Generate Timestamps noise issue

Support & Feedback Forum
1

G’day Comrades,

Just reporting some consistent failures I’m having with “generate timestamps”. In these lessons (one, two, three), for example, the process says that it is generating timestamps, and then says that it has done so successfully.

What has actually happened is either nothing has changed, or the first 2-3 sentences have been timed, and then it simply gives up. Everything else, the next ~20 minutes or so of audio, is left un-timed. As such, it is not tracked when I’m listening, and the sentence mode relies on AI audio.

I presume this is due to the background music for the podcast introduction. Given that a huge array of YouTube footage and podcasts have this, it would be great to have a process that’s a bit more resilient to this sort of thing. Even, if it can’t get some portions, just does a quick skip ahead and tries again (even 2-3 times of 10-15 seconds would probably overcome most of these).

At worst, if this isn’t possible, it would be better to have the process report that is has not succeeded, rather than thinking it’s sorted (this has to be done on desktop/browser), going somewhere I only have my phone, only to find my planned study is substantially disrupted.

Happy to provide any more info needed.

Travis.

2

Additional info - listening time also doesn’t count on these lessons. As an example, I listened to an entire 20 minute lesson yesterday and it didn’t contribute anything to listening time or score. On iOS, the files show as not having any length (i.e. where audio duration would usually be, it just says “–:–”.

I have used ffmpeg to turn one of the files into a .wav to see if this helped, and it didn’t. So I presume it is indeed background noise or something about the files themselves rather than an encoding issue.

Would appreciate any feedback on this - listening to real language from podcasts etc is another one of the core functions of LingQ that drew me in to sign up, it’s disappointing to have it not work.

Travis.

3

Thanks, we will look into this and see if there’s anything that can be done.