Manuals

General Information

Edinburgh Speech Tools provides a set of executables, which offer access to speech tools functionality in the form of a stand alone program. As far as possible, the programs follow a common format in terms of assumptions, defaults and processing paradigms.

The following are generally true of most speech tools programs.

• Arguments to functions can be specified in any order on the command line. Most programs can take multiple input files, which by default have no preceding argument.
• Output by default is to standard out. The -o flag can be used to specify an output file.
• Often programs can read many formats of input file. Wherever possible, this is done automatically. If this can't be done, the -itype flag can be used to specify the input file format.
• The output file format is specified by -otype.

Specific documentation for each executable is given on separate pages:

• ch_wave
• ch_track

• tilt_analysis
• tilt_synthesis

• sigfv_manual
• spectgen
• sigfilter
• designfilter_manual
• pda_manual
• pitchmark

• dp
• ngram_build
• ngram_test
• viterbi

• na_play
• na_record

• wagon
• ols
• ols_test

• wfst_build
• wfst_train
• wfst_test_manual

• scfg_make
• scfg_train
• scfg_parse
• scfg_test

• siod
• fringe_client
• bcat
• xml_parser