pkubowicz / vocab-tts Goto Github PK

Using SSML-compatible reader:

./ssml.sh to1000 | xclip

Using MacOS say:

./mac.sh to1000 Luca

Using Mary TTS:

./mary.sh to1000 it

(outputs out.wav, you can use it to replay)

Using own TTS:

node to1000.js | sed 's/<[^>]*>/ /g' | tee /dev/stderr | xclip

• tel – telephone number
• tel-fast – telephone number, spoken fast
• tel-zh– telephone number, in Mandarin
• to1000 – numbers between 11 and 999
• year - years between 1880 and 2060, skewed towards recent years
• year-uni - years between 1880 and 2060, uniform distribution
• zh – numbers between 11 and 91

Readers:

• https://ttsmp3.com/text-to-speech/Chinese%20Mandarin/
• https://www.getwoord.com/ssml-editor (ctrl+shift+v)

Spec:

• https://docs.aws.amazon.com/polly/latest/dg/supportedtags.html
• https://www.w3.org/TR/speech-synthesis

Anki: File > Export, Notes as plain text, just 1 deck,, no HTML no tags, add note type.

Training listening:

cat ~/tmp/chinese.txt | ./chinese-input.sh | ./ssml.sh from-input | xclip

Training sentence building from random words:

cat ~/tmp/chinese.txt | ./build-sentence.sh

and from latest words:

cat ~/tmp/chinese.txt | ./newest-sentence.sh