junghanacs

#한글: #소리내어읽기 #이맥스 #텍스트음성변환 #한국어 #활용법

4 min read

Read Aloud: A Text to Speech Voice Reader

소리 내어 읽기: 텍스트 음성 변환 음성 리더

espeak-ng -v mb-us1 "Hello world"

Beneath its curmudgeonly exterior it is perhaps the most accessible software application out there.

딱딱한 외관을 가진 이 소프트웨어는 아마도 가장 접근하기 쉬운 소프트웨어 애플리케이션일 것입니다.

espeak-ng

[2023-11-20 Mon 06:17] https://emacspeak.sourceforge.net/

결론 일단 다음으로 실행하자 영어 읽어준다.

espeak-ng -v mb-us1 "Hello world"
sudo apt-get install espeak-ng-data espeak-data

https://github.com/espeak-ng/espeak-ng#documentation

espeak-ng/espeak-ng/blob/master/docs/mbrola.md - github.com

 
apt-cache search mbrola
 
sudo apt-get install mbrola mbrola-us1 mbrola-us2 mbrola-us3 mbrola-hn1
 
 
espeak-ng -v mb-en1 "Hello world"

설정

[2023-11-20 Mon 06:39]

clients  modules  speechd.conf
jhnuc :: /etc/speech-dispatcher » ll
합계 40
drwxr-xr-x   4 root root  4096  7월 14 09:55 ./
drwxr-xr-x 187 root root 12288 11월 20 06:33 ../
drwxr-xr-x   2 root root  4096  7월 14 09:55 clients/
drwxr-xr-x   2 root root  4096  7월 14 09:55 modules/
-rw-r--r--   1 root root 13953 11월 25  2022 speechd.conf
jhnuc :: /etc/speech-dispatcher » pwd
/etc/speech-dispatcher
 
sudo apt-get install speech-dispatcher speech-dispatcher-audio-plugins  speech-dispatcher-espeak-ng  speech-dispatcher-espeak espeak-ng-espeak espeak-ng-data --reinstall

에디터

[2023-11-20 Mon 06:47] eSpeak: Text To Speech Tool For Linux - itsfoss.com

sudo apt install espeakedit

Speech Dispatcher ; speechd - #리눅스 관리자

손대야 하니까 싫다.

Speech dispatcher - ArchWiki - wiki.archlinux.org

최신 신경 텍스트-음성 시스템 Piper를 사용하려면 Piper-tts-bin AUR 과 해당 언어에 대한 음성 패키지 중 하나(예: Piper-voices-en-us AUR) 를 설치하십시오. 아래 전용 섹션에 설명된 대로 Piper를 사용하도록 음성 디스패처를 구성합니다. 위의 AUR 패키지 대신 Flatpak을 통해 배포되는 자동화된 그래픽 설치 프로그램인 Pied를 사용하여 음성 및 음성 디스패처 구성과 함께 Piper를 설치할 수 있습니다.

spd-say "Arch Linux is the best"

piper : 한국어 지원 안됨

[2024-10-29 Tue 12:35]

Voices are trained at one of 4 “quality” levels: x_low - 16Khz audio, 5-7M params low - 16Khz audio, 15-20M params medium - 22.05Khz audio, 15-20M params high - 22.05Khz audio, 28-32M params

# Make sure you have git-lfs installed (https://git-lfs.com)
git lfs install
git clone https://huggingface.co/rhasspy/piper-voices
 
# If you want to clone without large files - just their pointers
GIT_LFS_SKIP_SMUDGE=1 git clone https://huggingface.co/rhasspy/piper-voices

Pied - #flatpak 패키지로 piper 관리

[2024-10-29 Tue 12:16]

https://github.com/Elleo/pied?tab=readme-ov-file

Pied makes it simple to install and manage text-to-speech Piper voices for use with Speech Dispatcher. Pied installs and configures the Piper neural text-to-speech engine to work with Speech Dispatcher. It can then be used to download and manage different voices.

 
flatpak install com.mikeasoft.pied.flatpak
 
# Required runtime for com.mikeasoft.pied/x86_64/master (runtime/org.freedesktop.Platform/x86_64/24.08) found in remote flathub
# Do you want to install it? [Y/n]:
 
/home/junghan/.var/app/com.mikeasoft.pied/data/pied/piper/piper --model /home/junghan/.var/app/com.mikeasoft.pied/data/pied/models/en_US-ryan-high.onnx --output_raw\n

piper korean

아래 녀석을 변환하면 될 것 같은데

 
git clone https://huggingface.co/Xenova/mms-tts-kor
 
samples
Add bryce, john, norman
5 months ago
MODEL_CARD
498 Bytes
Add bryce, john, norman
5 months ago
en_US-john-medium.onnx
63.5 MB
LFS
Add bryce, john, norman
5 months ago
en_US-john-medium.onnx.json

echo ‘Welcome to the world of speech synthesis!’ | ./piper —model en_US-lessac-medium.onnx —output_file welcome.wav echo ‘김정한 안녕하세요’ | ../piper/piper —model ko_KR-kim-medium.onnx —output_file welcome.wav

vosk

[2024-10-29 Tue 13:06]

pipx install vosk

References

Backlinks