Friday, February 27, 2009

Research Update #1

After learning about Tartini/difference tones, I was interested to learn how these could affect harmonies in music--I'm trying now to make a connection between the rigid, quantitave mathematics of physics, and the opinionated, qualitative aesthetics of music. Accordingly, in this post I'll be using some music-based ideas such as the qualitative 'dissonance' and 'consonance' as well as "chord qualities" such as 'Major', 'Minor', and 'Perfect'.

**One important note is that although I call dissonance and consonance qualitative, they are, in fact, biologically different. When the brain processes consonant harmony, neurons fire at different rates; when the brain processes dissonant harmony, neurons fire simultaneously. I guess what I mean by qualitative is the level of dissonance--some harmonies are obviously dissonant to everyone (such as the minor and major 2nd), and some harmonies which are classified as dissonant simply sound excessively dark and brooding to some people (such as the diminished 5th a.k.a. the Tritone). The book This is Your Brain on Music: the Science of a Human Obsession is where I read the information about dissonance and consonance being actual differences which occur in the mind. A bibliographical entry for this book is available at the end of the post about Hypersonic Sound; also, I'll be publishing a full bibliography for the information in this post in one of the next few posts.

**Chord qualities are not always meant to convey a certain mood--major is not always positive and minor is not always negative; however, often they are said to have a certain mood attributed to them. This depends greatly on whether or not the chords are consonant or dissonant and will be explained further in the chart which compares the 2 sounding tones and their difference tone.

**Mathematically, to find the pitches of each note you would take the appropriate logarithmic scale which passed through 130.8 and 261.6 Hz and divide the x-axis into 12 parts including these tones. The y-coordinate is the frequency of the tone x number of half-steps above C3. This is based on the "A=440" scale which means that the "A" above "Middle C" is at a frequency of 440Hz. The "3" placed next to the "C" denotes it's octave relative to a piano's range. The space between 2 subsequent divisions is called a "half tone".

So, because there is an octave between 130.8 and 261.6 (i.e. 261.6 has twice the frequency of 130.8) and this space has been divided into 12 parts which are each a half tone apart, the octave contains 12 half tones (the upper octave tone is the 12th half tone, and the lower octave is not counted in this case). These are named (starting on C) "C-Dflat-D-Eflat-E-F-Gflat-G-Aflat-A-Bflat-B-C". From now on, i'll be using "b" to indicate "flat" as this resembles the actual musical notation for a flat symbol -- Eb = Eflat.

**Also, notice that there is no "Fb" or "Cb"... this is because an "Fb" is the same pitch as an "E" and a "Cb" is the same pitch as a "B". The reason for this is simply because although the scale is broken into 12 half-tone pitches, naming conventions which have developed over the centuries use 7 letters rather than 12 or 6. [In more advanced music and in music theory, sometimes Fb and Cb are notated depending on the scale and chord which is being used; also, sometimes Csharp (C#) would be written instead of Db for the same reason (this goes for other notes too, such as G# instead of Ab)... this is a naming convention issue that doesn't affect the sound of a chord... regardless of name, an Fb sounds the same as an E, and a C# sounds the same as a Db.]

Although at this point I'm not sure, I think that the decision to use 7 pitches instead of 12 has something to do with how a typical major scale is constructed: it uses 7 different tones. The C-major scale contains the notes C-D-E-F-G-A-B-C; notice that it contains no flats or sharps. I find it interesting that the only major scale which contains no flats or sharps starts on the letter "C" rather than "A" (which would make logical sense since "A" starts the alphabet); however, the only minor scale which contains to flats or sharps is the A-minor scale, which contains the notes A-B-C-D-E-F-G-A. Perhaps the reason for this is that in musical history the minor scale developed prior to the major scale or had more importance than the major scale. Then, maybe, from these scales the other ones were named and 'adding' a flat or sharp to a note was easier to remember than trying to remember which letters were in each scale. i.e: it was easier to write "G, A, B, C#, D, E, F#, G," rather than naming each half step "A, B, C, D, E, F, G, H, I, J, K, and L," and trying to construct the same scale from that based on the number of half-tones between each scale tone, which would lead to confusion as some letters would be entirely excluded. In our major and minor scales, every letter A through G is used once, and then depending on the scale a certain number of sharps or flats are added to some notes.

