The audio for saved sentences could be significantly improved:
Currently, it’s generated word by word, which results in unnatural and often incorrect pronunciation. A major improvement would be to use a more advanced Text-to-Speech engine that processes entire sentences at once.
Additionally, when original audio is available, it’s not being exported. That should be the preferred option over synthetic audio.
In my humble opinion, repeating sentences with original audio is one of the best ways to improve pronunciation, especially for languages with pitches/tones. I would love to see this feature added to LingQ.
This issue was first raised back in 2013 in this thread:
Thanks in advance, I hope you all have a great day!
If you’re not using the sentence mode, reader can not cut the exact word part from the audio. (Since the sentence is cut by the subtitle’s timing information)
If you want to try a more advanced TTS model. Look around my script:
Thanks a lot It’s really helpful.
But honestly, this kind of feature should already be part of the app. Having to use external tools, or pay extra (even if it’s just a small amount), shows that the app is not well integrated. And that’s the main point: something so basic should be built in by default.
As I said in another comment, many users have been asking for this for a long time. And the other user is right. Saving vocabulary sentences without the original audio is already a big issue. Replacing it with a robotic voice , especially in a language like Chinese, which I’m studying ,ruins the sentence. A lot of the time, I can’t understand what it says. Not because I don’t know the words, but because the pronunciation is so bad. I tested the same sentence with a paid TTS like Natural Reader, and I understood it right away. So the problem is not the learner; it’s the audio quality.
The bigger issue is that we constantly have to go outside the app to solve things that should already work inside. That’s what we mean when we say the app lacks integration. Everything should work smoothly inside the app, without needing extra tools or fixes.
From my experience learning Chinese, there are three serious problems:
No proper TTS. In Chinese, you’re learning both the spoken and written language. Without good audio, it’s almost impossible to read — especially in early stages, but even later on.
-Phrase collections without real audio are useless. Robotic voices often make things worse, not better.
-Fixing imported content is slow and frustrating. Sentence timing is usually off, editing is clunky, and sometimes changes don’t even save.
If the app wants to be truly useful, it needs to solve these core issues from the inside.
I’ve only used LingQ in English so I didn’t really feel any inconvenience with the tts. I totally agree with your opinion.
I think the reason why they don’t adopt AI-based TTS is the pricing. their pricing is more than x5 of mechanical ones. (Of course, as a consumer we don’t need to consider about them)
