Exam (elaborations)
COMD 5070 - FINAL EXAM QUESTIONS WITH COMPLETE SOLUTIONS
COMD 5070 - FINAL EXAM QUESTIONS WITH COMPLETE SOLUTIONS
[Show more]
Preview 3 out of 28 pages
Uploaded on
August 18, 2024
Number of pages
28
Written in
2024/2025
Type
Exam (elaborations)
Contains
Questions & answers
Institution
COMD 5070
Course
COMD 5070
$12.49
100% satisfaction guarantee
Immediately available after payment
Both online and in PDF
No strings attached
COMD 5070 FINAL EXAM STUDY GUIDE
QUESTIONS WITH COMPLETE
SOLUTIONS
WhatsissScience?s-sans--1.sEmpirical:sBasedsonsdatas
2.sDeterministic:sObeyssphysicalslawssofscausesandseffects(notsrandom)s
3.sPredictive:sifsyousdosthis...thensthatswillshappen;sitscansinfluencesanothersones
4.sParsimonious:susessimplestsexplanationspossiblesrathersthansmakesunnecessarilyscom
plicateds-sboilsitsdownstositssessence,swithoutsdumbingsdown.
Semitoness(howsmanyssemitonessinsonesoctave,stwosoctaves,setc.)s-sans--
Octave:schangessmathematicalsrelationshipsbwsfrequencies;sdoublings(up)/halvings(down
)sofsfrequencys400=s800s(up)s&s200s(down)s
-12ssemitone=s1soctaves
200-400Hz=1soctaves=s12ssemitones
200-800Hz=2soctavess=s24ssemitones
SamplingsRates(Howsdoessitsrelatestosplaybacksquality,sfilessize,sfrequenciesssavedsinsRec
ording,snyquist,setc.?)s-sans---
SamplingsRates(snapshots):s(specifiedsinsHz)sfrequencysofsnumberssstored/writtenstosacts
asssampledsanalogssignal:sdiscretessnapshotssinsrapidssuccession.sLinkedswithsqualityslev
el.sNumberssrepresentsamplitudesvalues;smoressamples,sbetterscansrepresentsoriginalssig
nal.sRapidlyschangingssignalsneedssmoressamples/dosstosreflectsallschangessinsoriginal.sT
ooslowssamplingsrates&schangessmissed.sThinksmoviesshots@s30fpsslooksscontinuous.sAn
yslowersthoswouldslooksjerky.s
-
HighersSamplingsRatesselection:sgivessbettersfidelity;sBUT=sfiless&sstoragesdisksspacesus
essmoresmemorysprocessingstime,s&scomputationalspower.s
NyquistsFrequency=Halfsofssamplesrate.sHighestsfreQscansrecords&splaysbacksaccurately.s
Mustssamples@2xsratesofstheshighestsFreQ.sEx:sCDssamplesrate=44,100sampssorssnapsh
ots/secs&sstoresssignalsupstos22,050Hzs(correspondsstosupperslimitsofsmostspplsshearing).s
Tosrecordsupstos100Hz,sSR=200Hzs1/2sSamplesRate=NyquistsFreQs(100HZ=originalsamo
unt).
Filterstypess(Whatsdoshighsorslowspasssorsbandspasssorsbandsrejectsfiltersdo?s-sans--
1.sHighsPasssFilters-
sAllowssHighsfrequenciessthrusbutsremovessorsattinuatessorsholdsslowersFrequencies
2.sLowsPasssFilters-
sAllowsslowersfrequenciessthrusbutsremovessorsattinuatessorsholdsshighersFrequencies
,3.sBandsPasssFilters-
sAllowssasbandsofsfrequenciessthrusbutsremovessorsattinuatessorsholdsshighersandslowersFr
equencies
4.sBandsRejectsFilters-
sAllowssasbothshighersandslowersfrequenciessthrusbutsremovessorsattinuatessorsholdssasba
ndsofsFrequencies
FouriersTransforms/Analysiss(WhatsissresultsofsFourierstransformsofssound?sWhatsdoessitsr
evealsaboutscomplexssound?)s-sans--
Camesupswithswaysofsanalyzingscomplexssignalssandsdecomposingsorssplittingsthemsintos
asseriessofsindividualscomponents.
1.Allsperiodicssoundssaresmadesofsascombinationsofssinceswaves.
2.Youscansbreaksthemsup
-amplitudessvarys(howsbigstheysare)
-phasesanglessvarys(wherestheysaresinscycle)
-frequenciessvarys(manysdifsfreq)
FourierstaughtsussthatsEvenscomplexssoundsscansbesbrokensdownsintostheirsind.ssinoudals
components.
