-
Notifications
You must be signed in to change notification settings - Fork 5
/
activlev.m
179 lines (176 loc) · 6.03 KB
/
activlev.m
1
2
3
4
5
6
7
8
9
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
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
function [lev,fso]=activlev(sp,fs,mode)
%ACTIVLEV Measure active speech level as in ITU-T P.56 [LEV,FSO]=(SP,FS,MODE)
%
%Inputs: SP is the speech signal (with better than 20dB SNR)
% FS is the sample frequency in Hz (see also FSO below)
% MODE is any combination of the following:
% r - raw: do not apply the input filter 0.2 to 5.5 kHz
% a - all: use all the samples rather than subsampling to 694Hz
% d - give outputs in dB rather than power
% l - give the "long term level" as well as the active speech level
%Outputs:
% LEV gives the speech level in units of power (or dB if mode='d')
% if mode='l' is specified, LEV is a row vector with the "long term
% level" as its second element (this is just the mean power)
% FSO is a column vector of intermediate information that allows
% you to process a speech signal in chunks. Thus:
%
% fso=fs; for i=1:inc:nsamp, [lev,fso]=activlev(sp(i:i+inc-1),fso,mode); end
%
% is equivalent to: lev=activlev(sp(1:nsamp),fs,mode)
%
% but is much slower. The two methods will not give identical results
% because they will use slightly different threshods.
%For completeness we list here the contents of the FSO array:
%
% 1 : sample frequency
% 2 : subsampling increment
% 3 : hangover increment
% 4-6 : smoothing filter coefs
% 7-12 : 200Hz HP filter numerator
% 13-18 : 200Hz HP filter denominator
% 19-24 : 5.5kHz LP filter numerator
% 25-30 : 5.5kHz LP filter denominator
% 31-33 : count,sum and sum of squares of filter speech
% 34 : upper limit of largest amplitude bin
% 35 : samples to skip for subsampling
% 36-37 : smoothing filter state
% 38-42 : 200Hz HP filter state
% 43-47 : 5.5kHz LP filter state
% 48-55 : thresholded sample counts
% 56-63 : hangover counts
% Copyright (C) Mike Brookes 1998
%
% Last modified Thu Apr 30 17:28:29 1998
%
% VOICEBOX home page: http://www.ee.ic.ac.uk/hp/staff/dmb/voicebox/voicebox.html
%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% This program is free software; you can redistribute it and/or modify
% it under the terms of the GNU General Public License as published by
% the Free Software Foundation; either version 2 of the License, or
% (at your option) any later version.
%
% This program is distributed in the hope that it will be useful,
% but WITHOUT ANY WARRANTY; without even the implied warranty of
% MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
% GNU General Public License for more details.
%
% You can obtain a copy of the GNU General Public License from
% ftp://prep.ai.mit.edu/pub/gnu/COPYING-2.0 or by writing to
% Free Software Foundation, Inc.,675 Mass Ave, Cambridge, MA 02139, USA.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
if nargin<3 mode='0'; else mode = [mode '0']; end
lev=zeros(1,1+any(mode=='l'));
thf=0.5;
nb=8;
fso=[fs;zeros(64-1-length(fs),1)];
if length(fs)==1
ni=max(floor(fs*all(mode~='a')/694),1);
ti=ni/fs;
nh=ceil(0.2/ti)+1;
g=exp(-ti/0.03);
% s-plane zeros and poles of high pass 5'th order chebychev2 filter with -0.25dB at w=1
szp=[0.37843443673309i 0.23388534441447i; -0.20640255179496+0.73942185906851i -0.54036889596392+0.45698784092898i];
szp=[[0; -0.66793268833792] szp conj(szp)];
zl=2./(1-szp*tan(200*pi/fs))-1;
al=real(poly(zl(2,:)));
bl=real(poly(zl(1,:)));
sw=1-2*rem(0:5,2).';
bl=bl*(al*sw)/(bl*sw);
fso(2:19-1)=[ni;nh;[1; -2*g; g^2]/(1-2*g+g^2);bl(:);al(:)];
if fs>14000
zh=2./(szp/tan(5500*pi/fs)-1)+1;
ah=real(poly(zh(2,:)));
bh=real(poly(zh(1,:)));
bh=bh*sum(ah)/sum(bh);
fso(19:19+11)=[bh(:);ah(:)];
end
end
% need to calculate offset correctly
% subsample the signal
ns=length(sp);
if ns
% input filter goes here
if all(mode~='r')
[sp,fso(38:42)]=filter(fso(7:12),fso(13:18),sp(:),fso(38:42));
if fso(25)
[sp,fso(43:47)]=filter(fso(19:24),fso(25:30),sp,fso(43:47));
end
end
sp=abs(sp(1+fso(35):fso(2):ns));
na=length(sp);
fso(35)=fso(35)+na*fso(2)-ns;
if (na)
fso(31:33)=fso(31:33)+[na;sum(sp);sp.'*sp];
[sp,fso(36:37)]=filter(1,fso(4:6),sp,fso(36:37));
% determine new peak
th=max(sp);
s=zeros(nb,1);
h=s;
if fso(34)>0
if th>fso(34)
nt=floor(log(th/fso(34))/log(thf));
th=fso(34)*thf^nt;
if nt>-nb
s(1-nt:nb)=fso(48:47+nb+nt);
h(1-nt:nb)=fso(56:55+nb+nt);
end
else
th=fso(34);
s=fso(48:55);
h=fso(56:63);
end
end
fso(34)=th;
nh=fso(3);
if nh<na
nm=na-nh+2;
for i=1:nb
th=th*thf;
b=sp>th;
np=h(i);
s(i)=s(i)+np+sum(cumsum([sum(b(1:np+1));b(2+np:na)]-[zeros(nh-np,1);b(1:na-nh)])>0);
ff=find([1;b(nm:na)]);
nm=nm-1+ff(end);
h(i)=nm+nh-na-2;;
end
else
for i=1:nb
th=th*thf;
b=sp>th;
np=h(i);
ff=find(b);
if length(ff)
h(i)=ff(end)+nh-na-1;
s(i)=s(i)+na+min(0,np+1-ff(1));
else
h(i)=max(0,np-na);
s(i)=s(i)-h(i)+np;
end
end
end
fso(48:63)=[s;h];
end
end
ssq=fso(33);
if ssq>0
s=fso(48:55);
thresh=(fso(34)*thf.^(1:nb).').^2;
% 15.9 dB = 38.9 power ratio
mar=38.9;
st=s.*thresh*mar/ssq;
ff=find(st>1);
if length(ff)
nf=ff(end);
ls=log(s(nf)/s(min(nb,nf+1)));
lev(1)=ssq/s(nf)*st(nf)^(ls/(ls-2*log(thf)));
else
[ls,nf]=max(st);
lev(1)=ssq/s(nf);
end
if length(lev)>1 lev(2)=ssq/fso(31); end
end
if any(mode=='d')
lev=10*log10(max(lev,1e-200));
end