Comps Study Guide PDF

`Physics, Physiology, and Psychology of Sound: Unit 1: ​ Sine waves: sinusoids, pure tones ○​ Things have mass, Things have springiness, Things are acted upon by force → creates simple harmonic motion ○​ Amplitude (A or v): always a pos (+) #, difference between two...

`Physics, Physiology, and Psychology of Sound: Unit 1: ​ Sine waves: sinusoids, pure tones ○​ Things have mass, Things have springiness, Things are acted upon by force → creates simple harmonic motion ○​ Amplitude (A or v): always a pos (+) #, difference between two points ○​ Frequency (f): determines period ​ 1/period ○​ Phase ○​ Period (T): how much time it takes to complete one repetition of a sine wave ​ 1/frequency ​ Complex waves: every sound signal is made up of pure tones, no matter how complex ○​ In-phase - sum of amplitudes ○​ Out-of-phase - amplitudes cancel eachother out (how noise-cancelling tech works) ○​ Fixed phase - click train/pulse, starting phase known and precise ○​ Random phase - noise, same tones as click train/pulse but starting phase unknown and imprecise ○​ Gaussian ​ Distribution → bell curve ​ white noise (broadband) → flat frequency spectrum bc all frequencies are equally represented but random at any given point in time -​ Sample for short time → frequency spectrum looks like white noise -​ Sample for infinitely long period of time → flat frequency spectrum ​ FFT (fast fourier transform): converts signal from original time domain into spectral (frequency) domain ​ Longitudinal waves - displacement of molecules is parallel to the propagation ○​ Waves go L→R, molecules go L→R ○​ Ex. sound waves ​ Transverse waves - propagation and displacement of molecules are perpendicular to each other ○​ Waves go L→R, molecules go up→down ○​ Ex. earthquake movement ​ Surface waves - combination of transverse and longitudinal waves ○​ Waves go L→R, molecules move in a circle ○​ Ex. water waves ​ Propagation: ○​ When sound encounters a new medium it is:  ​ Reflected (reverb), absorbed (friction or heat dissipation), and transmitted ○​ Inverse square law: intensity decreases as a function of distance ​ Function of friction and energy is dispersing more (must be distributed over a larger area) ○​ Monopole - propagation expands in all directions ​ Ex. balloon expanding and contracting from pump ○​ Dipole - displacement occurs in one direction ​ Max displacement on axis in direction of propagation, min displacement on axis perpendicular to propagation ​ Ex. loudspeaker ○​ Variables of propagation: if you fix one variable, and change another, the third HAS TO CHANGE) ​ Speed of sound ​ Wavelength (λ) - period for propagating sound, can be expressed as distance when speed is constant ​ Frequency ​ Reflection of waves: ○​ Theoretical fixed (hard)) boundary - could never happen in real life ​ Displacement remains zero ​ Phase flips (180° phase change) (polarity changes) ​ All energy is reflected back and there is no loss to absorption or transmission ○​ Theoretically free (soft) boundary ​ Restoring force is zero ​ No phase change (polarity remains the same) ​ More displacement in new medium, displacement is ‘free’ in new medium ○​ Low to high density (similar to fixed boundary) ​ Reflected sound is out of phase, transmitted sound is in phase ​ Some energy is reflected back, some is transmitted through ○​ High to low density (similar to a free boundary) ​ Reflected and transmitted sound is in phase ​ deciBel (dB) ○​ Based on log scale ○​ Weber’s law: k = ΔI/I ​ As overall intensity (I) increases, the amount of change required to notice also increases ​ ΔI ≈ 1 dB (just noticeable difference) ​ Frequency = octave scale (octave = doubling in frequency) ○​ dB = 10log(I2/I1) ​ I2 is sound you are trying to report intensity of ​ I1 is average threshold for healthy individuals at given frequency ○​ dB HL (hearing level) - 0 dB HL means no difference between patient being tested and threshold or average healthy population  ​ Not universal, based on transducer ○​ dB SPL (sound pressure level) - 20 μpascals ​ Completely universal ○​ dB SL (sensation level) - based on particular patient’s thresholds ​ Expressing sound level ○​ Pure tones: ​ All acoustic energy or power is concentrated at one frequency ​ The level is the dB at that frequency ○​ Complex sounds: ​ Acoustic energy is distributed across more than one frequency ​ Must add underlying linear intensities then convert to dB ○​ When you double an intensity → 3 dB increase (for any intensity or reference) ○​ For any # of sounds that are equal in level: ​ dB increase = 10 log # sounds ​ Overall level = dB increase + original dB ​ Filters - change the intensity of some frequencies in a signal ○​ Low pass filter (low freqs NOT filtered) ​ Lots of attenuation at high freqs ○​ High pass filter (high freqs NOT filtered) ​ Lots of attenuation at low freqs ○​ Bandpass filter  ○​ Stopband filter ○​ Cut off freq (aka 3dB down point aka half powerpoint) ​ Low pass - point below which freqs are not filtered at all ​ High pass - point above which freqs are not filtered at all ○​ Slope of filter → dB/octave (tells you how well the filter is working) ​ Signals ○​ Analog signals ​ Continuous in time and amplitude ​ Ex. record player, cassette ○​ Digital signals ​ Discrete steps in both time in amplitude ​ Information is lost when signal is digitized ​ Sampling rate or frequency → how often you are going to sample the amplitude in time ○​ Converting signals ​ Analog to digital (A/D) - converting continuous time values into discrete time values ​ Sampling period: 1/SR = time between samples ​ Lower sampling rate = worse signal ​ Digital to analog (D/A) - ​ Nyquist theorem: if you want to accurately represent a signal of a given frequency, then you must sample at at least double that frequency ​ Open to aliasing or quantization errors → can be fixed using anti-aliasing filter (turns into sine wave) ○​ Quantization: converting a range of analog values into discrete set of values ​ Trade off between precision and storage ​ Psychophysics ○​ Weber's law: a just-noticeable difference is proportional to the magnitude of the stimulus ​ k = Δi/I ○​ Fechner's law: sensation is proportional to the logarithm of the stimulus intensity ​ S = k logR ○​ Formalized psychophysical techniques: used to measure both absolute (quiet threshold) and difference thresholds  -​ Thresholds obtained with any of the classical methods may not be a good estimate of sensitivity because the threshold estimate can be influenced by sensory factors (sensitivity) and non-sensory factors (response bias) 1.