FOURIERsTRANSFORMs-
sWestakesastimesdomainswaveforms(likesmicrophonessignal)sandswesanalyzesitsandscreates
asspectrumsfromsthestimesdomainswaveform
-sAnalyzesascakestoslearnsitssingredients.
Timesdomainsx-timesy-
amplitudesTHENsfourierstransforms=Frequencysdomains(spectrum:slicesinstime)sx-
frequencysy-amplitude
TimesDomainsDatas(TDD)s-sans---aswaveformsrepresentsssoundsdirectly
-(airspressure)schangessoverstime.
x-sTIMEsy-sAMPLITUDE
TypessofsSpectra:sLine,sFFT,sLPC,s(theysrevealsdifferentsfeaturessofsspeech;sWhatsisseac
hsonesbestssuitedsfor?)s-sans---Waveform:sTimesdomainsviewsofssound.s
-
Spectrum:sFreQsdomainsviewsofssound.sIndivsingredients@ssinglespointsinstime.s(X=FreQ;s
Y=Intensity)sSpectra:sPluralsformsofsspectrum
-
LinesSpectrum:s(TypesofsFDD)sshowssFreQscomponentssofsperiodicssound.sSingelsvertica
lsline=Sineswave.sEasverclslinesisssinglesfreQ=
NyquistsFrequencys-sans--
ThesNyquistsfrequencyscomessintosplaysduringsthesrecordingsorsdigitization.sIfsyousselectsa
ssamplesratesof,ssay,s16,000sHz,sthensthesNyquistsisshalfsofsthat,sors8,000sHz.sThissmeanssi
nspracticalstermssthatsfrequenciessupstos8,000sHzsinsyoursincomingsanalogssignals(fromsth
, esmicrophone)swillsbescorrectlysrecordedsandscansbesplayedsback.sButsfrequenciessaboves
8,000sHzswillsnotsbessaved.sMostsmodernsrecorderssandscomputersswillsautomaticallyslow-
passsfiltersyoursincomingssignalsatsthesNyquistsfrequencystospreventsaliasing,swhichswould
scontaminatesyoursrecordingswithsincorrectlyssavedsfrequenciessthatsaresabovesthesNyqui
st.sSosit'ssduringsrecordingsthatswesaresinterestedsinsthesNyquist.
FouriersTransforms=sFrequencysdomainsdisplays-sans---
AslinesSPECTRUMsshowssthesfrequencyscomponentssofsasperiodicssound.
-Frequencysdomainsdescriptionsofsthessignal
-hassharmonicssthatsaresmultiplessofsthesfundamental
-hassnothingsbwsthesliness(theslinessrepresentsthesharmonicsfrequencies)
x-FREQUENCYsys-sAMPLITUDE
Hasssplitsupseachsindividualssoundsfromsthestotalssosuscanshearseachsindividually.
Benefits-
sisstosshowsusswhatsthosesind.scomponentssaresandsthesrelativesproportionssofseachsheigh
tsofseachsbarssayssthesstrengthsofseach.
Thississlikesourshumansspeechsandsitsiss'NEARLYsPERIODIC'
Thesuppersharmonicssgetsprogressivelysweakersassyousgosupsinsfrequency.
WhatsdoesseachsTimesvssfrequencysdomainsdisplayssshowssussaboutssound.s-sans--
1.sTimesdomains-sshowssussthestimesview
x-stimesy-Amplitude
2.Frequencysdomains-sshowssussthesfrequencysview.s-
sineswave:ssingleslinesonsasspectrum
-complexsperiodicssignals:smultipleslines
x-frequencysy-Amplitude
3.Whatswouldsnoiseslookslikesinsasspectrums?
-allsfrequencies
-equalsamplitude
-randomsphase
Spectrums-(FFT)sFastsFouriersTransforms-sans--
1.sClearlysshowssussrangesofsHARMONICsenergy
2.seachspeaksissasharmonica
3.sBUTswillsnotsshowsuss/slesssclearsatsshowingsformants.
4.smoresrevealingsofsthes(source)s=sproblemswithsvoice
Spectrums-s(LPC)sLinearspredictivescodings-sans--1.sShowssSPECTRALsENVELOPEs
2.sgoodsatsrevealingsformants
3.sdoesn'tsshowsharmonics
4.moresrevealingsofsfilters=sarticulationsproblems