Are you also encountering this issue: a section of the German audio I upload for transcription isn’t being transcribed properly?
I’ve noticed that approximately the first five minutes appear correct, but then the audio and the generated text stop aligning. Specifically, there simply isn’t enough text to match the audio.
After about one or two minutes of the audio playing, the text and audio synchronize again, which shows that a segment of that duration was completely missed during transcription. Is anyone else experiencing this particular problem?
I witnessed the same issue in English, too. However, since my listening skills are quite decent so I could understand the missing part even without looking at the transcript. But I guess it must be quite annoying for beginners in a foreign language. Today I have also come to know that there is a cap of 600 minutes of audio transcription in a month. So, I will have to wait until July to upload audio for the transcription and do Listening-Reading.
Update: The issue is still ongoing. I’m just realizing that almost three minutes of my audio was not transcribed! Is there any update on the fix yet please?
Thank you for checking in with your team regarding the status.
I’ve noticed a significant problem once again with another lesson: a substantial portion of my 40-minute German lesson has not been transcribed.
I’m seeing entire time periods of at least three or four minutes where the transcription is completely missing. I’m attaching the link to the lesson here if you’d like to take a look. Login - LingQ
I would really appreciate it if you could provide an update on this issue’s status or if the team has identified the cause. I’m very hesitant to upload another long audio file, whether it’s 30 or 40 minutes, only to find large sections of it remain untranscribed.
@kadirkaratas Our testers team tried to reproduce the problem with different files but weren’t able. However, we provided the audio file you had a problem with and lesson URL to our developers and they are trying to figure out what went wrong. We will hopefully identify the problem soon.
I just uploaded a new audio file, and I’m still encountering problems. In the lesson I mentioned above, at the specific timestamp I provided, the text completely stops matching the audio. This indicates that a significant portion of the audio has not been properly transcribed.
Even before that exact timestamp, perhaps one or two minutes earlier, the synchronization between the audio and the text is off, meaning the correct text isn’t highlighted.
However, at the specified timestamp, the audio and text completely diverge, and there’s simply not enough text to match the remaining audio.
This is incredibly frustrating for me to experience because a huge chunk of the audio isn’t being transcribed. I’ve been dealing with this issue for the past two or three weeks now, and it happens with every German audio file I upload for transcription!!
As a result, this transcription feature is becoming largely unusable for me.
Please, at least tell me that this is an acknowledge issue and some people are working on it…
Is this problem caused by the translation is done in background and the translation can take long time? Once when I imported a text some, but not all text was translated after few minutes. When I a looked at the lesson some day later, previous missing transcript parts was gone.
Thank you for letting me know the team is working on this. To make it easier for them to identify and solve the problem, I’ve included the specific lesson link and the exact time the issue appears. I am currently not uploading any new audio because of this transcription problem. It’s a major inconvenience when I’m listening to the audio and the corresponding text isn’t there. This would be an even greater problem for longer files, say 30 minutes or more. I just want to be clear that this is an important issue for me, and I hope this information helps. Cheers.
I believe I am still experiencing the issue. That is why I did not get anything transcribed lately in order to not have an audio with a half-transcribed piece of text. I am waiting for an official response from the team, saying that they have diagnosed the problem properly and dealt with it.