There is audio recorded by native speakers available on JapanesePod101, which is the one used by JDIC. It's under the URL http://assets.languagepod101.com/dictio … ;字. Just replace かな and 漢字 by the kana and kanji, respectively.

I wrote an Anki plugin for myself to download audio for cards in my deck and write the filename in a field. Anki shared plugin upload seems to be disabled for now, so I uploaded it here: http://pastebin.com/ZYQerXFA

It works for me, but haven't tested it that much, so use at own risk and try it on a small deck first. You need to edit the file to set model and field names correctly for your own deck first.

@Tori-kun: That is strange. Is the model name correct? It should find some cards normally.

If you want to you can send me the deck and your configuration, then I'll try it out myself.

Fixed the error found by Tori-kun, the new version of the script is here: http://pastebin.com/Cz63NEHu

A note for usage: Be sure that the specified Kana field just contains the Kana reading of the word. It will fail if there is nothing or Kanji with Furigana in Brackets. For Kana-only words both the Kanji and Kana field need to contain the Kana.

I updated the plugin to strip all non-kana characters from the kana field string before querying for the audio. That should work for you in most cases, except when the reading is written like 駆け寄[かけよ]る which becomes けかけよる.

New version here: http://pastebin.com/L2JauhXY

Good it works for you now :-)

As to not download audio repeatedly, the plugin skips cards for which either the audio file is already present (so if you have bogus files they won't get updated) or for which the audio field contains something other than whitespace. So if you leave the "audio not found" message empty, the plugin will try to re-download the missing media again every time, which takes some time. But you could just change it to one character like - or set the font color to the background color in Anki if it annoys you.

Writing the kana in a new field is definitely possible, it's just a matter of coding ^^ Wasn't much work based on my script, so I made a Kana update script: http://pastebin.com/pTwnQDgW

Just set NEW_KANA_FIELD to the new field name, and run the plugin once to fill the field. Then you can point KANA_FIELD of the audio plugin to that field.

You can clear the Audio field from Anki by selecting all the cards to update in the card browser, then choosing Edit -> Find and Replace from the menu, entering .* as the Find query, selecting the Audio field as In, leaving Replace With empty and checking Treat input as regular expression. Then pressing Ok should clear the Audio field.

After that with an empty media folder the plugin should try to download audio for all cards. This probably takes some time, so don't exit Anki in between.

Last edited by cryptica (2012 August 04, 11:22 am)

I don't really have any clue, but it could be that languagepod101 has a download restriction.

For the cards that didn't get mp3 files, did the plugin write anything in the audio field? If not, you could try to start the plugin again later to resume downloading. If it works then, the problem would with languagepod101, otherwise I guess it could be a plugin error.