Repository navigation
Text-to-Speech (TTS) support using device's built-in engine for ebooks #1916
standhaftsohnsergius
started this conversation in
Ideas
Replies: 1 comment
|
This would be a very useful addition, especially for ebooks that do not have an audiobook version. Using the device’s native TTS engine feels like a practical approach because it avoids bundling voice models and lets users rely on their existing language/voice settings. Syncing TTS position with reading progress would make the feature feel much more seamless. |
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
Currently, Audiobookshelf app lets users read ebooks (EPUB, PDF, CBR, CBZ) but there's no way to have the text read aloud automatically. It would be great to add Text-to-Speech support that uses the device's native TTS engine (e.g. built-in Android TTS, iOS AVSpeechSynthesizer) to read ebook text out loud.
Why this would help:
Many ebooks in a personal library don't have a corresponding audiobook version — TTS would let users "listen" to them anyway, similar to how many other ereader/ebook apps already work.
Using the OS-level TTS engine (instead of building a custom one) keeps the app lightweight — no need to bundle voice models, and users can pick any voice/language pack they've already installed on their device.
It would let people continue their audiobook-listening habit even when only the ebook is available, e.g. while driving, doing chores, etc.
Suggested implementation ideas:
Add a "Listen" button (separate from "Read") inside the ebook viewer, similar to how audiobooks have a Play button.
Use platform TTS APIs directly: TextToSpeech class on Android, AVSpeechSynthesizer on iOS.
Sync playback position with reading progress, so users can switch between reading and listening seamlessly.
Allow basic controls: play/pause, speed adjustment, and voice selection (using whatever voices are already installed on the device).
Optionally, remember TTS position separately from manual reading progress, in case the two diverge.
Related discussions:
#1189
#1743
advplyr/audiobookshelf#601
I understand this has come up before, but I wanted to specifically highlight the approach of using the device's already-installed TTS engine rather than a custom/cloud-based solution, which could make this simpler to implement and maintain.
All reactions