Hey, just two requests.
- Will there be audio added to the vocab words once we type them in correctly?
- Is there a way to set all Translate review types to reveal and grade?
4 Likes
I can’t answer the first question but i can the second; not really, at the moment you have to go into the deck that you are studying, go into deck settings and change the review type to flashcard, then press the blue (or red?) button underneath that says change all to flashcard. This only does one deck though, so if you are working from multiple you have to do them individually. I think there’s plans to make a universal setting though soon.
2 Likes
@I-k-d Answered the 2nd one! Thanks.
Yup Universal setting coming real soon.
I’ll let @Asher answer the first one, I’m not sure.
3 Likes
We are planning on adding audio for the words themselves in the near future. However, this will be for the words only, not the whole sentences.
There is a chance that we will record the first 4 sentences with a native speaker, or do all 8, potentially using professional TTS software that sounds very natural. At the moment we are not 100% decided, as the first option will take an unbelievable amount of time to record, while the second option is not so ideal as there are a lot of people that don’t like TTS on principle, no matter how good it is 
.
Either way, yes, it’s on our to do list haha 
.
11 Likes
Thank you!
2 Likes
I tried the jpdb website, they are using TTS …
They use TTS for words + TTS for sentences (most of them, some sentences don’t have any audio).
At first, I thought it was a human xD Even the pitch accent is correct (with some rare exceptions).
I don’t know how expensive professional TTS are, but if they are cheap you could start with that (better than no audio, especially if the pitch is correct) and if the future move to a native speaker if you have more time 
2 Likes
The service JPDB uses for TTS was bought when the license was very cheap (comparatively) and now the same TTS costs far more, I think. Equally, the pitch accent aspect isn’t automatic and the JPDB dev had to generate all that himself. I am sure the Bunpro devs could contact him if they want to know what he did to get such good results but it isn’t as simple as just plugging it into the same TTS service (to get the same quality as JPDB), I believe.
I do second the opinion that the JPDB TTS is basically the best around though. I actually frequently play it to native speakers to compare their intonation etc with the TTS and no one has ever commented on it being TTS.
3 Likes
@CursedKitsune @JCalandr This is the software we have. There are 6 voices that have automatic settings, but you can also change the level of emotion etc to get more customized results.
Some of the voices are more theatrical, while others are more suited to reading plain texts. You may already know some of these voices. I’ve heard a few on regular TV commercials in Japan, and a lot of Japanese youtubers use it.
3 Likes
From the video, it looks great and surprisingly versatile! 
Summary
Looks … I mean … sounds
3 Likes
@asher why not use the TTS for all eight sentences then ask for feedback on those that sound most weird or unatural. Then focus on using native speaker for that list of priority weird or unatural sentences. This way you have the best of both world.
2 Likes
Only logical move is to voice the sentences using the Hatsune Miku software. It’s the only thing that makes sense.
4 Likes