**Interference beats occur at a rate which is equal to the difference between the two sounding tones. The number of beats per second between the two tones 200Hz and 203Hz is 3. Because humans cannot hear pressure fluctuations below 20Hz, when two tones which are less than 20Hz apart sound simultaneously, it is usually perceived simply as a fluctuation in sound volume; over 20Hz, and a Tartini tone is heard with a frequency of the difference between the two sounding tones. The Tartini tone resulting from 200Hz and 250Hz is an audible tone at 50Hz. As you should remember, this is the basis for HSS.

Initially I created a chart which compared the musical tones between and including 130.8 Hz to 261.6 Hz--C3 to C4. The chart compared the base tone C3 to each of the 12 subsequent chromatic (going up by half-step) scale tones. I then included in this chart the frequency and note name of the resulting Tartini tone. My hypothesis was: If the difference in the two sounding tones is below 20Hz (because humans would perceive this as a rapid volume fluctuation rather than as an added third tone (the Tartini tone)), then the resulting harmony will be recognized as being dissonant.

Using the major 2nd, a very dissonant interval, as an example, I noted that between C3 and the note one whole step higher (two half steps), D3, the difference in pitches was a mere 16Hz--Slow enough to not produce an audible Tartini tone, and yet fast enough to not really be perceived as a volume fluctuation. According to my hypothesis, this was the reason for the dissonance--the fact that it didn't produce a Tartini tone and yet was more than just a volume fluctuation.

What you should quickly realize, however, is that as you move up in octaves, keeping this same musical (or mathematical if you're following the same logarithmic scale) interval of a Major 2nd (to reiterate: 2 half steps), the pitches move farther and farther apart frequency-wise. Between C3 and D3 there are 16 Hz of separation, however between C7 and D7 there are 256 Hz of separation. And yet, C7 and D7 are still extremely dissonant despite producing a substantial Tartini tone. So, my initial hypothesis was shown to be wrong.

**One note which I must make is that in music, chords which are in the lower octaves do tend to sound more "muddled" than chords in higher octaves. It is standard music compositional practice to allow about a Perfect 4th (5 half steps) between the lowest tone and the next tone above it--sometimes more or less depending on how deep the lowest tone is. I'm almost sure that this has to do with the tones being so close in frequency combined with the human ear not being able to differentiate such relatively small differences as easily as larger differences.

Here is the chart that I made comparing C3 to all of it's possible half tone harmonies up to C4:

Note ; # of 1/2 steps ; Hz difference ; harmony name ; Note name of Tartini tone (approx)
C3......... 0.................. 0............. Perfect unison.................... --

Db........ 1.................. 7.78.......... minor second................... C--

D.......... 2................ 16.02.......... major second ...................C0

Eb........ 3............... 24.75........... minor third...................... G0

E ...........4............... 34.00......... major third ...................Db1

F ...........5 ...............43.80 .........Perfect fourth ..................F1

Gb........ 6............... 54.19 ..........Tritone............................. A1

G.......... 7............... 65.19.......... Perfect fifth ......................C2

Ab........ 8............... 76.84.......... minor sixth..................... Eb2

A.......... 9............... 89.19.......... major sixth ..................F2/Gb2

Bb .......10............. 102.27 .........minor seventh................. Ab2

B......... 11.............. 116.13 ..........major seventh................. Bb2

C4 .......12 .............130.81 .........Perfect octave ..................C3

**I'll be using the term "inversion" in the next few paragraphs. An inversion is when you take two notes, such as a C and an E, and raise the lower note one octave. Ex: The base (not bass) tone is C3 and the harmonizing tone is E3--there are 4 half steps between these two tones and the interval is a major third. The C3 is raised to C4 so that now the base tone is E3--There is a distance of 8 half steps between these two notes and the interval is a minor sixth.

