Status of voices on different platforms (TTS)

[Fukurouonthesea]

I wanted to ask about what voices Asterics does or doesn't recognize, and whether it's possible to expand that.

Someone in a server I'm in said that for apple, personal voices don't show up, which made him not be able to use Asterics since they're very useful to apple people.
Also, on my Android tablet (Lenovo Tab M11) every app is a little different in what device voices it finds or doesn't find and how they sound, but Asterics is the most limited? (At least it doesn't echo though, which acapela and some others do a lot.) And I have no clue why.
I understand that voices are probably complicated to add, but it's frustrating to find a great voice in one app only for the other (generally for me better) app to not have that one.

I made videos of the different voices, but I have to add them from tablet after this post.
Thank you for the work!

[Fukurouonthesea]

screenrecorder-20260102-231233.mp4

Coughdrop (this is what it offers as default device voice in addition to the premium ones, so no clue which it is. But it's gotta be on here.)

screenrecorder-20260102-232332.mp4

Speech Assistant

screenrecorder-20260107-130228.mp4

TypingTalk

screenrecorder-20260102-232454.mp4

Asterics Grid

[klues]

Yes, voices are probably the biggest issue for us right now. It's a littlebit complicated. You can find some general background info here: https://github.com/asterics/AsTeRICS-Grid/blob/master/docs/documentation_user/11_voices.md but in short - there are online and offline voices:

• offline voices depend on platform (iOS, Android, ...)
• online voices can be the same on all platforms - but have the drawback that you need internet access

For offline voices:

• Apple / iOS:
  • Apple limits the voices which are available for webapps (like Asterics Grid), so e.g. some high quality voices are not available, although they are available on the system. Apple wants it like this - see my request and Apple's answer here: https://developer.apple.com/forums/thread/723503
  • I think there is no way of installing additional offline voices from other providers on iOS
• Android
  • it's possible to install different "speech engines" from different providers, so you're not limited to Google like you're limited to Apple on iOS. There is a setting called Preferred engine in Android. It's possible to install additional engines with apps like:
    • Acapela - paid voices
    • AhoTTS - I think for Spanish / Basque - is free, can be used to test if it works in general.
    • the default provider is "Google" on many phones, where you can also install additional voices in the settings, which then can be used in AsTeRICS Grid
    • on Samsung there is also the Samsung speech provider, which has great voices, however sadly they decided to remove them for non-Samsung apps in the latest update - so it's the same as on iOS, so bad. See: https://us.community.samsung.com/t5/Galaxy-S25/Can-t-use-Samsung-TTS/td-p/3131886
  • Sadly - there is a bug in Chrome - so if you need to switch languages and not only use the default voice set in the settings - you have to use Firefox on Android.
• Windows:
  • here you can install many relatively good offline voices in the Windows settings. You have to add a language in the settings and if it has this symbol, then there is a voice available:

• if you're using the Edge browser (or also Firefox in the latest version) you also get plenty of Microsoft's online voices, which have very good quality (but require internet access)
• it's also possible to install additional (paid) voices on Windows, like Acapela or CereProc which can be used in Asterics Grid

Online voices:

• currently we've implemented access to voices from responsivevoice - we're using their free plan, which now came to it's limits (and was not reliable many times) - so we'll have to remove this integration
• we're in contact with Microsoft and ElevenLabs and probably will be able to offer high quality voices from them for free, because of their non-profit programs - but we need some time to get these integrations done.
• ElevenLabs also offers free access to good voices (I think even voice clones), for people with diseases like ALS etc. - but you have to apply for it, see this blog post from ElevenLabs - we'll also integrate this to Asterics Grid, but it will also take some time.

So TL;DR:
It's a mess on all platforms and some big companies sadly don't give access to their good voices for web-based AAC apps. However, some quite good improvements will come via Microsoft and ElevenLabs cooperations.

[klues]

Regarding the apps where you've done the videos: all of them probably have some kind of mixture of the possibilities I've mentioned above, maybe using system voices, maybe using online voices. And for native apps (installed via PlayStore or AppStore) there are also other possibilities of integrating voices or maybe using more system voices.

[Fukurouonthesea]

Oh, thank you for the long answer!
I'm using chrome on android (and edge on windows for editing) - because firefox can't display the battery and displays the text in bold, which makes it very overwhelming for me. Sigh. I'll see what firefox voices look like.
The voices I had so far are Google TTS - had a poke around and installed the voices it offered in the languages I need. Unfortunately, it's not a samsung device, so I can't use those - and acapela is echoing badly on this tablet, no idea why. And I can't use online voices, because this device doesn't have internet access except for wifi, and service here is spotty.
I noticed the bug about the switching! Sigh.
I was thinking about trying Sherpa TTS (https://huggingface.co/spaces/k2-fsa/text-to-speech) but it looks really complicated.

As far as I know, all of these apps use system voices (well, except paid coughdrop), online and offline - so it was just a bit confusing what they were picking up or missing.

[deinhofer]

The voice engine of Sherpa TTS should actually work on Android devices, but I could not manage to get the voices listed within Asterics Grid.
I installed the voice engine of sherpa from here: https://k2-fsa.github.io/sherpa/onnx/tts/apk-engine.html
You must

1. manually download an .apk and install the file locally.
2. Select the speech engine in the text-to-speech system settings. On my phone it's "TTS Engine: Next gen Kaldi"
3. Then the engine and voices should be available within Asterics Grid, but I can't see them and cannot select them, neither in chrome on Android nor in Firefox on Android.

@klues any idea?

[Fukurouonthesea]

Oh okay, thank you - then I don't need to figure it out for now. I hope you get it working!

Oh, and I looked on firefox - well, waterfox, but they're supposed to work the same - and did not see any difference in voices to chrome. so... not sure what's going on there. Should I get and try "real" Firefox?