​ Method of limits a.​ Find ascending and descending thresholds and average b.​ Carhart-Jerger method = modified method of limits 2.​ Method of adjustment a.​ Subject adjusts the stimulus level to the point where they can just barely hear the sound 3.​ Method of constant stimuli (most used) a.​ Random or quasi-random presentation of stimulus to find threshold ​ Signal detection theory → attempts to address problem of non-sensory factors influencing threshold (used to separate sensory and non-sensory factors) ​ Fundamental detection task: ○​ Subject responds if signal was present or not ○​ Response matrix: ​ Signal detected + Questions: 1.​ Main purpose of a cochlear implant? a.​ To restore hearing via electrical stimulation of the cochlear nerve by bypassing damaged inner and outer hair cells. 2.​ How does a CI work? a.​ External processor collects sound and converts it to an electrical signal which is sent to the internal components via radio waves. The internal electrode in the cochlea releases electrical stimulation directly to the cochlear nerve bypassing inner and outer hair cells. 3.​ What do CIs keep vs give up in sound signals? a.​ Cochlear implants keep the amplitude envelope and removes the temporal fine structure. 4.​ What are the 3 steps in envelope extraction in CI speech processing? a.​ Bandpass filtering b.​ Rectification - make negative peaks positive (full-wave) or 0 (half-wave) c.​ Amplitude modulated pulse train 5.​ Main limitations of CIs? a.​ Current spread - cochlea is a fluid-filled space so electrical stimulation spreads out b.​ Frequency warping - the array of electrode is shorter than the basilar membrane (and cochlear nerve) so low frequency sounds are forced into mid-frequency part of the nerve 6.​ What is upward spread of masking? a.​ A tone will mask frequencies above it rather than below it because of basilar membrane displacement 7.​ What are the two types of masking? Examples? a.​ Energetic masking - simultaneous masking (tone and masker presented at the same time) b.​ Temporal/informational masking - forward or backwards masking (tone presented either before or after masker) 8.​ What is psychoacoustics? a.​ Study of the physical sound stimulus and perception of that stimulus and the interaction between them 9.​ How does the brain localize sound in the horizontal plane? In the vertical plane? a.​ Horizontal - ITDs (interaural time differences) and ILDs (interaural level differences) i.​ Phase wrapping in ITDs ii.​ Head shadow effect in ILDs b.​ Vertical - shape of the pinna 10.​What is the inverse square law? a.​ Energy/intensity decreases as a function of distance (because of friction and dispersal of energy over an increased space)... decrease of 6dB with a doubling of distance 11.​What is the FFT? What is a test that uses it?  a.​ Converts time domain to frequency domain. b.​ TEOAEs 12.​Best sound sensitivity? a.​ 2kHz-5kHz 13.​Three things that effect pitch perception? a.​ Frequency b.​ Intensity c.​ Duration 14.​What are the different types of filters? a.​ Low-pass b.​ High-pass c.​ Band-pass d.​ Band-stop 15.​What is the Volley theory? What coding mechanism does it fall under? What frequency is it best for? a.​ Collective group of neurons firing toegther so that all positive peaks are represented (not every neuron can fire at every peak because of the refractory periods of the neurons) b.​ Rate mechanism c.​ Best for low frequencies (until 4kHz ish) 16.​ Frequencies for place mechanisms? a.​ Higher frequencies (beginning at 4-5kHz) 17.​What are standing waves? a.​ Combination of two waves that can either sum or cancel out (can combat using warble tones) 18.​What is fundamental frequency? a.​ Lowest frequency of a wave form (first harmonic) 19.​What are harmonics? a.​ Whole number multiples of a fundamental frequency 20.​What is gaussian white noise a.​ Noise evenly distributed across all the frequencies (may not seem this way at any single point in time, but will be the case over an infinitely long period of time) - flat frequency spectrum across time 21.​Where in the auditory pathway are ILDs being processed? ITDs? a.​ LSO (lateral superior olive) b.​ MSO (medial superior olive) 22.​Binaural summation? a.​ 3dB gain associated with binaural hearing 23.​Minimum audiblity curve? a.​ Average threshold for healthy individuals at every frequency 24.​SL? HL? SPL? a.​ Sensation level - compared to average healthy individuals b.​ Hearing level - compared to individual prior hearing thresholds c.​ Sound pressure level - compared to 20 micropascals 25.​Most speech sounds are produced within which frequencies? a.​ 1kHz - 4kHz 26.​Whats the Nyquist theory? a.​ In digital to analog conversion, if you want to accurately represent a signal of a given frequency, then you must sample at at least double that frequency 27.​What is tonotopy? a.​ Organization by frequency in cochlea and auditory pathway 28.​What is loudness recruitment? a.​ When the difference between threshold of hearing and UCL is small - a reduced dynamic range, outer hair cells amplifies softer sounds more than louder sounds 29.​What is an example of an anolog signal? a.​ Records, AM & FM radio wave, pure tones 30.​What is an example of an digital signal? a.​ computers 31.​What does a decibel always need? a.​ A reference