bQ ML iidcen
`
`4eor Phs ednlv
`Professor LG Mismann
`nstitut fur Theoret Nachrichter.tchn1k
`Universit(cid:31)t
`Hannover
`Appeistrasse
`D3000 Hannover
`B.R.D
`
`ISO/IEC JTC1/SC2/WG8
`MPEG 90/013
`
`atd o.pt
`
`aht
`
`-af zecrefl
`
`BCL
`ADC
`AR29B06/90
`
`onderw re
`conc betr
`
`doorkiesnummer
`acces
`
`intern dir
`
`in-diaking
`durchwohi
`
`0407 32801
`
`datum date
`
`19900216
`
`Dear Prof Musmann
`
`one page was missing in the description of
`Unfortunately
`last year
`the end of
`MUSICAN which was sent
`to you at
`
`Please find page
`
`and
`
`of
`
`the text
`
`in enclosure
`
`Furthermore the table ItDifferential scale
`incorrect
`was
`
`factor coding
`
`The updated table is included
`
`in enclosure
`
`With kind regards
`
`De Wit
`
`Copy
`
`Messrs Chiarigi ione
`Dehery
`Stoll
`Takahashi
`
`CSELT
`CCETT
`IRT
`MEl
`
`Torino
`Cesson Sevigne
`M(cid:31)nchen
`Kadoma
`
`Enclosures
`
`Nederlandse
`
`Philips Bedrijveri B.V
`
`Eindhoven
`
`Handeisregister
`
`Nederiand
`Eindhoven
`
`no 8551
`
`Teex 35000
`phtc ni
`Tnat
`040791111
`nt 31 40
`7911
`
`11
`
`1
`
`HP 1031
`
`

`

`iocaicn is de from zri nasriq threshold whC i-
`1i ci
`ulaad eh 24 ins
`is r1erV
`thresto1d
`The maskina
`is calcu1aed by
`the power density
`soectrura
`escirnate
`to the subafld
`FFT The FFT is calcuiate
`in parallel
`1024point
`audio sna1 is
`fiierinc 3afore
`FFT
`the
`calculation
`of
`the
`Hanningwindow
`windowed
`by
`Technical data
`sampling rate of t2 48 kHz
`FT up to 24 kHz with
`21.3 ins with
`of
`window
`to
`Samples
`corresponding
`48 kHz
`resciutton
`
`46.88 Hz with 512
`
`frequency
`
`samples
`
`up to
`
`2.024
`
`Lt2
`Frequency
`24 kHz
`Non overlapping window
`
`The Hanningwindow
`hi
`
`0.5
`
`is given by the following expression
`cos2pii/nl
`ni
`
`1024
`
`The power density spectrum is estimated as
`n-i
`hl sl
`xk
`exp_jkl2 /n12 dB
`k0....5i2
`
`10
`
`log
`
`10
`
`CALCULATIoN OF THE MASKING THRESHOLD
`
`calculation
`The
`following steps
`
`of
`
`the masking
`
`thresholds
`
`consists
`
`of
`
`the
`
`and non-tonal components
`tonal
`Finding of
`individual masking thresholds
`Computation of
`the global masking threshold
`Computation of
`in each
`the signal-tomask-ratio
`Calculation of
`
`subband
`
`These
`
`steps will be discussed in the four
`
`following sections
`
`Finding of
`
`tonal
`
`and nontonal
`
`components
`
`the FFT-spectrUm
`first step it is necessary to find out of
`In
`for calculating the
`components and noise components
`the tonal
`global masking threshold
`
`2
`
`

`

`cne xaozion ol Lo.a1 cone.nt5
`sarts
`nat
`is knwn
`enerimeflzs
`isvchcacoutiC
`It
`the cwer than in the
`ear is better
`resolution
`frcxency
`resolution df cr
`i1i9r frequency
`region The frequency
`is given by
`the tonal
`conponents
`calculating
`
`3.75O Hz
`140.63
`HZ
`2l.25 Hz
`562.50
`Hz
`
`df
`df
`df
`
`in the frequency
`the frequency
`lfl
`in the frequency
`in the frequency
`
`3.0 kHz
`range up to
`6.0 kHz
`range up to
`range up to 12.0 kHz
`range up to 24.0 kHz
`
`To
`
`label
`
`operations
`
`lines xk that are tonal
`the spectral
`are performed
`
`the following
`
`local maxima
`Labelling of
`line xk is labelled as
`local maximum if
`spectral
`xk xkl
`xk xkl
`
`and
`
`Labelling of
`components
`tonal
`local maximum is labelled as
`xk xkj
`
`dB
`
`tonal
`
`component if
`
`for
`for
`for
`for
`
`63
`127
`255
`
`63
`127
`255
`500
`
`6.. 2..
`12.. 2.. 12
`If xk is labelled as
`expression are set
`
`to
`
`component
`tonal
`minimum level of
`
`the xkj of
`-160 dB
`
`the above
`
`Labelling of nontonal
`components
`The nontonal noise components are calculated from the remaining
`components from these
`lines To calculate the nontonal
`spectral
`lines xk the critical band rates bk are determined
`spectral
`BAND PATES The
`AND CRITICAL
`using the table FREQUENCIES
`lines within one critical band are replaced by one
`spectral
`the critical
`the geometric centre of
`line located at
`spectral
`the
`band The power of
`the sum of
`line equals
`this new spectral
`line is
`The newly created spectral
`the old ones
`powers of
`nontonal component
`labelled as
`
`3
`
`

