readsfs.m

this is speech recognition file

function [y,fs,hd,ffx]=readsfs(ff,ty,sub,mode,nmax,nskip,xpath)
%readsfs  Read a .SFS format sound file [Y,FS,HD,FFX]=(FF,TY,SUB,MODE,NMAX,NSKIP,XPATH)
%
% The SFS (Speech Filing System) is a package mostly written by Mark Huckvale
% and is available for UNIX and PC systems from http://www.phon.ucl.ac.uk/resource/sfs/
%
% Input Parameters:
%
%  FF gives the name of the file or alternatively
%                 can be the ffx output from a previous call to readsfs
%  TY gives the type of data item required e.g.:
%              0 Main header, 1 Speech data, 2 Laryngograph, 5 Annotation
%  SUB specifies the instance of type TY in the file: 0 for first (default), -1 for last or else
%      it can specify the start of the processing history field as a string (e.g. 'hqtx')
%  MODE		specifies the following (*=default):
%
%           File I/O: 'f'    Do not close file on exit
%             'd'    Look in data directory: voicebox('dir_data')
%           Int Format: 'i'  Force integer data to be at least 16 bits
%                              (some sfs files have a header error which falsely indicates 8-bit data)
%           Create item: 'c' Create item if necessary
%
%  NMAX     maximum number of samples to read (or -1 for unlimited [default])
%  NSKIP    number of samples or frames to skip from start of file
%               (or -1 to continue from previous read when FFX is given instead of a filename [default])
%  XPATH    (used with 'c' option) gives the full name of the program needed to generate the data or
%           the directory containing it.
%
% Output Parameters:
%
%  Y        data matrix or cell matrix whose format depends on TY:
%        TY=0: empty
%			TY=5: cell array {nf,3} = {position length annotation}
%        TY=1,2: column vector containing data
%        TY=11: data array with one row per frame
%  FS       sample frequency in Hz
%  HD     cell matrix whose format depends on TY:
%        TY=0: cell{14,1}
%              {1} row vector
%                {1}(1) = serial_date (see DATENUM() for format)
%                {1}(2) = file_number
%                {1}(3) = machine_type
%              {2} = File type (= 'UC2')
%              {3} = username of creator
%              {4} = site of creator
%              {5} = source
%              {6} = database
%              {7} = speaker name
%              {8} = session code
%              {9} = session date (as a string)
%             {10} = name of token
%             {11} = token repetition code
%             {12} = recording conditions
%             {13} = archiving details
%             {14} = general comments
%        TY>0: cell{4,1}
%              {1} = (1,14) array:
%                 {1}(1)  = processdate (see DATENUM() for format)
%                 {1}(2)  = datatype: 1=speech, 2=lx, 3=tx cycle lengths, 4=fx freq
%                                     5=annotations, 6=phonetic, 7=synthesiser, 8=words
%                                     9=grey-level, 10=voicing, 11=energy, 12=formants
%                                     13=energy, 14=lpc, 15=markov, 16=acoustic, 17=?,
%                                     18=geometry, 19=aerodynamics, 20=articulatory
%                                     21=source, 22=physiological, 23=rational filter
%                                     24=poles/zeros, 25=glottal flow, 26=excitation model
%                                     27=nose, 28=calibration, 29=area
%                 {1}(3)  = subtype
%                 {1}(4)  = floating: 1=float, 0=int, -1=structure+
%                 {1}(5)  = datasize in bytes
%                 {1}(6)  = framesize in units of datasize
%                 {1}(7)  = numframes
%                 {1}(8)  = length in bytes of data
%                 {1}(9)  = frameduration=1/sample_rate
%                 {1}(10) = datapresent: 0=deleted, 1=present, 2=link
%                 {1}(11) = time offset
%                 {1}(12) = windowsize
%                 {1}(13) = overlap
%                 {1}(14) = lxsync
%              {2} = processing history
%              {3} = parameter field
%              {4} = comment
%
%  FFX     cell array containing:
%              {1} = filename
%              {2} = (1,4) = [fid byte_order item_row values_read]
%              {3} = (nitem,5) = one row per item [type subtype length position byteorder]
%              {4} = {nitem,3} = cell: one row per item {processing parameters comment} text strings
%

