Prof. Greg Francis1PSY 310: Sensory and Perceptual ProcessesPurdue UniversitySoundPSY 310Greg FrancisLecture 28Why doesn’t a clarinet sound like a flute?Purdue UniversityOther senses Most of this course has been about visual perception Most advanced science of perception Perhaps the most important human sense Relatively easy to show many effects Many topics apply to other senses For the rest of the course we discuss other senses, inless detail Sound (up to third exam) Touch Smell TastePurdue UniversityThe perceptual processAttended stimulusEnvironmental stimulusActionRecognitionPerceptionProcessingTransductionStimulus onreceptorsPurdue UniversityAuditory perception As for visual perception, auditory perception (hearing)provides information about the world around us Sounds can be heard even when objects are out of sight Behind other objects Night time We are quite good at recognizing sounds demonstrationPurdue UniversitySound stimulus The stimulus of sound is pressure changes Usually in the air, but can also be produced in water or othermediums Vibration makes sound e.g., hitting a tuning forkPurdue UniversitySound stimulus Air around the fork is pushed back and forth This makes small changes in the pressure of particles in theair Pressure is the density of particles in the air When particles are pushed, they move, and push theirneighbors. The energy travels along.Prof. Greg Francis2PSY 310: Sensory and Perceptual ProcessesPurdue UniversitySound stimulus Air around the fork is pushed back and forth This makes small changes in the pressure of particles in theair Pressure is the density of particles in the air When particles are pushed, they move, and push theirneighbors. The energy travels along.Purdue UniversitySound stimulus There are both compressions and rarefactions to produce asound wave Pushed away from source in compressions “Sucked” back in during rarefactions It doesn’t really matter what the source of the pressurechanges is, a string is a good as a forkPurdue UniversitySound stimulus If you could see the particles in the air with a soundwave, they would look like thisPurdue UniversitySound stimulus So sound is a pressure wave We can describe a simple sound with a sine wavePurdue UniversitySound stimulus Different properties of the wave generallycorrespond to different perceptual aspects of soundPurdue UniversitySound stimulus The frequency of the wave corresponds (roughly) to pitch Here is the wave for middle C Play the sound A Hertz is the number of cycles of the wave in a second of timeProf. Greg Francis3PSY 310: Sensory and Perceptual ProcessesPurdue UniversitySound stimulus Increasing the frequency of the wave changes the pitch Here is the wave for middle A Play the soundPurdue UniversitySound stimulus Increasing the frequency of the wave changes the pitch Here is the wave for high A Play the soundPurdue UniversitySound stimulus What we perceive as loudness corresponds (roughly) to theamplitude of the wave How much pressure is changed This is a matter of how much energy is involved in pushing the air A sound spreads out across an area, and the energy is constant So sound waves have a smaller amplitude with increasing distance Inverse squarePurdue UniversityMeasuring amplitude The easiest way to physically measure amplitude would seemto be just measure the changes in pressure Micropascals (one-millionth of a pascal) Pascals are used to measure pressure for atmospheric changes But this proves to not be useful because variations inamplitude are perceived variations in loudness People have varying sensitivity to sound wave amplitude We discriminate small differences when the amplitude is small Whisper versus talk quietly We cannot discriminate big differences whenthe amplitude is big Jet engine versus jet engine and someone shoutingPurdue UniversityDecibels A measure that is relatively similarto human perception of loudnessis the decibel scale Take the ratio of a sound’s amplituderelative to some fixed soundamplitude Take the logarithm Multiple by 20 (or whatever, it justscales the number)€ dB = 20logpp0      Purdue UniversityDecibels Note, you can get negativedecibels When your sound amplitude is smallerthan the reference Psychologists use a reference of 20micropascals, which is just about thefaintest sound you can hear When the decibel value equals zero,the sound amplitude is the same asthe reference€ dB = 20logpp0     Prof. Greg Francis4PSY 310: Sensory and Perceptual ProcessesPurdue UniversityDecibels Many people (including our textbook) suggest that the decibelscale is better than just amplitude because it allows us to workwith a smaller range of numbers It is true that sounds in the environment cover a very large rangeof values Faintest sound is 20 micropascals Whisper is 630 micropascals Loud radio 6,300,000 micropascals Jet engine 6.3 billion micropascals But changing the numbers is not the reason for using the decibelscale Who can’t work with big numbers but can work with logarithms of ratios? The real motivation for the decibel scale is that the numbers itassigns to different pressure amplitudes follow a pattern fairlysimilar to our perceptions of loudnessPurdue UniversityDecibels Consider different sounds and how loud they are Loudness demoPurdue UniversityDifferent wave forms We hear lots of sounds that are not sine waves Demos of flute, clarinet If sound is a vibration that produces a sine wave, then whatproduces non-sine wave sounds? Fourier analysis / synthesis All wave patterns can be broken down into sine waves A lot of different sine waves combined together produce a wave that isnot a sine wavePurdue UniversityFourier analysis A function can be described either in space (x) Or in terms of the Fourier coefficients (a0, a1,…,b1, b2,…) Each of the coefficients refer to a sine wave of a given Hertz€ f (x) =xπ€ an=−4nπ( )2€ a0=12€ bn= 0€ x€ nPurdue UniversityComplex sounds Different sounds can be described by the Fourier coefficientsthat correspond to the shape of the wave400 Hz sine wave400 Hz square wavePurdue UniversityComplex sounds Different sounds can be described by the Fourier coefficientsthat correspond to