What happened to audio generation?

Hey,

It seems like the option to generate audio has disappeared via card generation. Am I missing something?

Cheers

2 Likes

Ah I have figured it out. The layout of the card gen is borked, so the audio generation button doesn’t line up any more. Can this be fixed?

2 Likes

Hey!

That doesn’t seem right, I’ll get it fixed asap. Sorry about that.

Thanks for reporting the issue!

2 Likes

Thanks!

1 Like

Some words are also not generating any audio. I’ve been trying to generate one for “夜市” for a couple of hours now. Are there specific dictionary entries that truly can’t have audio? I’ve had some other words in the past too but I’ve made it work by inputting the hiragana instead. That didn’t work for 夜市.

Sorry to piggyback on this post!

1 Like

I did notice a few errors coming through from the Google API used to generate the audio. By now I’d have expected Kitsun to have cached almost all possible japanese audio files, but seems like that one wasn’t generated before yet.

I’ll let you know once it’s resolved! :slight_smile:

Thank you! :slight_smile:

1 Like

Audio generation should be working again :smile: Sorry for the wait!

Yay, it is!! Thanks so much for looking into it :slight_smile:

1 Like

I didn’t check when I generated cards using the dictionary and some have an incorrect audio and some don’t even have any audio !
I wanted to fix this but I didn’t find any way to add the audio afterwards.
Could you please tell me how I could do this ?

The incorrect audio can be because of the Kanji vocabulary having different readings (the TTS audio generation takes one that is most likely to be the correct one), if you play it during the card creation you can check it first and if incorrect you can specify a reading and generate the audio from that.

Not completely sure where the problem with no audio comes from, but I’m guessing it was due to the above problem.

You can fix your cards this way:

For cards that have incorrect audio:

  1. Change the audio value to be the hiragana version, so for example:

[sound:https://data.kitsun.io/kitsun-audio/japanese/丸腰-D.mp3] -> [sound:https://data.kitsun.io/kitsun-audio/japanese/まるごし-D.mp3]

  1. Check if the audio works, if not go to the dictionary page, search anything (doesn’t matter what) and open the card generation popup.

  2. Check the “Specify Reading” box and enter the hiragana version (so まるごし in the example above). Doing this will generate the audio file on the Kitsun hosting and after that it will work for your existing card as well. No need to save the card from the popup, you can just close it when you are done (or generate more audio for the cards with the other problems).

For cards that have no audio:

  1. Copy paste the audio value of another card and change the vocabulary value in it to match with your current card.
  2. If the audio does not work yet, it means it wasn’t generated yet. This time you can just search for the word and open the popup. The audio should automatically get generated as soon as you open the popup :slight_smile:

Please let me know if you have need further help! :slight_smile:

1 Like

Awesome ! Thank you very much ! :bowing_man:
I’m not able to try it yet because I came to the forums after doing my reviews :sweat_smile: but I can’t see why it shouldn’t work :stuck_out_tongue:
I will fix them in the mean time and the next batch should be perfect, thanks again ! :slight_smile:

(btw the ones without audio had [sound:null])

Edit : @Neicudi it worked ! The [sound:null] ones didn’t have audio generated. I generated them using the dictionary page as you said and it worked. Thanks :wink:

1 Like

Glad to hear :smile:

1 Like