READSPH Read a SPHERE/TIMIT format sound file [Y,FS,WRD,PHN,FFX]=(FILENAME,MODE,NMAX,NSKIP)
Input Parameters:
FILENAME gives the name of the file (with optional .SPH extension) or alternatively
can be the FFX output from a previous call to READSPH having the 'f' mode option
MODE specifies the following (*=default):
Scaling: 's' Auto scale to make data peak = +-1 (use with caution if reading in chunks)
'r' Raw unscaled data (integer values)
'p' * Scaled to make +-1 equal full scale
'o' Scale to bin centre rather than bin edge (e.g. 127 rather than 127.5 for 8 bit values)
(can be combined with n+p,r,s modes)
'n' Scale to negative peak rather than positive peak (e.g. 128.5 rather than 127.5 for 8 bit values)
(can be combined with o+p,r,s modes)
Format 'l' Little endian data (Intel,DEC) (overrides indication in file)
'b' Big endian data (non Intel/DEC) (overrides indication in file)
File I/O: 'f' Do not close file on exit
'd' Look in data directory: voicebox('dir_data')
'w' Also read the annotation file *.wrd if present (as in TIMIT)
't' Also read the phonetic transcription file *.phn if present (as in TIMIT)
Eac line of the annotation and transcription files is of the form: m n token
where m and n are start end end times in samples and token is a word or phoneme test descriptor
The corresponding cell arrays WRD and PHN contain two elements per row: {[m n]/fs 'token'}
These outputs are only present if the corresponding 'w' and 't' options are selected
NMAX maximum number of samples to read (or -1 for unlimited [default])
NSKIP number of samples to skip from start of file
(or -1 to continue from previous read when FFX is given instead of FILENAME [default])
Output Parameters:
Y data matrix of dimension (samples,channels)
FS sample frequency in Hz
WRD{*,2} cell array with word annotations: WRD{*,:)={[t_start t_end],'text'} where times are in seconds
only present if 'w' option is given
PHN{*,2} cell array with phoneme annotations: PHN{*,:)={[t_start t_end],'phoneme'} where times are in seconds
only present if 't' option is present
FFX Cell array containing
{1} filename
{2} header information
{1} first header field name
{2} first header field value
{3} format string (e.g. NIST_1A)
{4}(1) file id
(2) current position in file
(3) dataoff byte offset in file to start of data
(4) order byte order (l or b)
(5) nsamp number of samples
(6) number of channels
(7) nbytes bytes per data value
(8) bits number of bits of precision
(9) fs sample frequency
(10) min value
(11) max value
(12) coding: 0=PCM,1=uLAW + 0=no compression,10=shorten,20=wavpack,30=shortpack
(13) file not yet decompressed
{5} temporary filename
If no output parameters are specified, header information will be printed.
To decode shorten-encoded files, the program shorten.exe must be in the same directory as this m-file
Usage Examples:
(a) Draw an annotated spectrogram of a TIMIT file
filename='....TIMIT/TEST/DR1/FAKS0/SA1.WAV';
[s,fs,wrd,phn]=readsph(filename,'wt');
spgrambw(s,fs,'Jwcpta',[],[],[],[],wrd);0001 function [y,fs,wrd,phn,ffx]=readsph(filename,mode,nmax,nskip) 0002 %READSPH Read a SPHERE/TIMIT format sound file [Y,FS,WRD,PHN,FFX]=(FILENAME,MODE,NMAX,NSKIP) 0003 % 0004 % Input Parameters: 0005 % 0006 % FILENAME gives the name of the file (with optional .SPH extension) or alternatively 0007 % can be the FFX output from a previous call to READSPH having the 'f' mode option 0008 % MODE specifies the following (*=default): 0009 % 0010 % Scaling: 's' Auto scale to make data peak = +-1 (use with caution if reading in chunks) 0011 % 'r' Raw unscaled data (integer values) 0012 % 'p' * Scaled to make +-1 equal full scale 0013 % 'o' Scale to bin centre rather than bin edge (e.g. 127 rather than 127.5 for 8 bit values) 