For Windows operating systems: The Windows Command-Line prompt can be accessed by clicking the 'Start' button which is normally at the bottom left of the desktop, then click 'Run' then type ' cmd.exe ' within the text field excluding the single quotation marks, then press enter and the Windows Command-Line prompt will open in a new window.
For Mac and Linux operating systems: Simply open the terminal which should be located on the desktop to use these commands.
- Code: Select all
Usage: swift [OPTIONS...] [STRINGS | FILES...]
The "swift" application is a command-line utility for Text-to-Speech
synthesis. Given text or Speech Synthesis Markup Language (SSML), it
can render audio containing the material as spoken content to file or.
to an audio device.
Options:
-n <string> Specify a voice by name to use as the current voice.
This has precedence over the "-d" option.
-d <string> Specify a path to a directory containing a valid voice to
use as the current voice
-f <list> Synthesize text files or play audio in <list>. To
use standard input, use "-f -". Multiple file names
may be given. Without the "-f" switch, the default
is to synthesize text arguments on the command line.
When taking input from a pipe, the engine will always
treat two newlines as an input delimeter, and start
speaking text up to that point without waiting for the
end of the file.
-o <string> Write output to file specified. Without this option,
the default is to output to the default local audio
device.
-m <string> Mode to assume for input: one of "text", "ssml",
"phone", or "wave". The default is "ssml".
-e <string> Text Encoding to assume for input. Common encoding types
include: "utf-8", "utf-16", "iso8859-1", "iso8859-15",
and "us-ascii". The default is "us-ascii".
Note: This does not cause swift to convert text to the
specified encoding, but rather tells swift to expect the
input text to be of the specified encoding.
-x <string> Read a file containing special effects (SFX) filters to
be applied to the output audio. Overrides any default.sfx
file in the voice's directory.
-p <string> Set Swift TTS Engine parameters. Multiple parameters can
be specified as a comma-delimited string. For a list of
available paramters, run 'swift --params'. Syntax:
PARAM=VALUE[,PARAM=VALUE...]
-l <string> Load lexicon entries from the specified file. These
entries are appended to the voice's internal lexicon and
those from the voice's lexicon.txt file. In the event of
repeated entries, those found in this file override those
found in the voice's lexicon.txt file and the voice's
internal lexicon.
-t Display sentence text as it is synthesized.
-V Display copyright and version information for Swift, the
current voice, and the language and lexicon components
used by the current voice. You can use the -n and -d
options to use a particular voice. Otherwise, the
default voice is used.
--events Display a table of synthesis events.
--help Display a usage message, including descriptions
of all available options and usage examples.
--params List and describe all Swift TTS Engine parameters
available for use with the -p option.
--voices Display a table containing all available voices
on the system, including columns for the voice
name, version, license status, gender, age,
language, and sample-rate of each voice.
--search <string> Search for voices matching string.
--order <string> Loosely sort voices based on string criteria.
Voices matching string are placed at the top of
the list of voices.
The following can be used as search and order
criteria:
- name
- path
- version
- sample-rate
- license/key
- language/tag
- language/name
- language/version
- lexicon/name
- lexicon/version
- speaker/name
- speaker/gender
- speaker/age
To provide multiple criteria for search or order,
provide them as a semicolon-seperated list, such
as "speaker/gender=male;speaker/age=30"
--say-prgm-version Synthesize version information for Swift TTS
using the current voice. You can specify the -n
and -d options to use a particular voice.
Otherwise, the default voice is used.
--say-voice-version Synthesize voice name, language, and version for
the current voice, using the current voice.
You can specify the -n and -d options to use a
particular voice. Otherwise, the default voice
is used.
--say-all-voices Synthesize voice name, language and version of
every voice found on the system, using each voice
to announce itself. A table containing this
information is printed to the screen as the
voices are announced.
--reg-voice Enter a voice license key. This may be used
with the "-n" or "-d" options to specify a
voice to license, or with the Additional
Licensing Options shown below.
--reg-ports Enter a concurrency license key to allow for
simultaneous speech synthesis. This may be used
with the Additional Licensing Options shown below.
Additional Licensing Options:
These additional options are supplemental to the --reg-voice
and --reg-ports options.
--customer-name <string> Used with --reg-voice or --reg-ports.
Your name as it appears in your invoice.
--company-name <string> Used with --reg-voice or --reg-ports.
Your company name as it appears in your
invoice. If no company name is
specified in your invoice, use:
[--company-name ""].
--voice-name <string> Used with --reg-voice Only.
The name of the voice to license.
--num-ports <string> Used with --reg-ports Only.
Number of ports for concurrency license.
For Unlimited Concurrency, use:
[--num-ports ""].
--license-key <string> Used with --reg-voice or --reg-ports.
Your license key.
Licensing Examples:
swift --reg-voice
Register a voice interactively.
swift --reg-ports
Register port concurrency interactively.
swift --reg-voice --customer-name "John Doe" --company-name
"Initech" --voice-name David --license-key "XX-XXXXXX-
XXXXXX-XXXXXX-XXXXXX-XXXXXX"
Register a voice using command-line parameters.
swift --reg-ports --customer-name "John Doe" --company-name
"Initech" --num-ports 96 --license-key XX-XXXXXX-XXXX
XX-XXXXXX-XXXXXX-XXXXXXX
Register port concurrency using command-line parameters.
Examples:
swift --voices --search "speaker/age=30" --order "sampling-rate=8000"
swift "Cepstral: We Build Voices."
Speak the contents of the quoted string. If the quotation marks
are not used, the system will concatenate all text into a string
to synthesize.
swift -l /path/to/mylex -f /path/to/myfile.ssml
Add the lexicon in the file "mylex" to the text processing,
and then render the contents of the SSML file. The default
mode for both file and command-line input is SSML.
swift -n William -m text -f /path/to/file.txt -o file.wav
Use the voice William, assume the mode is raw "text" rather
than SSML, render the contents of "myfile.txt" to audio,
and save it to the given output file.
Cepstral Swift 4.2.0, February 2007. Copyright (C) Cepstral LLC.
If any assistance is needed please feel free to post a reply within this topic and I will try my best to help.