cat

来自「speech signal process tools」· 代码 · 共 199 行

SDCOMP(1-ESPS)                              ENTROPIC RESEARCH LAB



NNNNAAAAMMMMEEEE
     sdcomp - listening program to compare two sampled data files

SSSSYYYYNNNNOOOOPPPPSSSSIIIISSSS
     ssssddddccccoooommmmpppp [ ----bbbb _s_h_i_f_t-_v_a_l_u_e ] [ ----cccc _c_h_a_n_n_e_l ] [ ----dddd _d_e_l_a_y ]  [   ----
     {{{{ffffpppprrrrssss}}}} _r_a_n_g_e ] [ ----gggg _g_a_i_n ] [ ----iiii ] [ ----RRRR _n_r_e_p_e_a_t ] [ ----nnnn _s_t_e_p ]
     [ ----wwww _w_i_d_t_h ] [ ----xxxx _d_e_b_u_g-_l_e_v_e_l ] _s_d_f_i_l_e [ _s_d_f_i_l_e ... ]

DDDDEEEESSSSCCCCRRRRIIIIPPPPTTTTIIIIOOOONNNN
     _S_d_c_o_m_p cycles through one or more ESPS FEA_SD(5-ESPS)  files
     playing  a  portion  from  each  file  by means of the local
     play(1-ESPS) program.  _S_d_c_o_m_p assumes that  the  local  play
     program has the same command line options as _m_c_p_l_a_y.

     After _s_d_c_o_m_p starts and after a portion is played from  each
     file, the user is prompted for the next action.  The follow-
     ing commands are then accepted (<CR> refers to the  "return"
     key):

          <CR>      play next portion from each file
          r<CR>          repeat current portion from each file
          b<CR>          back up and play previous portion from each file
          c<CR>          continue playing portions without stopping for commands
          q<CR>          quit


EEEESSSSPPPPSSSS PPPPAAAARRRRAAAAMMMMEEEETTTTEEEERRRRSSSS
     The ESPS parameter file processing is the same as  that  for
     _m_c_p_l_a_y.

EEEESSSSPPPPSSSS HHHHEEEEAAAADDDDEEEERRRRSSSS
     ESPS headers are treated as in _m_c_p_l_a_y.

OOOOPPPPTTTTIIIIOOOONNNNSSSS
     The flags bbbbccccggggiiiirrrrwwwwxxxx are passed to the local play program every
     time  that  program  is invoked.  In addition, the following
     options are supported:

     ----{{{{pppprrrrssssffff}}}} _r_a_n_g_e
          Selects the first portion that will be played from each
          file,  with  syntax as in _m_c_p_l_a_y. The default is ----ssss0:1.
          If the part of the range specification after the  colon
          is  omitted  (giving  only  the  starting  point),  the
          defaults are as follows:

               ----pppp<start>:+7999
               ----rrrr<start>:+7999
               ----ssss<start>:+1
               ----ffff<start>:+79

          These give approximately the  same  length  (1  second)
          when  the  sampling  frequency  is  8000  Hz.,  but not



        Copyright Entropic Research Laboratory, Inc. 1G1        1






SDCOMP(1-ESPS)                              ENTROPIC RESEARCH LAB



          otherwise.

          The effect of the ---- {{{{ppppssssrrrrffff}}}}  option  is  to  select  the
          starting  position  and  length  of  the portion.  This
          length remains in effect for subsequent  portions  that
          are  played, while the starting positions of subsequent
          portions are determined by the ----nnnn option.  Note that ----pppp
          and ----rrrr are synonyms.

     ----dddd _d_e_l_a_y
          If ----dddd is not specified, _s_d_c_o_m_p passes all the  filename
          arguments  to  one  invocation of the play program each
          time a portion is played from the files.   If   ---- dddd  is
          specified,  then  each filename is passed to a separate
          invocation of the play program, which is followed by  a
          pause  of  _d_e_l_a_y seconds.  (Note that specifying ----dddd0000 is
          not equivalent to omitting the option entirely.)

     ----wwww _w_i_d_t_h [100]
          This option defines  the  frame  width.  The  width  is
          ignored unless the ----ffff option is also specified.

     ----nnnn _s_t_e_p
          Selects the amount by which the  starting  position  is
          changed  from one portion of each file to the next por-
          tion of each file.  If the _s_t_e_p is less than  the  por-
          tion length (determined by ----{{{{pppprrrrssssffff}}}} ), the portions will
          overlap.  If it is greater than the portion length, the
          portions  played  will skip segments of the files.  The
          _s_t_e_p must be expressed in the same units as the  _r_a_n_g_e.
          That  is, if _r_a_n_g_e is specified with ----pppp or ----rrrr (points),
          the _s_t_e_p is assumed to be points, etc.  The default for
          _s_t_e_p equals the length of the initial portion specified
          by the ----{{{{pppprrrrssssffff}}}} option.



FFFFUUUUTTTTUUUURRRREEEE CCCCHHHHAAAANNNNGGGGEEEESSSS
     The user should be able to change the order in  which  files
     are played.

     There should be an option that causes _s_d_c_o_m_p to  select  the
     order  in which files are played and report that order after
     all of the speech has been played.

     The user should be able  to  change  the  position,  portion
     length, and step size dynamically.

     The user should be able to specify a separate gain (---- gggg)  or
     shift-value (----bbbb) for each input file.





        Copyright Entropic Research Laboratory, Inc. 2G2        2






SDCOMP(1-ESPS)                              ENTROPIC RESEARCH LAB



SSSSEEEEEEEE AAAALLLLSSSSOOOO
     play(1-ESPS), FEA_SD(5-ESPS)

WWWWAAAARRRRNNNNIIIINNNNGGGGSSSS
BBBBUUUUGGGGSSSS
     _A _r_a_n_g_e or _s_t_e_p value in seconds is  restricted  to  integer
     values (this reflects a bug in _p_l_a_y ).

AAAAUUUUTTTTHHHHOOOORRRR
     Manual page by John Shore; program by Rodney Johnson













































        Copyright Entropic Research Laboratory, Inc. 3G3        3