**"Perfect" harmonies are considered to be entirely consonant, however have basically no 'mood' associated with them. The Major 3rd is considered a "happy" or "positive" chord--the Minor 3rd a "sad" or "negative" chord. Both the minor and major 2nd are considered very dissonant tones. The major 6th, while considered fairly positive, is not as "happy" as a major 3rd. This most likely stems from the Major 6th being an inversion of the Minor 3rd. The minor 6th is exactly the opposite; additionally, the minor sixth is considered slightly dissonant and dark or brooding. The minor seventh is typically a slightly uplifting interval. This is most likely due to it "wanting" to resolve to--or lead into--another note; in other words, it creates anticipation which is normally interpreted positively. The Major seventh, being an inversion of the minor second, is fairly dissonant.

After making this chart I saw some interesting correlations between dissonance, consonance, and chord quality, and the tartini tone produced. Some of the relations made sense... some didn't:

First of all: dealing with the Perfect intervals, the unison produces no Tartini tone because the pitches are the same; the octave produced a Tartini tone equivalent to C3 because it's a doubling of that note and thus the frequency difference is equal to the base tone. The perfect 4th (F3) produced a Tartini tone of F2, and the perfect 5th (G3) produced a Tartini tone of C2. This is interesting because, as can be seen, all of the other intervals produced tones besides just the 2 harmonizing tones. Another note to make, unrelated to the Tartini tone, is that a perfect 4th is an inversion of a perfect 5th and thus has a similar, although just slightly more dissonant, quality.

Secondly: dealing with the thirds and sixths, the minor third (Eb) produced a Tartini tone of G0, which forms a perfect harmony with C and thus, having relatively no dissonance or other quality, shouldn't affect this chord at all--and yet the minor third is considered a "sad"-sounding harmony. The major third (E), however, was interesting in that the Tartini tone produced by it was Db1. This note is dissonant to the C when transposed upwards a few octaves to be more relative to C3, and forms a minor third with the E when transposed upwards. And yet, despite the dissonance with the C and the "sad", minor quality with the E, the Major third is widely considered a "happy" interval. The minor sixth (Ab) produced an Eb as a Tartini tone. This is interesting as a minor third is considered to have an ominous or "minor" sound, and is fairly dissonant. It is interesting to note, however, that the Eb does form a perfect interval with the Ab. Remember that a MINOR 6th is the inversion of a MAJOR 3rd, and a MAJOR 6th is the inversion of a MINOR 3rd.

Thirdly: the sevenths and seconds. The minor second (Db) produced the C three octaves below the base tone C3, which is an inaudible frequency. What is interesting, is that although this is exactly what happened with the non-dissonant perfect intervals, the minor second is one of the absolute most dissonant intervals in music. The major second (D)also produced a C, this time only two octaves below C3--which is still inaudible (although at higher frequencies this interval would produce an audible Tartini tone.)-- and once again, although this is exactly what the perfect intervals did, the major second is one of the most dissonant intervals in music. The minor seventh (Bb), and inversion of the major second, is less dissonant than the major second by far, and, as I've said, is considered to be positive probably due to it's tendency to create anticipation. This interval produced a Tartini tone at Ab-- a tone that is very dissonant with the Bb, and also somewhat dissonant with the C. In fact, depending on the inversion, an Ab can form a minor third with a C which, as you should remember, is considered a "sad" interval--and yet despite this, the minor 7th is considered a very "positive"-sounding interval! The major 7th (B), being an inversion of the minor 2nd, is still very dissonant, although not quite as much as its inverted counterpart. It produced a Tartini tone of Bb2; although a Bb is not very dissonant with a C, when a C, B and the Tartini tone Bb are all played at the same time, 3 tones which are all only one half step apart are sounding and, as I've already said, the minor second (half step) is one of the most dissonant tones.



Lastly: The tritone (Gb). This is considered one of the most dissonant intervals in music--most likely due to the fact that it splits the octave evenly in half (6 half tones). While personally I don't consider it the most dissonant, it does sound very dark, ominous, and still has a lot of dissonance. In medieval times, musicians were not allowed to write this interval into music or play it because of a superstition that playing it would call the Devil to come forth from Hell (Sadly, I'm not making this up...). The reason for this dissonance is unclear, while the Tartini tone A does form a minor third with the Gb, this is not a particularly dissonant interval--more just "sad"-- and also the A is not very dissonant with the C itself, forming a major sixth.














Because there doesn't seem to be any real correlation on the surface between the Tartini tone produced and the amount of dissonance or the quality produced by an interval--especially because the Tartini tones are sometimes audible and sometimes inaudible depending on the octave--I plan on investigating further what, both in physics and in neurology, could be causing humans to perceive such a distinct difference between these characteristics.


