I’ve posted on here a long time ago about RTranslator, which is a phenomenal, real-time translation app for conversations. It’s a beautifully designed app that allows for on-device transcription and translation.
RTranslator has a few different modes that make it great for bilingual conversations. It can do simple text translations very well, like a handful of other apps. I actually use this one for that purpose because it has system context menus and OCR for image translation.
Where RTranslator really shines is in translating live conversations.
When it first came out a couple years ago, it felt like sci-fi. You can set it up to walkie-talkie mode, and the phone will automatically detect multiple speakers, identify the language they’re speaking, and translate to the other language. If you have TTS enabled on your phone, you can even hear the other half of the conversation (more on that in a moment).
The really amazing feature is if you have translator installed on two separate devices, then you can communicate with each other over Bluetooth. Even with your phone in your pocket, you can easily have a complete conversation with another person. Did I mention that it does all of this without any internet?
When you first set up the app, you’ll have to access the internet to download language models, but after that, it never goes online. The file size and processing power might be heavy on older or cheaper phones, but runs smoothly and easily on modern phones (I have a Pixel 7, which is hardly bleeding edge).
I like things that happen on-device for privacy purposes, but it’s also immediately practical for something like this. If you need on the spot translation, there’s a good chance you’re traveling and may not have access to reliable (or free) internet.
Most Android devices have built-in text-to-speech (TTS). So after you’ve installed RTranslator and selected your languages to download, then you’re good to go.
If you’re using an alternative OS (like graphene, which I use), then you’ll have to select one yourself. Note: I think that Google’s TTS also works on device without an internet connection, but I’m not 100% sure.
That’s where SherpaTTS comes in. Like the translator, Sherpa will need some space on your device and uses some processing power (although, the footprint is much smaller). It takes a little bit of time to set up, but most of that time is choosing your selected voice. There are literally hundreds of models to choose from.
You can set it up as your system default TTS, and then select it through the menu or during setup in RTranslator. Now you have modern, offline, live translation almost instantaneously.
Both apps are easy to set up and walk you through the process pretty quickly.
All software here mentioned is open source. Some of the models used by this software might be derived from proprietary sources, but the models themselves are considered open source. Nobody asked me to do this, but I like sharing great open source software. Projects like these don’t have an advertising budget, and rely on word of mouth.
You are not logged in. However you can subscribe from another Fediverse account, for example Lemmy or Mastodon. To do this, paste the following into the search field of your instance: [email protected]
5. Engage respectfully: Harassment, flamebaiting, bad faith engagement, or agenda posting will result in your posts being removed. Excessive violations will result in temporary or permanent ban, depending on severity.
6. Memes are not allowed to be posts, but are allowed in the comments.
7. Posts from clickbait sources are heavily discouraged. Please de-clickbait titles if it needs to be submitted.
8. Submission statements of any length composed of your own thoughts inside the post text field are mandatory for any microblog posts, and are optional but recommended for article/image/video posts.
I just tried out with two languages I speak fluently and it works pretty well.