Raw Data : 185.000 GERMAN PRONUNCIATIONS FROM THE ENGLISH WIKTIONARY (3.5 GiB).
Download: 3.52 GB folder on MEGA
UPDATE: An .mdx dictionary was made with this raw data. See the new post HERE.
99.99% of audios are in .ogg format and are of very high quality. They are suitable to make a Pronunciation Dictionary for GoldenDict.
However, the audios can be used directly on GoldenDict after decompressing. Just Open the "Menu > Dictionaries > Sound Dirs " and choose the path of the folder containing all the sounds.
The naming of the sounds is very simple. For example, the German word “Haus” is named “Haus.ogg” .
Scraping was done on Linux after obtaining a .JSON file containing all the audio URLs:
The English Wiktionary contains more than 900,000 audios in many languages (not only English). Those pronunciations can be scraped thanks to this Source Code: