synthesize: Synthesize Speech

In cloudyr/aws.polly: Client for AWS Polly

Synthesize Speech

Description

Pass text to the synthesis API and return an audio file

Usage

get_synthesis(
  text,
  voice,
  format = c("mp3", "ogg_vorbis", "pcm"),
  rate = deprecated(),
  lexicon = deprecated(),
  ssml = FALSE,
  lexicon_names = NULL,
  sample_rate = NULL,
  engine = c("standard", "neural"),
  language = NULL,
  ...
)

synthesize(text, voice, ...)

Arguments

text	Either a plain text character string (maximum 1500 characters) or a character string containing SSML (ssml should be set to TRUE).
voice	Name of an AWS Polly voice. See list_voices.
format	Output file format: one of ''mp3'‘ (default), '’ogg_vorbis'' or ''pcm''.
rate	Deprecated. Use sample_rate instead.
lexicon	Deprecated. Use lexicon_names instead.
ssml	A logical indicating whether text contains SSML markup.
lexicon_names	Optional character vector (max length 5) specifying the names of lexicons to apply during synthesis. See get_lexicon.
sample_rate	Audio frequency in Hertz. One of '24000', '22050', '16000' or '8000'. If 'NULL' (default), 'AWS' will pick an appropriate value depending on the voice, engine and output format.
engine	Engine type: either ''standard'‘ (default) or '’neural''. Not all voices support the neural engine, see list_voices.
language	Optional language code. This is useful for bilingual voices.
...	Additional arguments passed to pollyHTTP.

Value

get_synthesis returns a raw vector (i.e., the bytes representing the audio as the requested file format). synthesize is a convenience wrapper around that, which returns an object of class “Wave” (see Wave).

Examples

## Not run: 
hello <- synthesize("hello world!", voice = "Geraint")
if (interactive() & require("tuneR")) {
    try(play(hello))
}

## End(Not run)

cloudyr/aws.polly documentation built on June 18, 2022, 3:45 a.m.