Another thing that I want to investigate, which is related to the topics of quality and dissonance, is the development of musical scales over the past several millenia. For instance, while all of the Western scales, and several Eastern scales as well, are broken into a combination of the 12 half tones, many Eastern, Middle Eastern, African, and Pacific (most people lump all of these categories simply into "Eastern" scales) scales are broken into combinations of quarter tones instead. To most western ears, this would sound very dissonant and out-of-tune--most would probably perceive it as a mistake by the musician.


So:
1) Why do humans perceive consonance and dissonance? Is it physics, neuroscience, or a combination of the two?



2) Is there a deeper correlation that can be found between the half-step intervals and their dissonance/quality in relation to the Tartini tone which is produced? (One interesting thing that I plan on investigating soon is the Tartini tones' relation to the overtone series of a fundamental C... at a very quick glance, the progression of tones seems to follow this series fairly closely.)


3) Why were the various Western and Eastern scales developed the way they were? An exploration of the origins of melodic music, evolution of instruments, and spread of musical styles/genres will be conducted.



3.b) Was there a "base scale"-- a type of fundamental scale which most ancient cultures seemed to use from which the other scales were possibly derived as embellishments and minor changes were made? If so, why were those embellishments or changes being made the way they were?

Terms which I will be investigating so far are:

Aulos, Kithara, Lyre, Tetrachord, Quarter Tone, Pentatonic Scale, Terpander, Constantin Brailoiu, Al-Farabi, and Hebrew Music (which is what it is presumed most ancient music stemmed from).

Friday, February 13, 2009

Hypersonic Sound

Hypersonic Sound (HSS) is the term used to describe the process by which audible sound waves can be produced using ultrasonic sound waves that are free from non-linearity. The first attempts at hypersonic sound were made in the 1960’s using underwater sonar. In the 1970’s it was proven that mathematically HSS could be produced in air, but by the 1980’s the technology was abandoned because of problems with distortion. In the late 1990’s HSS was again researched because of advances in sound production technology and in 1998 the first working, commercial prototypes were made under the name “Audio Spotlight”.

The advantage of using ultrasonic sound is that sound transmissions can be focused into a narrow, far-reaching beam that resists diffusion and attenuation; therefore, the beam can be transmitted over greater distances with pinpoint accuracy. Additionally, this sound beam can be targeted to only a single object or person, leaving the surrounding environment free of noise pollution. Already, this technology is being put to use in the advertising and automobile industries, and the United States military.

First, it is important to understand what a sound wave is. A sound wave is a series of alternating high (condensation) and low (rarefaction) pressures created by some object disturbing the environment through which the sound wave is traveling. This pressure wave, then, is received by the eardrum which converts it through the inner ear into an electric signal which the brain can process. The key thing to recognize in the case of hypersonic sound is that each of these small pressure changes is a different micro-environment; the small portions which are low-pressure have different densities (atmospheric density is related to pressure) than those that are high-pressure. This is extremely important to note when dealing with the transmission of a sound wave across distances.
Next, it is important to understand the terms diffusion and attenuation, which describe the behavior of a sound wave over time. Diffusion is the process by which a sound wave expands outward, and attenuation is the process by which a sound’s intensity diminishes. These two characteristics of a sound wave are very interrelated; as a sound wave expands and increases its area occupied, its intensity (which is inversely proportional to area occupied) decreases. Additionally, a sound wave’s absorption into the surrounding environment as well as its reflection off of objects and particles in the environment decreases its intensity and thus contributes significantly to its attenuation.

