Hello,
I am experiencing an issue with the audio playback feature that occurs on Windows and Android, but interestingly, it does not happen on iOS.
The problem is that the audio reader starts reading the text from the next page before the interface actually turns the page. Because of this lag in page-turning, I have to manually flip the pages quickly myself to keep up with the audio, which defeats the purpose of the auto-flip feature.
I’ve noticed a few differences in how it works on different platforms:
On iOS: It works perfectly. It highlights word-by-word (rather than just the first word of a phrase) and flips the page exactly when it should.
On Android/Windows: The voice quality and pronunciation are actually better, but the synchronization is off.
Suggested Fix:
Would it be possible to make the reader pause at the end of the current page, trigger the page transition, and then start reading the next page? Or perhaps you have a smarter technical solution to keep the audio and UI in sync.
It’s a fantastic feature with great benefits, and I’d love to be able to use it smoothly on all my devices.
Thank you!
When you say the synchronization is off, what happens for you exactly? Does it highlight word-by-word? Or does it only highlight the first word of each sentence?
I forgot to mention that the continuous reading of the sentence before moving to the next page only happens if the last sentence carries over to the next page. However, if the page ends exactly where the sentence ends, the reading of the next page won’t start until it actually turn to it. Also, I find the text-to-speech quality on Windows/Android to be more natural and less robotic than on the iPhone. I don’t prefer listening to the iPhone reader, even though it solves the problem.
Thanks for the detailed report. I’ve now shipped a fix for this.
What was happening: some voices (including the Google Español voice) don’t emit word-by-word timing events, so the reader couldn’t tell when the spoken audio had crossed onto the next page. That’s also why those voices only highlight the first word of each sentence.
Readlang now detects, on the first playback with a given voice, whether it emits those word timings. If it doesn’t, sentences that span a page break are split at the page boundary, the page turns, and a new utterance starts on the next page.
Please refresh your Readlang browser tab and let me know if it’s working better for you.