CSD 303 - Exam 2 || With 100% correct answers.
- auditory perception: sound transformation up to the auditory pathway 1. sound waves (air)
2. ossicular vibrations (bone)
3. basilar membrane waves (fluid)
4. hair cell/ stereocilia movement 5. neural firing correct answers how is sound transformed along the auditory pathway?
- OUTER: pinna and air canal resonance - pinna creates the range of sensitivity on equal loudness curve correct answers how is the auditory representation affected by the transformations?
- individuals are more sensitive to sound in middle frequencies of the curve and less sensitive to higher and lower frequencies of the curve - most sensitive represented by the lowest dip in the curve - We have to play low- and high-frequency sounds at greater amplitude to hear them at the same loudness as mid-frequency sounds. correct answers equal loudness curve
frequency: cannot hear below 20 Hz or above 20,000 Hz
loudness: sound is inaudible below a certain threshold correct answers absolute limits
frequency: hard to tell apart 1001 Hz and 1002 Hz
loudness: hard to tell apart 80 and 81 dB
timing: hard to tell the order of tones (or whether they are simultaneous) if they are less than 20 ms apart
duration: hard to tell 200ms from 220ms
direction: hard to tell is sounds are less than 1-15 degrees apart horizontally (vertical localization is even harder) correct answers discrimination limits
- the auditory representation of loudness is NONLINEAR and DEPENDS ON FREQUENCY
- the auditory representation of frequency is NONLINEAR, or more sensitive to changes at the low end than at the high end (this is due to the cochlear/basilar membrane, which has a nonlinear representation of frequencies along its length) correct answers linear vs. nonlinear auditory representations of loudness and frequency
bark: an auditory (or psychoacoustic) frequency scale (corresponds with the bandwidth of the
cochlear filter; think of it as similar distance along the cochlea)
Hz: an acoustic frequency scale (cycles/s); properties of sound - lower frequencies are more sensitive to change, so they are higher in bark (vice versa for higher frequencies) - more of the bark scale is lower frequencies because this is what one hears most - low frequency sounds spread apart, and high frequency sounds are squished together correct
answers frequency scales - Hz (linear) vs. Bark (nonlinear)
neurons need time to "recharge their batteries"; neurons will not respond as strongly to sounds that follow right after other sounds correct answers neural saturation
- each neuron responds to multiple frequencies
- if a neuron is already firing to one frequency (e.g. 1000 Hz), it cannot increase firing much to another (e.g. 1100 Hz)
- sensitivity is reduced in the presence of a masker, so sounds will not necessarily be heard as
well (sounds that are followed by or follow silence are more easily heard) correct answers frequency masking correct answers temporal masking
- when visual and auditory speech signals conflict, it can change the way we hear speech sounds - the integration of conflicting cues
- this proves that visual information contributes to speech perception - speech is not just acoustics - race, gender, and stereotypes also effect perception correct answers McGurk Effect
the vowel prototype acts like a magnet, pulling your perception towards it. It is harder to tell apart stimuli near the prototype because they are all perceived similarly the closer you get, and are therefore less different from each other correct answers Perceptual Magnet Effect
- how we perceive sounds that vary along a particular dimension
- between-category discrimination is good (discriminating /p/ from /d/)
- within-category perception is poor (discriminating /p/ from /b/)
- this sensitivity to sounds around the category boundary can be changed by language experience, however, there are probably some innate properties of the auditory system that result in natural sensitivities to certain VOT regional therefore, languages "prefer" to put boundaries in certain places to make them more easily discriminable 1. snake-shaped graph: identification of a singular sound 2. peak graph: are two sounds the same or different?
- regardless of the VOT it will always be perceived as the same, but at that category boundary
we are good at discriminating (very high at category boundary) correct answers Categorial Perception
- how many times was the phoneme 'x' heard as 'y'?
