Sentences as flashcards: audio?

[Selek]

Hi all,

I recently discovered that we can convert Pleco's example sentences into flashcards. I love this feature, because I like to learn a language through phrases, not just individual words. In fact, I already use Anki to quiz myself on sentences from Spoonfed Chinese, and I also use the mobile app Immersive Chinese for sentences. But what I like about Pleco's sentence-flashcards is that I can pick for myself which sentences I want to practice. Some, for me, are more useful than others. At my stage, I especially prefer simple sentences that test me on one or two words or phrases.

My only complaint, though, is that the text-to-speech audio in the sentences isn't replicated on the flashcards. Instead of a fluid (if robotic) readout of the sentence that I get when I play it at the dictionary definition, in the flashcard I get the slow character-by-character readout. Is there a way to transfer the audio as well as the sentence characters to the flashcard?

Many thanks in advance.

 

[mikelove]

Go into Settings / Audio - is 'use TTS if no recording' turned off? Turn it on if so, that should get it back to using TTS.

 

[Selek]

Many thanks for your reply. That worked, mostly! Now when I encounter a sentence-flashcard, the female TTS does indeed work -- but it alternates with the male voice reading each character one by one. So if I listen to a card with TTS, then do the next card, the next card has the male character-by-character voice unless I tap "listen" again. Not a huge deal, but it would be nice if I could just disable that male voice altogether for sentence-flashcards. (For normal flashcards, I love having the alternate male and female voice recordings, of actual humans.)

 

[mikelove]

That's unfortunately not possible at the moment unless you have one of our male TTS add-ons, since it doesn't recognize sentence flashcards as being a distinct type of thing from word flashcards. (you could of course set it to female only but then that would apply to regular words too)

However, you could try our 4.0 beta (sign up at pleco.com) which has a lot more flexibility about this sort of thing. There's a 'Pleco Legacy' app you can install along with it which lets you keep using the old version at the same time.

 

[Selek]

Thanks for that information. I will check out the male TTS add-ons as well as the 4.0 beta. Both sound promising!

Thanks also for maintaining and improving this awesome app. I use it more than any other app on my phone. It's great!

 

[thph2006]

Mike, you advise "(you could of course set it to female only but then that would apply to regular words too)". I've tried that and still get the alternating female/male voices. I'd be absolutely happy to live with female-only voice to get the phrases to play in a natural tone and cadence. If that can't work having a button to replay the female voice on a flashcards would be nice. It gets difficult waiting for the one character at a time voice to slowly get through a long phrase.

I have:
Settings>Audio>Preferred gender Female
Settings>Audio>Use TTS if no recording ON
Settings>Audio>Voice Apple Mandarin Female.
The flashcards I'm reviewing are phrases.
Pleco version is 3.2.77
IOS version 18.4.1

 

[mikelove]

Go into "Add-ons" and delete the audio recordings add-ons and any male TTS engines from there - does that help?

 

[thph2006]

Thanks! Yes, deletion of all the recorded voices fixed the issue (except that I now only have female voice rather than being able to alternate female/male).

I assume I can get the addons back if I need them but if you could confirm it would make me more comfortable.

I deleted:
Audio Pronunciation (female)
Audio Pronunciation (male)
Extended Audio (female)
Extended Audio (male)

I have no TTS addons but Apple Mandarin Female is enabled under settings>SENTENCE AUDIO>Voice

 

[mikelove]

Yes, you can always re-download them if you need them.

 

[thph2006]

It appears now that using the Apple Female TTS, some cards containing sentences with numeric content are spoken incorrectly. I don't recall an issue when the individual character recordings were used but since I deleted them I can't confirm.

Here's an example of the issue:
From HSK1 Examples Flashcard set with Apple Female TTS enabled:

Card Info:
他今年24岁。
Ta1 jin1nian2 24 sui4
He is 24 this year.
Voice Output:
Ta1 jin1nian2 er4 shi2 er4 su4 sui4
(something to the effect of "He is 22 4 years old this year")

In multiple cards digits were reversed or randomly added to the translation.

Questions:
- Is this a known problem with the Apple TTS?
- Will it also happen with the paid TTS if I order it?
- Will the paid TTS upgrade to the neural TTS when I move to 4.0 or be an extra charge?

 

[mikelove]

It's actually not so much about the Apple TTS as about example sentences - particularly in imported flashcard lists or dictionaries - not always being coded to have the pinyin and numbers perfectly in sync; the TTS thinks it doesn't have a pinyin reading for one of the '2's and pronounces the character instead.

I don't have the flashcard file you're using handy, but I would expect that the problem would occur with the paid TTS as well since that wouldn't change the data. If you edit the card to insert an inverted question mark ¿ in the place where the pinyin and characters are out of sync - I'm guessing in this case it would be after the 2 in the characters - that should hopefully fix it, but you would potentially need to make a similar fix for any other sentence you import with numbers in it (or letters, or other things that confuse the TTS).

We're not planning to charge for the neural TTS upgrade - we already rolled it out on Android and we don't charge for it on there.