Audio files for the vocab example sentences [Google Drive]

Bunpro
#1

Link to main google drive folder : bunpro vocab sentences - Google Drive

I’ll be adding at least 100 words of vocab per week, this is the speed I had planned on going for my own needs. I might try and get in more words to get through it quicker though, now that it is a community resource as well.

Current words : 100
Total audio time : 56m44s

If you encounter any mistakes, missed sentences or problems with pitch accents, feel free to let me know and I’ll fix it asap.

Original post :

So, I just started making audio files for all of the example sentences for the new vocab on bunpro using voicevox because I wanted to add all the example sentences to anki. I thought people might be interested in having these audio files available without people having to make them themselves (though I can highly recommend voicevox, it’s really insanely easy to use).

If people are interested I can upload them to google drive whenever I add new sentences.

@Jake @mrnoone @Asher @Fuga (and any other bunpro staff I am forgetting at the moment), would I be allowed to share these on the forums, or would you prefer I don’t at all?

Until I get the go ahead from the bunpro staff and know there’s any interest, here is the folder for the first word, so people know what to expect :

https://drive.google.com/drive/folders/12AzhcyxcSwWcucdKkTYeAkr-PcKK8qhQ?usp=sharing

18 Likes
#2

Sure, I don’t see any reason why not.

4 Likes
#3

Wow, I was completely unaware such a high-fidelity TTS program even existed. Thanks a lot!

3 Likes
#4

Great :grinning_face_with_smiling_eyes: shared!

Yeah, I also only discovered it recently, it really is amazing :slightly_smiling_face:

2 Likes
#5

Thanks so much!

3 Likes
#6

My pleasure! :grinning_face_with_smiling_eyes:

2 Likes
#7

Everything up to word 100 (サーフィン) is up now , 4 days ahead of schedule :partying_face:

2 Likes
#8

Thanks for the effort you put into this @mathijsdm! I actually tried out the software myself to get an idea of the limitations.

I noticed that the pronunciation itself was great, but that the software really struggles to find where natural pauses should go in sentences. However, if you take the example sentences, and chuck a few extra commas in there where natural pauses would be, the result is far far closer to real intonation.

Here is an example of one I added a bit of extra punctuation to.
http://sndup.net/wgyg

This does take a little bit longer, and I am not sure specifically what you are using the sentences for, but just thought that I would recommend giving this method a go (if you’re not already) (and if it doesn’t kill too much of your time!). Thanks again for the share! The hard work would be appreciated by many 🙇🏼‍♂️😊

1 Like
#9

I hadn’t thought about incorporating superfluous commas to get natural pauzes, that’s a really great idea! I could definitely incorporate that, as the result is indeed a lot better. It would slow the pace at which I can put out new sentences a lot though. I’m fine with either type of sentences myself, and don’t mind slowing down the speed I add new sentences to my deck personally, so I can absolutely do it. As it doesn’t matter all that much to me which way I’ll go, I will leave the choice up to the community.

  • Sentences as before (No natural pauzes) → Output of roughly 100 words (800 sentences) per week.
  • Natural pauzes incorporated → Output probably around 20 - 40 words (160 - 320) per week

0 voters

Especially the first weeks will be a bit slower to put out, as natural pauzes are something I haven’t put all that much thought in before, so I’d like to put a bit of attention into that so I get it right when I start outputting, so output might be even under 20 words the first week or two.

If incorporating natural pauzes, would people prefer it if I redid the first batch too?

  • Yes
  • Doesn’t matter to me
  • No

0 voters

I will await the results of the poll and make up my mind at the start of the weekend based on the results :grinning_face_with_smiling_eyes:

2 Likes
#10

Only do it if it will be useful for you too mate, and not feel like a chore! It’s for yourself first and foremost afterall.

Again thanks for the great resource :muscle:

2 Likes
#11

So, just a short update :

I will try and incorporate natural pauzes as best as I can. Some sentences this is a cinch and it doesn’t take that much longer to check than doing a normal sentence (roughly twice as long only ^^), though on some sentences I spend a good five - ten minutes and still not get a satisfying result, so those ones will either be without natural pauzes added or “finished” in the state they were when I ran my last test. In total it took me the same time to do 8 words as it took me last week to do 40, but I might be able to speed it up some more once I get more of a feeling on how it works software wise and stuff. If this is my max speed however, I will be keeping to roughly 20 words per week for the foreseeable future, though I might be able to free some more time once a work project ends in a few months.

on numbering convention

The sneaky buggers at bunpro seem to have expanded the N5 deck since me starting last week :stuck_out_tongue: Because I am too lazy to renumber all the folders you will start to see ###b or ###c and so on appear every so often, this is to add folders at the proper location, following the words they normally would appear behind. Unless they are planning another expansion this naming convention will be very minimal (8 words), so shouldn’t be too big a bother.

2 Likes