0014 % (can be combined with n+p,r,s modes) 0015 % 'n' Scale to negative peak rather than positive peak (e.g. 128.5 rather than 127.5 for 8 bit values) 0016 % (can be combined with o+p,r,s modes) 0017 % Format 'l' Little endian data (Intel,DEC) (overrides indication in file) 0018 % 'b' Big endian data (non Intel/DEC) (overrides indication in file) 0019 % 0020 % File I/O: 'f' Do not close file on exit 0021 % 'd' Look in data directory: voicebox('dir_data') 0022 % 'w' Also read the annotation file *.wrd if present (as in TIMIT) 0023 % 't' Also read the phonetic transcription file *.phn if present (as in TIMIT) 0024 % Eac line of the annotation and transcription files is of the form: m n token 0025 % where m and n are start end end times in samples and token is a word or phoneme test descriptor 0026 % The corresponding cell arrays WRD and PHN contain two elements per row: {[m n]/fs 'token'} 0027 % These outputs are only present if the corresponding 'w' and 't' options are selected 0028 % 0029 % NMAX maximum number of samples to read (or -1 for unlimited [default]) 0030 % NSKIP number of samples to skip from start of file 0031 % (or -1 to continue from previous read when FFX is given instead of FILENAME [default]) 0032 % 0033 % Output Parameters: 0034 % 0035 % Y data matrix of dimension (samples,channels) 0036 % FS sample frequency in Hz 0037 % WRD{*,2} cell array with word annotations: WRD{*,:)={[t_start t_end],'text'} where times are in seconds 0038 % only present if 'w' option is given 0039 % PHN{*,2} cell array with phoneme annotations: PHN{*,:)={[t_start t_end],'phoneme'} where times are in seconds 0040 % only present if 't' option is present 0041 % FFX Cell array containing 0042 % 0043 % {1} filename 0044 % {2} header information 0045 % {1} first header field name 0046 % {2} first header field value 0047 % {3} format string (e.g. NIST_1A) 0048 % {4}(1) file id 0049 % (2) current position in file 0050 % (3) dataoff byte offset in file to start of data 0051 % (4) order byte order (l or b) 0052 % (5) nsamp number of samples 0053 % (6) number of channels 0054 % (7) nbytes bytes per data value 0055 % (8) bits number of bits of precision 0056 % (9) fs sample frequency 0057 % (10) min value 0058 % (11) max value 0059 % (12) coding: 0=PCM,1=uLAW + 0=no compression,10=shorten,20=wavpack,30=shortpack 0060 % (13) file not yet decompressed 0061 % {5} temporary filename 0062 % 0063 % If no output parameters are specified, header information will be printed. 0064 % To decode shorten-encoded files, the program shorten.exe must be in the same directory as this m-file 0065 % 0066 % Usage Examples: 0067 % 0068 % (a) Draw an annotated spectrogram of a TIMIT file 0069 % filename='....TIMIT/TEST/DR1/FAKS0/SA1.WAV'; 0070 % [s,fs,wrd,phn]=readsph(filename,'wt'); 0071 % spgrambw(s,fs,'Jwcpta',[],[],[],[],wrd); 0072 0073 % Copyright (C) Mike Brookes 1998 0074 % Version: $Id: readsph.m 713 2011-10-16 14:45:43Z dmb $ 0075 % 0076 % VOICEBOX is a MATLAB toolbox for speech processing. 0077 % Home page: http://www.ee.ic.ac.uk/hp/staff/dmb/voicebox/voicebox.html 0078 % 0079 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% 0080 % This program is free software; you can redistribute it and/or modify 0081 % it under the terms of the GNU General Public License as published by 0082 % the Free Software Foundation; either version 2 of the License, or 0083 % (at your option) any later version. 0084 % 0085 % This program is distributed in the hope that it will be useful, 0086 % but WITHOUT ANY WARRANTY; without even the implied warranty of 0087 % MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the 0088 % GNU General Public License for more details. 