Next, it is important to understand what it means to be non-linear and how or why a sound wave is non-linear. Non-linearity simply means that as the wave advances through the environment and time elapses, the conditions of the environment in which the wave exists do not remain constant. Explaining how or why a sound wave is non-linear is a little more complicated and requires the piecing-together of some facts which have already been noted. Because a wave’s frequency depends on the speed of sound, and the speed of sound depends on the density of the environment through which the wave is traveling, and the density of a fluid (fluids are gasses and liquids) environment depends on the pressure—which is fluctuating due to the nature of the wave—of the fluid, a wave’s frequency depends greatly, although transitively, on the pressure of the fluid. As a wave moves through various pressures, its frequency and speed change. Because the wave’s speed changes, the rate of diffusion changes as a result of its rate of expansion changing. Because of both the rate of diffusion changing, and because of the amount of particles to reflect off of (because of the compression and rarefaction, where lower densities have fewer particles and vice-versa) changing, the rate of attenuation changes. All of these factors are even further affected as the sound wave travels outwards because of the diminished intensity and conversely the diminished compression, rarefaction, attenuation, and diffusion. Thus, a sound wave is non-linear both in small segments (from one micro-environment to the next) and as an entire segment (as its intensity diminishes from the source at point A to the target at point B).

Because of sound’s non-linearity, it is extremely difficult to project a sound across long distances, and when a sound is projected across long distances, it becomes extremely distorted. So, logically, to counteract these effects, sound has to be given a linear quality. To do this, ultrasonic sound waves are used. Ultrasonic sound is sound that is above the human range of hearing (20,000Hz); in HSS, frequencies in the hundreds-of-thousands of hertz are used both to improve the linearity of the sound and to prevent harm to animals whose range of hearing exceeds that of humans. Because the frequency of ultrasonic waves is hundreds or even thousands of times faster than audible waves, and frequency is a measure of number of pressure fluctuations per second, the pressure fluctuates between rarefaction and compression hundreds or even thousands more times per second. Because the micro-environments are now hundreds or thousands of times smaller than with audible sound, the effects of the micro-environments on the propagation of the sound wave become negligible and thus the non-linear characteristics which were present in audible sound waves are not present in ultrasonic waves; additionally, because of the new nature of the pressure differences, the air through which it is traveling loses its non-linearity (because it is essentially “part of the wave”) and thus fails to make the sound “audible” (being “audible” would cause the sound to lose intensity and attenuate). This is what enables HSS to travel over incredibly long distances without losing intensity, becoming distorted, or propagating spherically outward rather than in a straight line. When the wave then hits an object that is non-linear, such as a wall or a human, the wave disturbs that object (because a wave is a disturbance of the surrounding environment and the object is its new environment through which to propagate) and uses that object to once again become “audible”.

In the case of these ultrasonic sound waves, though, “audible” does not really mean audible (ultrasonic is by definition inaudible); rather, what it means is that the wave once again becomes non-linear so that a theoretical human ear capable of hearing over 20kHz would be able to decipher it. To make the ultrasonic waves audible, a phenomenon known in music as the “Tartini tone” or in physics as the “difference tone” is employed.

A difference tone is a frequency that is generated when two other frequencies interact (in music: form a chord). This phenomenon is a result of both physical interaction between the two frequencies and neurological processing of the two frequencies. This phenomenon is almost like interference beats, which are caused when two frequencies of the same pitch are sequenced out-of-phase and thus cause the amplitude to fluctuate. A difference tone, however, is caused when two pure tones (perfect sine waves) are played in-phase but at different frequencies. When done at differences in frequency of over about 100Hz and not including the pitch which is at twice the frequency of the lower tone (in music: the octave)—which would be inaudible when using ultrasonic tones anyway--, a new tone with the frequency of the difference between the two original tones is created. For example: a tone at 440Hz (in music: the note “A”) and 660Hz (in music: the note “E” a perfect fifth higher) would produce the tone of 660Hz – 440Hz, which is the tone 220Hz (in music: an “A” an octave lower than the original “A”). What would not work is playing the tone 440Hz and it’s doubling at 880Hz, because the resultant difference tone would have the same frequency as the original 440Hz tone. Due to the capabilities of the human ear and aural processing centers in the brain, frequency differences which are almost a doubling of the original tone and frequency differences that are so small that the two tones are almost the same tend not to work; the ideal differences in tones are from 5:4 to 3:2 (in music, from a major third to a perfect fifth). This is because when the two tones are as such, the brain simply processes the sounds as being very dissonant, rather than allowing the difference tone to become clear.

