I’m currently trying to learn Vietnamese and wanted a quick way to be able to add audio to my cards. Unfortunately, while there is the option of batch adding audio from Google Translate to cards, the Vietnamese test-to-speech for it sounds very robotic and therefore isn’t the best to learn from. I was able to find two other TTS services (see: https://fpt.ai/tts# and https://vbee.vn/) that have much more natural sounding speech patterns, but they aren’t supported by Awesome TTS, so for now I’ve been manually exporting the sound files and adding them to my cards. Is there any way to request support for this TTS in the Awesome TTS add-on? Or is there a way that I could add it myself for just my version of the add-on?
Hi, i’m the maintainer of AwesomeTTS. Google Cloud TTS has neural-network trained Vietnamese voices (https://cloud.google.com/text-to-speech/docs/voices), have you tried those ? They are guaranteed to sound very natural. You do have to pay for them, or find an evaluation key.
One easy way to try out those voices is to become a fan on my patreon page: https://www.patreon.com/lucw
if you become a paid fan, you’ll get access to Google Cloud TTS, Azure and Forvo API keys for your personal use.
If you do want to go ahead with adding those services you mentioned, you’ll need to know how to write software in python or find someone who can do it for you. AwesomeTTS is open source, so adding a service is not complicated, but the difficulty may lie in integrating those services if they are not open to having other apps use them.
This one https://ankiweb.net/shared/info/1436550454 has all the functionality and is the one I would recommend. The add-on was not maintained for a period of time, so we ended up with some forks. But as of a month ago, we are trying to unify the codebase.
Thanks for the response! I did try Google Cloud TTS, and while it does sound quite natural, it only has Vietnamese available in the Northern accent. Since I’m trying to learn the Southern accent, it’s preferable for me to add Southern accent voice files to my cards.
I don’t have any experience in python or coding, but am willing to learn to get small projects done. Do you have any recommendations for how to get started?
Open an issue on the awesometts github page and we’ll get things started. The second api you mentioned requires payment. The first one seems to be free up to a certain amount of characters. If you want to implement the second one, we’ll need some help with paying for the access.
It doesn’t prevent other parallel versions from being present on the add-on page. The author of the “unofficial” version was the maintainer of awesometts for a while. He may want to keep his version to deploy updates more easily.
The unofficial version likely will be updated from time to time, though it might take some time before I could get back to it, but at the moment I’m not interested to fix or add new services, I won’t keep it in sync with the upstream repository and the add-on won’t receive updates from the upstream repository if something got updated or fixed.
Sorry I don’t know. If you want a stable experience, stay on Anki 2.1.49. Note that HyperTTS, (AwesomeTTS 2.0) is already compatible with Anki 2.1.50. Give a try: HyperTTS - Add speech to your flashcards - AnkiWeb
I did purchase your addon, but it has big problem when I do mass TTS, there is no option to skip card if its already have an audio in field.
Dont you guy understand that when a decks with hundred of cards constantly updating and adding new cards how should I add audio only to new cards? Its pain to wait for all card being re-record again. Please consider to add this option to ur addon thanks you