% Features yet to be implemented:
%
%   (1) If no output parameters are specified, header information will be printed.
%   (2) following link items
%   (3) MODE options:
%                Scaling: 's'    Auto scale to make data peak = +-1
%                         'r'    Raw unscaled data (integer values)
%                         'q'    Scaled to make 0dBm0 be unity mean square
%                         'p' *	Scaled to make +-1 equal full scale
%                Errors   'r'    Return if file is non-existant


%	   Copyright (C) Mike Brookes 1998
%      Version: $Id: readsfs.m,v 1.4 2007/05/04 07:01:39 dmb Exp $
%
%   VOICEBOX is a MATLAB toolbox for speech processing.
%   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
%   http://www.gnu.org/copyleft/gpl.html or by writing to
%   Free Software Foundation, Inc.,675 Mass Ave, Cambridge, MA 02139, USA.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

if nargin<7
    xpath=voicebox('sfsbin');       % path for sfs programs
end
EXESUF=voicebox('sfssuffix');                                              % suffix for executable O/S dependent
if nargin<4 mode='p';
else mode = [mode(:).' 'p'];
end
if nargout==0
    if nargin<2 | ty<=0
        [yy,fs,hd,ffx]=readsfs(ff,0,0,mode);
        fprintf(1,'File: %s\n',ffx{1});
        fprintf(1,'Database: %s, Speaker: %s, Date: %s, Token: %s\n',hd{6},hd{7},hd{9},hd{10});
        lst=ffx{3};
        for it=2:size(lst,1);
            [yy,fs,hd,ffx]=readsfs(ffx,lst(it,1),lst(it,2),mode,0);
            nf=hd{1}(7);
            fd=hd{1}(9);
            fprintf(1,'%3d.%02d %ss @ %sHz (%d frames): %s\n',lst(it,1),lst(it,2),sprintsi(nf*fd),sprintsi(1/fd),nf,ffx{4}{it,1});   
        end
    end
else
    it=[];
    xfid=[];                  % xfid will be non-empty second time around
    while (isempty(it))                 % may go round this loop twice
        if ischar(ff)           % If ff is a string we must read file
            if any(mode=='d')
                ff=fullfile(voicebox('dir_data'),ff);
            end
            fid=fopen(ff,'rb','b');
            if fid == -1 error(sprintf('Can''t open %s for input',ff)); end
            
            t=fread(fid,512,'uchar').';
            if (t(1:3)~='UC2')
                error(sprintf('%s is not an SFS file type UC2',ff))
            end
            itemlist = [0 1 0 0 t(512)];
            
            proglist={};
            for i=2:200
                pos = ftell(fid);
                [t,n]=fread(fid,512,'uchar');
                if (n < 512) break; end
                mm=pow2(1,8*([0 1 2 3]+(t(512)==0)*[3 1 -1 -3]));
                itemlist(i,:)=[mm*[t(389:392) t(393:396) t(413:416)] pos t(512)];
                if itemlist(i,1)>29
                    error(sprintf('%d is not a valid SFS item type',itemlist(i,1)))
                end
                proglist{i,1}=char(zerotrim(t(1:256)'));
                proglist{i,2}=char(zerotrim(t(257:384)'));
                proglist{i,3}=char(zerotrim(t(437:456)'));
                fseek(fid,itemlist(i,3),'cof');
            end
            ffx={ff; [fid 0 0 0]; itemlist; proglist};
        else
            ffx=ff;
            ff=ffx{1};
            fid=ffx{2}(1);
            if fid<0 
                fid=fopen(ffx{1},'rb',char('b'+(ffx{2}(2)~=0)*('l'-'b')));
            end
        end
        
        % now try to find the requested item