Because the frequencies of ultrasonic tones are so high, however, a ratio of 5:4, when the original tones are at 400,000Hz and 500,000Hz, would still produce a difference tone of 100,000Hz--5 times the highest tone that is recognizable to a human. However, due to the temporary non-linearity of ultrasonic sounds, the ratios can be shrunken so that the difference of the tones can be as little as about 200Hz, producing a tone well within the human hearing range. The only disadvantage of this method of sound transfer is that bass tones (lower than about 200Hz) are unable to be reproduced. In applications like music, this would affect the harmonies and could make a performance sound “top-heavy”. In speech, the absence of bass tones, while not detracting from the decipherability of what is said (which is primarily affected by the mid- and high-range tones), would alter the timbre of a voice, which would affect the audience’s recognition of familiar voices and their ability to judge which person is speaking when 2 or more speakers are present (such as in a stage performance).

Hypersonic Sound technology is already being used in advertising displays so that sound can be projected only at one person without disrupting people who are not in the beam’s path, by the US military both to convey messages over long distances and, in a more powerful form, as a sound stun gun (LRAD). Future uses for this technology could include installation into automobiles—each passenger could hear their own music; concert halls—true surround sound projected from a central location and reflected off of the walls--sound projections could even move around the room in real-time; laptop computers—listen to podcasts, videos, or music without disrupting anyone else; and even in megaphones—whisper a message to one person instead of yelling over an entire crowd. HSS, now only in its relative infancy, will soon become a technology that will be seen frequently in a myriad of applications as prices shrink and popularity grows.



Equations:


Frequency (f) (measured in Hz) equals wavelengths (λ) per second (s) through a certain point:

f = λ/s

Speed of sound (v) equals wavelength (λ) times frequency (f):
v = λf

Speed of sound through an ideal gas (v) equals the square root of {[(the ratio of specific heats at a constant pressure {γ}) times (Boltzmann’s constant {k}) times (temperature in Kelvin {T})] divided by mass of one molecule of the gas (m)}:



Intensity of spherically radiating sound (I) equals power (P) (measured in watts) divided by surface area of a sphere (4πr):



Bibliography + Annotations:


Feynman, Richard P., Robert B. Leighton, and Matthew Sands. The Feynman Lectures on Physics. Reading, Mass.: Addison-Wesley, 1963.
Chapter 47 of this book explores the topic of sound waves and eventually their relation to electromagnetic waves and atomic harmonics. Equations are given in calculus format.

Kock, Winston E. Sound Waves and Light Waves. Garden City, NY: Anchor Books, 1965.
This book provides the fundamentals of sound- and light-wave motion and delves into the topic of propagation and dissipation of waves.

Levitin, Daniel J. This Is Your Brain on Music : The Science of a Human Obsession. New York: Plume, 2007.
This is an excellent book about how the brain processes music and sound. In addition, a section of the book is devoted to explaining the basics of music notation and jargon.

"Sound Attenuation." Sound Attenuation. 4 May 2002. Silex Exhaust Systems. 6 Feb. 2009 . This PDF file discusses various issues related to sound, sound transfer, and sound dampening (forced attenuation). Equations and explanations are provided here.

Wolfe, Joe. "Interference Beats and Tartini Tones." Music Acoustics. University of New South Wales. 2 Feb. 2009. .
This website from the University of New South Wales's physics department provides an excellent discussion of interference beats and difference tones. Audio examples are also provided here.



Links:

A link to a video of Woody Norris, a pioneer in popularizing Hypersonic Sound, provided by TED. Woody talks about the development of HSS technology, demonstrates the technology, and talks about the future of the technology. 15 minutes.

http://www.ted.com/index.php/talks/woody_norris_invents_amazing_things.html

YouTube video of Woody Norris demonstrating HSS as featured on the May 10, 2006 television show "Future Weapons." Unlike in the TED demonstration, in this video the technology is audible. 2 minutes.

http://www.youtube.com/watch?v=5imaJwfJMZ8


The Wikipedia article on "Sound from Ultrasound."

http://en.wikipedia.org/wiki/Sound_from_ultrasound


University of New South Wales's physics department's website pertaining to interference beats and difference tones.

http://www.phys.unsw.edu.au/jw/beats.html#Tartini


The official website of Audio Spotlight.

http://www.holosonics.com/


The official website of Woody Norris's Hyper Sonic Sound.

http://www.atcsd.com/site/content/view/34/47/