0089 % 0090 % You can obtain a copy of the GNU General Public License from 0091 % http://www.gnu.org/copyleft/gpl.html or by writing to 0092 % Free Software Foundation, Inc.,675 Mass Ave, Cambridge, MA 02139, USA. 0093 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% 0094 0095 persistent BYTEORDER 0096 codes={'sample_count'; 'channel_count'; 'sample_n_bytes';'sample_sig_bits'; 'sample_rate'; 'sample_min'; 'sample_max'}; 0097 codings={'pcm'; 'ulaw'}; 0098 compressions={',embedded-shorten-';',embedded-wavpack-'; ',embedded-shortpack-'}; 0099 if isempty(BYTEORDER), BYTEORDER='l'; end 0100 if nargin<1, error('Usage: [y,fs,hdr,fidx]=READSPH(filename,mode,nmax,nskip)'); end 0101 if nargin<2, mode='p'; 0102 else mode = [mode(:).' 'p']; 0103 end 0104 k=find((mode>='p') & (mode<='s')); 0105 mno=all(mode~='o'); % scale to input limits not output limits 0106 sc=mode(k(1)); 0107 if any(mode=='l'), BYTEORDER='l'; 0108 elseif any(mode=='b'), BYTEORDER='b'; 0109 end 0110 if nargout 0111 ffx=cell(5,1); 0112 if ischar(filename) 0113 if any(mode=='d') 0114 filename=fullfile(voicebox('dir_data'),filename); 0115 end 0116 fid=fopen(filename,'rb',BYTEORDER); 0117 if fid == -1 0118 fn=[filename,'.sph']; 0119 fid=fopen(fn,'rb',BYTEORDER); 0120 if fid ~= -1, filename=fn; end 0121 end 0122 if fid == -1 0123 error('Can''t open %s for input',filename); 0124 end 0125 ffx{1}=filename; 0126 else 0127 if iscell(filename) 0128 ffx=filename; 0129 else 0130 fid=filename; 0131 end 0132 end 0133 0134 if isempty(ffx{4}); 0135 fseek(fid,0,-1); 0136 str=char(fread(fid,16)'); 0137 if str(8) ~= 10 || str(16) ~= 10, fclose(fid); error('File does not begin with a SPHERE header'); end 0138 ffx{3}=str(1:7); 0139 hlen=str2double(str(9:15)); 0140 hdr={}; 0141 while 1 0142 str=fgetl(fid); 0143 if str(1) ~= ';' 0144 [tok,str]=strtok(str); 0145 if strcmp(tok,'end_head'), break; end 0146 hdr(end+1,1)={tok}; 0147 [tok,str]=strtok(str); 0148 if tok(1) ~= '-', error('Missing ''-'' in SPHERE header'); end 0149 if tok(2)=='s' 0150 hdr(end,2)={str(2:str2num(tok(3:end))+1)}; 0151 elseif tok(2)=='i' 0152 hdr(end,2)={sscanf(str,'%d',1)}; 0153 else 0154 hdr(end,2)={sscanf(str,'%f',1)}; 0155 end 0156 end 0157 end 0158 i=find(strcmp(hdr(:,1),'sample_byte_format')); 0159 if ~isempty(i) 0160 bord=char('b'+('l'-'b')*(hdr{i,2}(1)=='0')); 0161 if bord ~= BYTEORDER && mode~='b' && mode ~='l' 0162 BYTEORDER=bord; 0163 fclose(fid); 0164 fid=fopen(filename,'rb',BYTEORDER); 0165 end 0166 end 0167 i=find(strcmp(hdr(:,1),'sample_coding')); 0168 icode=0; % initialize to PCM coding 0169 if ~isempty(i) 0170 icode=-1; % unknown code 0171 scode=hdr{i,2}; 0172 nscode=length(scode); 0173 for j=1:length(codings) 0174 lenj=length(codings{j}); 0175 if strcmp(scode(1:min(nscode,lenj)),codings{j}) 0176 if nscode>lenj 0177 for k=1:length(compressions) 0178 lenk=length(compressions{k}); 0179 if strcmp(scode(lenj+1:min(lenj+lenk,nscode)),compressions{k}) 0180 icode=10*k+j-1; 0181 break; 0182 end 0183 end 0184 else 0185 icode=j-1; 0186 end 0187 break; 0188 end 0189 end 0190 end 0191 0192 info=[fid; 0; hlen; double(BYTEORDER); 0; 1; 2; 16; 1 ; 1; -1; icode]; 0193 for j=1:7 0194 i=find(strcmp(hdr(:,1),codes{j})); 0195 if ~isempty(i) 0196 info(j+4)=hdr{i,2}; 0197 end 0198 end 0199 if ~info(5) 0200 fseek(fid,0,1); 0201 info(5)=floor((ftell(fid)-info(3))/(info(6)*info(7))); 0202 end 0203 ffx{2}=hdr; 0204 ffx{4}=info; 0205 end 0206 info=ffx{4}; 0207 if nargin<4, nskip=info(2); 0208 elseif nskip<0, nskip=info(2); 0209 end 0210 0211 ksamples=info(5)-nskip; 0212 if nargin>2 0213 if nmax>=0 0214 ksamples=min(nmax,ksamples); 0215 end 0216 end 0217 0218 if ksamples>0 0219 fid=info(1); 0220 if icode>=10 && isempty(ffx{5}) %#ok<AND2> 0221 fclose(fid); 0222 dirt=voicebox('dir_temp'); 0223 [fnp,fnn,fne,fnv]=fileparts(filename); 0224 filetemp=fullfile(dirt,[fnn fne fnv]); 0225 cmdtemp=fullfile(dirt,'shorten.bat'); % batch file needed to convert to short filenames 0226 % if ~exist(cmdtemp,'file') % write out the batch file if it doesn't exist 0227 cmdfid=fopen(cmdtemp,'wt'); 0228 fprintf(cmdfid,'@"%s" -x -a %%1 "%%~s2" "%%~s3"\n',voicebox('shorten')); 0229 fclose(cmdfid); 0230 % end 0231 if exist(filetemp,'file') % need to explicitly delete old file since shorten makes read-only 0232 doscom=['del /f "' filetemp '"']; 0233 if dos(doscom) % run the program 0234 error('Error running DOS command: %s',doscom); 0235 end 0236 end 0237 if floor(icode/10)==1 % shorten 0238 doscom=['"' cmdtemp '" ' num2str(info(3)) ' "' filename '" "' filetemp '"']; 0239 % fprintf(1,'Executing: %s\n',doscom); 0240 if dos(doscom) % run the program 0241 error('Error running DOS command: %s',doscom); 0242 end 0243 else 0244 error('unknown compression format'); 0245 end 0246 ffx{5}=filetemp; 0247 fid=fopen(filetemp,'r',BYTEORDER); 0248 if fid<0, error('Cannot open decompressed file %s',filetemp); end 0249 info(1)=fid; % update fid 0250 end 0251 info(2)=nskip+ksamples; 0252 pk=pow2(0.5,8*info(7))*(1+(mno/2-all(mode~='n'))/pow2(0.5,info(8))); % use modes o and n to determine effective peak 0253 fseek(fid,info(3)+info(6)*info(7)*nskip,-1); 0254 nsamples=info(6)*ksamples; 0255 if info(7)<3 0256 if info(7)<2 0257 y=fread(fid,nsamples,'uchar'); 0258 if info(12)==1 0259 y=pcmu2lin(y); 0260 pk=2.005649; 0261 else 0262 y=y-128; 0263 end 0264 else 0265 y=fread(fid,nsamples,'short'); 0266 end 0267 else 0268 if info(7)<4 0269 y=fread(fid,3*nsamples,'uchar'); 0270 y=reshape(y,3,nsamples); 0271 y=[1 256 65536]*y-pow2(fix(pow2(y(3,:),-7)),24); 0272 else 0273 y=fread(fid,nsamples,'long'); 0274 end 0275 end 0276 if sc ~= 'r' 0277 if sc=='s' 0278 if info(10)>info(11) 0279 info(10)=min(y); 0280 info(11)=max(y); 0281 end 0282 sf=1/max(max(abs(info(10:11))),1); 0283 else sf=1/pk; 0284 end 0285 y=sf*y; 0286 end 0287 if info(6)>1, y = reshape(y,info(6),ksamples).'; end 0288 else 0289 y=[]; 0290 end 0291 0292 if mode~='f' 0293 fclose(fid); 0294 info(1)=-1; 0295 if ~isempty(ffx{5}) 0296 doscom=['del /f ' ffx{5}]; 0297 if dos(doscom) % run the program 0298 error('Error running DOS command: %s',doscom); 0299 end 0300 ffx{5}=[]; 0301 end 0302 end 0303 ffx{4}=info; 0304 fs=info(9); 0305 wrd=ffx; % copy ffx into the other arguments in case 'w' and/or 't' are not specified 0306 phn=ffx; 0307 if any(mode=='w') 0308 wrd=cell(0,0); 0309 fidw=fopen([filename(1:end-3) 'wrd'],'r'); 0310 if fidw>0 0311 while 1 0312 tline = fgetl(fidw); % read an input line 0313 if ~ischar(tline) 0314 break 0315 end 0316 [wtim, ntim, ee, nix] = sscanf(tline,'%d%d',2); 0317 if ntim==2 0318 wrd{end+1,1}=wtim(:)'/fs; 0319 wrd{end,2}=strtrim(tline(nix:end)); 0320 end 0321 end 0322 fclose(fidw); 0323 end 0324 end 0325 if any(mode=='t') 0326 ph=cell(0,0); 0327 fidw=fopen([filename(1:end-3) 'phn'],'r'); 0328 if fidw>0 0329 while 1 0330 tline = fgetl(fidw); % read an input line 0331 if ~ischar(tline) 0332 break 0333 end 0334 [wtim, ntim, ee, nix] = sscanf(tline,'%d%d',2); 0335 if ntim==2 0336 ph{end+1,1}=wtim(:)'/fs; 0337 ph{end,2}=strtrim(tline(nix:end)); 0338 end 0339 end 0340 fclose(fidw); 0341 end 0342 if any(mode=='w') 0343 phn=ph; % copy into 4th argument 0344 else 0345 wrd=ph; % copy into 3rd argument 0346 end 0347 end 0348 else 0349 [y1,fs,ffx]=readsph(filename,mode,0); 0350 info=ffx{4}; 0351 if ~isempty(ffx{1}), fprintf(1,'Filename: %s\n',ffx{1}); end 0352 fprintf(1,'Sphere file type: %s\n',ffx{3}); 0353 fprintf(1,'Duration = %ss: %d channel * %d samples @ %sHz\n',sprintsi(info(5)/info(9)),info(6),info(5),sprintsi(info(9))); 0354 end 0355 0356 0357 0358