- confusion matrices and distance maps: "confusion scores" give us a way to think about the similarity between two phonemes
-when one phoneme is confused with another, we can say the two are similar: close in perceptual space
- auditory perception: sound transformation up to the auditory pathway 1. sound waves (air)
2. ossicular vibrations (bone)
3. basilar membrane waves (fluid)
4. hair cell/ stereocilia movement 5. neural firing correct answers how is sound transformed along the auditory pathway?
- OUTER: pinna and air canal resonance - pinna creates the range of sensitivity on equal loudness curve correct answers how is the auditory representation affected by the transformations?
- individuals are more sensitive to sound in middle frequencies of the curve and less sensitive to higher and lower frequencies of the curve - most sensitive represented by the lowest dip in the curve - We have to play low- and high-frequency sounds at greater amplitude to hear them at the same loudness as mid-frequency sounds. correct answers equal loudness curve
frequency: cannot hear below 20 Hz or above 20,000 Hz
loudness: sound is inaudible below a certain threshold correct answers absolute limits
frequency: hard to tell apart 1001 Hz and 1002 Hz
loudness: hard to tell apart 80 and 81 dB
timing: hard to tell the order of tones (or whether they are simultaneous) if they are less than 20 ms apart
duration: hard to tell 200ms from 220ms
direction: hard to tell is sounds are less than 1-15 degrees apart horizontally (vertical localization is even harder) correct answers discrimination limits
- the auditory representation of loudness is NONLINEAR and DEPENDS ON FREQUENCY
- the auditory representation of frequency is NONLINEAR, or more sensitive to changes at the low end than at the high end (this is due to the cochlear/basilar membrane, which has a nonlinear representation of frequencies along its length) correct answers linear vs. nonlinear auditory representations of loudness and frequency
bark: an auditory (or psychoacoustic) frequency scale (corresponds with the bandwidth of the
cochlear filter; think of it as similar distance along the cochlea)
Hz: an acoustic frequency scale (cycles/s); properties of sound - lower frequencies are more sensitive to change, so they are higher in bark (vice versa for higher frequencies) - more of the bark scale is lower frequencies because this is what one hears most - low frequency sounds spread apart, and high frequency sounds are squished together correct
answers frequency scales - Hz (linear) vs. Bark (nonlinear)
neurons need time to "recharge their batteries"; neurons will not respond as strongly to sounds that follow right after other sounds correct answers neural saturation
- each neuron responds to multiple frequencies
- if a neuron is already firing to one frequency (e.g. 1000 Hz), it cannot increase firing much to another (e.g. 1100 Hz)
- sensitivity is reduced in the presence of a masker, so sounds will not necessarily be heard as
well (sounds that are followed by or follow silence are more easily heard) correct answers frequency masking correct answers temporal masking
- when visual and auditory speech signals conflict, it can change the way we hear speech sounds - the integration of conflicting cues
- this proves that visual information contributes to speech perception - speech is not just acoustics - race, gender, and stereotypes also effect perception correct answers McGurk Effect
the vowel prototype acts like a magnet, pulling your perception towards it. It is harder to tell apart stimuli near the prototype because they are all perceived similarly the closer you get, and are therefore less different from each other correct answers Perceptual Magnet Effect
- how we perceive sounds that vary along a particular dimension
- between-category discrimination is good (discriminating /p/ from /d/)
- within-category perception is poor (discriminating /p/ from /b/)
- this sensitivity to sounds around the category boundary can be changed by language experience, however, there are probably some innate properties of the auditory system that result in natural sensitivities to certain VOT regional therefore, languages "prefer" to put boundaries in certain places to make them more easily discriminable 1. snake-shaped graph: identification of a singular sound 2. peak graph: are two sounds the same or different?
- regardless of the VOT it will always be perceived as the same, but at that category boundary
we are good at discriminating (very high at category boundary) correct answers Categorial Perception
- how many times was the phoneme 'x' heard as 'y'?
- confusion matrices and distance maps: "confusion scores" give us a way to think about the similarity between two phonemes
-when one phoneme is confused with another, we can say the two are similar: close in perceptual space
