comparison man/man1/genasa.1 @ 0:5242703e91d3 tip

Initial checkin for AIM92 aimR8.2 (last updated May 1997).
author tomwalters
date Fri, 20 May 2011 15:19:45 +0100
parents
children
comparison
equal deleted inserted replaced
-1:000000000000 0:5242703e91d3
1 .TH GENASA 1 "11 May 1995"
2 .LP
3 .SH NAME
4 .LP
5 genasa \- generate auditory spectral analysis
6 .LP
7 .SH SYNOPSIS
8 .LP
9 genasa [ option=value | -option ] [ filename ]
10 .LP
11 .LP
12 .SH DESCRIPTION
13 .LP
14 The genasa module of the AIM software performs a time-domain spectral
15 analysis on the input wave using a bank of auditory filters, and
16 summarises the information in a sequence of auditory spectra. The
17 spectral analysis converts the input wave into an array of filtered
18 waves, one for each channel of a gammatone auditory filterbank. The
19 surface of the array of filtered waves is AIM's representation of
20 basilar membrane motion (BMM) as a function of time. The sequence of
21 auditory spectra is produced by calculating the envelope of the BMM
22 and extracting spectral slices from the envelope every 'frstep_epn'
23 ms. The envelope is calculated continuously, by rectifing,
24 compressing, and lowpass filtering the individual BMM waves as they
25 flow from the filterbank.
26 .LP
27 The auditory spectrum produced by genasa is intended to simulate the
28 spectral representation of a sound as it occurs in the peripheral
29 auditory system just prior to neural transduction. As a result, the
30 frequency resolution of the analysis varies with the center frequency
31 of the channel, and the distribution of channels across frequency is
32 chosen to match that in the auditory system. The auditory spectrum is
33 a plot of the activity in each channel as a function of the centre
34 frequency of the auditory filter (in ERB's). The representation is
35 referred to as an auditory spectrum to distinguish it from the Fourier
36 energy spectrum (Patterson, 1994a). The suffix 'asa' is short for
37 'auditory spectral analysis'; it is used to distinguish this spectral
38 representation from three other spectral representations provided by
39 the AIM software ('epn' excitation pattern, 'sgm' auditory
40 spectrogram, and 'cgm' cochleogram).
41 .LP
42 The spectral analysis performed by genasa is the same as that
43 performed by genbmm. The primary differences are in the display
44 defaults and the inclusion of the Compression and Leaky Integration
45 modules used to construct the spectral slices from the BMM. As a
46 result, this manual entry is restricted to describing the option
47 values that differ from those in genbmm and the additional options
48 required to control the Compression and Leaky Integration.
49 .LP
50 .SH DISPLAY DEFAULTS
51 .LP
52 The default values for three of the display options are reset to
53 produce a spectral format rather than a landscape; specifically,
54 display=excitation, bottom=0 and top=2500. The number of channels is
55 increased to 128 to ensure reasonable frequency resolution in the
56 spectral display.
57 .LP
58 .SH COMPRESSION AND LEAKY INTEGRATION
59 .LP
60 Compression and lowpass filtering are activated and the neural
61 encoding stage that comes between them is turned off:
62 .LP
63 .LP
64 .SS "Compression"
65 .PP
66 Auditory spectra are usually produced via the functional route in
67 AIM. In this case, compress is set on
68 .LP
69 .TP 13
70 compress
71 Logarithmic compressor switch
72 .RS
73 Switch. Default: on.
74 .RE
75 .RS
76 .LP
77 Note: The compressor in the functional route of AIM is logarithmic and
78 it screens out negative BMM values before compression. This rectifies
79 the wave during the compression process and so the separate rectify
80 option is left off.
81 .RE
82 .LP
83 .RS
84 .LP
85 Note: The compressor in the physiological route of AIM is an integral
86 part of the tlf module, so when using this route to produce auditory
87 spectra, turn off the logarithmic compressor (i.e. compress=off). The
88 compressor in tlf does not screen out negative values so it is also
89 important to set rectify=on.
90 .RE
91 .RS
92 .LP
93 Full wave rectification is produced if rectify is set to 2. This is
94 useful when calculating envelopes with genasa.
95 .RE
96 .LP
97 .LP
98 .SS "Transduction"
99 .PP
100 .LP
101 .TP 13
102 transduction
103 Neural transduction switch (at, meddis, off)
104 .RS
105 Switch. Default: off.
106 .RE
107 .LP
108 .LP
109 .SS "Leaky Integration"
110 .PP
111 .LP
112 .TP 13
113 stages_idt
114 Number of stages of lowpass filtering
115 .RS
116 Default unit: scalar. Default value: 2
117 .RE
118 .TP 13
119 tup_idt
120 The time constant for each filter stage
121 .RS
122 Default unit: ms. Default value: 8 ms.
123 .RE
124 .LP
125 The Equivalent Rectandular Duration (ERD) of a two stage lowpass
126 filter is about 1.6 times the time constant of each stage, or
127 12.8 ms in the current case.
128 .TP 13
129 frstep_epn
130 The time between successive spectral frames
131 .RS
132 Default unit: ms. Default value: 10 ms.
133 .RE
134 .LP
135 With a frstep_epn of 10 ms, genasa will produce
136 spectral frames at a rate of 100 per second.
137 .LP
138 .TP 13
139 downsample
140 The time between successive spectral frames.
141 .RS
142 Default unit: ms. Default value: 10 ms.
143 .RE
144 .LP
145 Downsample is simply another name for frstep_epn, provided to
146 facilitate a different mode of thinking about time-series data.
147 .RE
148 .LP
149 .SH FILES
150 .LP
151 .TP 13
152 .genasarc
153 The options file for genasa.
154 .LP
155 .SH SEE ALSO
156 .LP
157 genbmm, gensgm
158 .LP
159 .SH BUGS
160 .LP
161 None currently known.
162 .SH COPYRIGHT
163 .LP
164 Copyright (c) Applied Psychology Unit, Medical Research Council, 1995
165 .LP
166 Permission to use, copy, modify, and distribute this software without fee
167 is hereby granted for research purposes, provided that this copyright
168 notice appears in all copies and in all supporting documentation, and that
169 the software is not redistributed for any fee (except for a nominal
170 shipping charge). Anyone wanting to incorporate all or part of this
171 software in a commercial product must obtain a license from the Medical
172 Research Council.
173 .LP
174 The MRC makes no representations about the suitability of this
175 software for any purpose. It is provided "as is" without express or
176 implied warranty.
177 .LP
178 THE MRC DISCLAIMS ALL WARRANTIES WITH REGARD TO THIS SOFTWARE, INCLUDING
179 ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS, IN NO EVENT SHALL
180 THE A.P.U. BE LIABLE FOR ANY SPECIAL, INDIRECT OR CONSEQUENTIAL DAMAGES
181 OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS,
182 WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION,
183 ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS
184 SOFTWARE.
185 .LP
186 .SH ACKNOWLEDGEMENTS
187 .LP
188 The AIM software was developed for Unix workstations by John
189 Holdsworth and Mike Allerhand of the MRC APU, under the direction of
190 Roy Patterson. The physiological version of AIM was developed by
191 Christian Giguere. The options handler is by Paul Manson. The revised
192 SAI module is by Jay Datta. Michael Akeroyd extended the postscript
193 facilites and developed the xreview routine for auditory image
194 cartoons.
195 .LP
196 The project was supported by the MRC and grants from the U.K. Defense
197 Research Agency, Farnborough (Research Contract 2239); the EEC Esprit
198 BR Porgramme, Project ACTS (3207); and the U.K. Hearing Research Trust.