`

`i-ic(cid:31)r
`
`Codwjrd
`
`Diferenti
`
`ie-tacur
`
`inr
`
`iir
`
`-53
`
`-62
`D1
`
`-56
`-55
`
`54
`-53
`-52
`
`-51
`-50
`
`49
`-48
`-47
`
`-46
`-45
`-44
`
`-43
`-42
`-41
`-40
`
`39
`-38
`-37
`
`-35
`-35
`-34
`-33
`-32
`
`-31
`-30
`-29
`-28
`
`27
`-26
`-25
`-24
`-23
`
`22
`-21
`-20
`-19
`-18
`-17
`-16
`-15
`-14
`-13
`-12
`-11
`-10
`
`-8
`
`-7
`
`-6
`
`-5
`-4
`-3
`-2
`-1
`
`iC1I0O10i000Iiii00
`101iC0l01000111101
`COCUOlOOOOUOOIC
`1000000100C000011
`10000001000000000
`10000001000000001
`t0000301000000110
`10000001000000111
`10000001000000100
`10000001000000101
`i01i00101001101010
`101100101001101011
`101100101001101000
`IO11CO1O100I1O1001
`1011001O1001101110
`101100101001101111
`101100101001101100
`101100101001101101
`101100101001100010
`101100101001100011
`101100101001100000
`101100101001100001
`101100101001100110
`101100101001100111
`101100101001100100
`101100101001100101
`101100101001111010
`101100101001111011
`101100101001111000
`101100101001111001
`101100101001111110
`101100101001111111
`101100101001111100
`101100101001111101
`101100101001110010
`101100101001110011
`101100101001110000
`101100101001110001
`10000001001101101
`1000000100110111
`1011001010000100
`1011001010001101
`10110010100011111
`1011001010001110
`1011001010001100
`10000001001100
`10110010100010
`1000000100111
`100000010001
`101100101010
`11100001000
`1011001011
`100000011
`10000000
`11100000
`1011000
`1110001
`101101
`10001
`11101
`1010
`1101
`1111
`
`110
`
`iOOi
`10111
`111001
`
`1000001
`10110011
`111000011
`101100100
`1110000101
`1000000101
`11100001001
`101100101011
`106000010010
`1000000100001
`10000001000001
`10110010100000
`1011001010000110
`1011001010000101
`10110010100001110
`iOilOOlOlOOlllOllO
`101100101001110111
`101100101001110100
`101100101001110101
`101100101001001010
`101100101001001011
`101100101001001000
`i01100101001001001
`101100101001001110
`101100101001001111
`101100101001001100
`101100101001001101
`101100101001000010
`101100101001000011
`101100101001000000
`101100101001000001
`101100101001000110
`101100101001000111
`101100101001000100
`101100101001000101
`101100101001011010
`101100101001011011
`101100101001011000
`101100101001011001
`101100101001011110
`i01100101001011111
`101100101001011100
`iOi100101001011101
`101100101001010010
`i01100101001010011
`iOllOOlOlOOlOl0000
`101100101001010001
`101100101001010110
`101100101001010111
`101100101001010100
`101100101001010101
`10000001001101010
`10000001001101011
`10000001001101000
`10000001001101001
`101100101000011110
`101100101000011111
`10000001001101100
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`20
`
`21
`22
`23
`24
`25
`
`26
`
`27
`28
`29
`30
`
`31
`32
`33
`34
`35
`36
`
`37
`38
`39
`40
`
`41
`42
`43
`44
`45
`46
`
`47
`48
`49
`
`50
`
`51
`52
`53
`54
`55
`
`56
`
`57
`58
`
`60
`61
`62
`
`63
`
`4
`
`