Bat Echolocation Flashcards
Why ultrasound
good for small objects
disadvantages of ultrasound
strong attenuation, limit range, so need to do louder/longer calls
Discovery of ultrasound and how calls affected by search
Griffin, higher rate/shorter duration as approach target, search/approach/terminal
Two (3) types of call
FM = freq.modulated, broadband.
CF = constant.freq, narrowband
CF-FM = both
Describe FM signals
short <5ms, quickly sweep down range of freqs, harmonics
Describe CF signals
Longer 30ms, narrowband
Distance estimation
FM signal Delay time of pulse-echo, can do within 10-15mm 75% accuracy threshold
Acuity of echo delay
Jitter experiment, good to 10ns, enough of 2um in target = acoustic texture, moth from wall
Why FM for distance
Broadband gives good range for time delay across points
Angular Size?/absolute
Amplitude/loudness
Get absolute size from distance (delay) + subtended angle (loudness)
Location cues?
Azimuth by using interaural timing, elevation by interaural intensity
How doppler shift works
If approach stationary = higher freq
Receding = lower freq
appraoching object = higher freq
Doppler shfit analysis (what 2 things analyze about target used for) mustached bat
Use CF signal since constant, good analysis
USE TO DETERMINE VELOCITY
Doppler shift compensate to KEEP ECHO IN ACOUSTIC FOVEA
Used fo FLUTTER DETECTION ALSO, doppler echo of flapping wings of insect (+/- 30Hz wow!, 0.04% change in freq heard)
FM vs CF
FM - broadband good for hunting on ground with vegetation, good for texture/distance
CF - narrowband, good for open fields, doppler shift/velocity/flutter
Explain harmonics of mustached bat CF-FM
2nd harmonic at 61kh is loudest, this is where their acoustic fovea is
1st harmonic is softest bc only they can hear it (allow for combo sensitive neurons to fire only for their calls/echos)
Adaptations of CF-FM mustached bat aud system
Acoustic fovea over-rep. of neurons for CF component of call
During doppler compensation they reduce sensitivity to loud call so can hear quiet echo
Describe q10db factor
Quality factor, how sharply tuned the neurons are
Display frequency and dB threshold graph, see char freq. point and go up 10Db, HOW WIDE IS the factor, higher factors are better weirdly
Most common in CF bats, FM bats don’t have as good sharp tuning since broadband
Bats and aud sensititvity HOW DO IT
Reduce sensitivity DURING call, so that they are ready to hear the soft echo
FM bats - contact middle ears to attenuate loudness
CF bats - cant contract since freq overlap, so call is LOWER then doppler echo, ECHO FALLS ON FOVEA, CALL DOESN’T
Aud Path in bats
Cochlear nucles - sup. olive - lateral lemniscus - Inferior colliculus - Medial geniculate nucleus - Aud cortex
Post-stimulus time histogram what show
Time from onset of stim and then neurons firing,
Louder sound = shorter latency
EXCEPT IC NEURONS - accurately encode pulse-echo delay so fire without wavering when hear sound even if louder no jitter
USE IC NEURONS FOR DISTANCE ESTIMATION
IC neurons
Accurately encode pulse-echo delay
Low spiking threshold sharply tuned to a freq. in an FM sweep
Respond to pulse-echo pair of particular delay
FM-FM area
axis for delay in vert columns and axis for harmonic pairing (FM1 to FM2/3/4),
Map of delay, Tuned to DELAY AND AMPLITUDE (respond only to soft echoes)
ENCODE SIZE (loudness/delay) AND DISTANCE
FM1 Pulse is actively SLOWED to reach MGN coincidence with echo at SAME TIME
columnar organization duh
CF-CF area
Encode velocity
Respond only with pulse and doppler echo, combination sensitive
Axis for Delay time and axis for pairs (CF1 to CF2/3), overlap so allow compute velocity
Doppler shifted CF area (DSCF)
Process doppler shifted CF calls
sharply tuned around 61khz
Tuned to FREQ AND AMPLITUDE OF ECHO, NOT CALL, acoustic fovea
columnar too, particular CF and amplitude axes
