltage in the ABR waveform but in physics and mathematics it means only the peak value in a sinusoid. I'll concede that some authorities do allow an extension to other vibrations, such as a modulated sinusoid, but all agree that amplitude denotes the peak value. What is unpardonable is the use of this word to denote the general value of a time-varying quantity. I have seen instantaneous amplitude used in more than one text. In my view this is a contradiction in terms that will lead to utter confusion in any subsequent analysis.
AO/Beck: For many of us in clinical practice, basic issues and answers are something we studied a long time ago -- and we rarely refer back to! Therefore, I was hoping you might address a few basic issues. For example, can you offer a pragmatic discussion on the decibel?
PMH: I've dealt exhaustively with decibels because of their great abundance in audiology. The subject is treated in the main text and again, with numerous examples in an appendix. The thing to bear in mind is that at heart, a decibel is a way of expressing sound power on a logarithmic scale. This conveniently leads to a scale in which equal scale intervals correspond to equal power ratios. To use the scale in an absolute sense we need to agree on a zero-value. That's all there is to it except to say that a Bel is a scale interval corresponding to a 10-fold change in power and a decibel is one-tenth of this interval. It corresponds to the sound power ratio 101/10 = 1·258. Adding decibels could mean adding intervals on a logarithmic scale. In effect this is multiplication, as done on a slide-rule.
AO/Beck: Could you please discuss directional microphones as they apply to hearing aids?
PMH: The subject is dealt with briefly on pages 166-172. So far as I know, directional microphones used in hearing aids are responsive to the gradient of the sound pressure, that is, to the rate at which pressure changes with position in the field. Sound pressure itself is omnidirectional but pressure gradient depends on the direction in which the wave is travelling. I think it's usually the case (and certainly true for plane waves) that the pressure gradient is greatest in the direction of travel. A pressure-gradient response can be obtained by supplying the microphone from two ports separated by a small distance. One port communicates with the front of the diaphragm and the other with the rear. The pressure difference - which drives the diaphragm - is then proportional to the change in pressure over the distance between the ports. The cunning trick is to delay the arrival of sound from the rear port relative to that from the front port. This is done, in effect, by inserting an acoustic resistance and compliance (analogous to an electrical resistance-capacitance element) in the line to the rear port. The polar response of the microphone depends on the delay. It can, if desired, be cardoid, supercardoid or hypercardoid.
AO/Beck: Would you please discuss the difference between admittance and impedance, as they relate to acoustics?
PMH: One is merely the reciprocal (the inverse) of the other. You might say the basic concept is that impedance is the ratio of a force or force-like quantity such as sound pressure to a velocity or its equivalent. So, for example, the acoustic impedance of the ear is the ratio of the sound pressure in the ear canal to what is called the volume velocity. The latter is the rate of flow of air in the canal as a consequence of the presence of sound. It is particle velocity multiplied by the cross-sectional area of the canal. In terms of their amplitudes or rms values, the ratio of these quantities is equal to the magnitude (modulus) of the impedance. In general, sound pressure and volume velocity are not in phase. They would be in the same phase in a plane progressive wave, but in the ear canal we find that the pressure lags the velocity when the system is dominated by its elastic rather than its mass properties. This is the usual situation at low frequencies. The full specification of impedance supplies both the magnitude and the phase information.
At the low frequencies usually used in tympanometry, the wavelength of sound is very large compared to the length of the ear canal. For this reason the sound pressure is virtually the same everywhere in the canal and it is therefore equal to the pressure at the tympanic membrane. The volume velocity (the airflow) is the sum of the flow of air in and out of the tympanometer probe and the equivalent flow at the tympanum as it moves in response to pressure. Now think of sound pressure as electrical voltage and volume velocity as electric current; they are exact analogues. The ratio of voltage to current is electrical impedance. So now, in the analogue, you have two electrical impedances, one representing the ear canal and the other the eardrum. They are connected in parallel because the same voltage (sound pressure) exists across each and the total current (volume velocity) is the sum of the current in each one. Now forget the phase. If voltage and current were in phase impedance would be the same as resistance. Remember from the theory of elementary electrical circuits that when resistors are in parallel you find the resistance of the combination R by the formula: 1/R = 1/R1 + 1/R2. This is where admittance comes in. Admittance equals 1/(impedance). It's obviously easier to work with admittance than with impedance because we can add directly instead of having to find the inverse of each term. 1/R = Y = Y1 + Y2.
AO/Beck: Regarding the state of the art of electroacoustic characteristics of hearing aids, is there some major change you would personally like to see in the near future? Is there something that we're missing or something that needs to be changed or updated as soon as possible?
PMH: As a general answer I'd have to say that I don't know enough about it. One thing I notice, though, is that a great deal of hard work goes into test box and real-ear measurement for fitting hearing aids. Part of the problem is matching the fitting, based on coupler/ear measurements, to the audiogram. Some hearing aid manufacturers produce hearing aids in which the audiogram is more or less obtained directly in situ by having the hearing aid generate test tones. It's true that this can't give accurate hearing levels (because there isn't a normal reference) but it does permit a good estimate of sensation level which should be sufficient for fitting purposes. Perhaps more can be done in this area. My rather pessimistic view is that the work done in the 1940s has not really been improved on very much. It may be a fact of physiology that no amount of tinkering with the incoming signal can compensate for the sensory defect. Despite all the wonderful technology we still find the aided performance is generally about where it was decades ago. That's a provocative view to upset your readers. Those sending me hate-mail won't receive a reply!
AO/Beck: Thanks very much Peter, I appreciate your knowledge and time, and I want to thank you for speaking with me today.
PMH: Thanks very much indeed for giving me the opportunity to take part in this discussion.
