Is there a way to disable or hide the suggested responses?
I've been a Duolingo French learner for 11months and I'm adept at reading French aloud or at least well enough to have it recognized by speech to text. However, my response synthesis skill is severely under-developed - I have a difficult time choosing how to respond when spoken to. I can see the suggested responses being useful to someone practicing their pronunciation, but for what I want to improve upon, they act as a crutch.
Additionally, is text-to-speech on the roadmap? It would be nice to have a fully audio conversation, even to the extent of disabling the text output of the agent or instituting touch-to-reveal. Like a lot of language learners, I am quite a bit more adept at reading and writing than I am at speaking and listening.
Finally, a small bug(?). When I've selected a language to practice, speech to text shouldn't generate text which is outside of the practice language. I purposefully got sloppy a few times and Portuguese and Arabic were generated from my speech. I understand this isn't a deal-breaker, but it broke the immersivity of the experience for me.
Thank you! I've had a few people request an option to disable suggestions but you're the first to present the use case, which is super useful to me!
I'll add an option to minimize them.
> is text-to-speech on the roadmap?
You mean an option to hide the text, right? Because there's already text-to-speech (with the teacher responding out loud, unless maybe if your phone is on silent.)
Will definitely implement this. People requested it on my previous product and I'm a fool for not having implemented it yet. Text shouldn't get in the way of listening.
> I purposefully got sloppy a few times and Portuguese and Arabic were generated from my speech
I'll add a layer of checks to get rid of this for good :)
> my response synthesis skill is severely under-developed
A million times this. I can read at pretty much an 18 year old level, but i can barely synthesize at a 14 year old level, and this is where i need pushing.
> speech to text shouldn't generate text which is outside of the practice language
Another take: the speech recognition is too good with my sloppy banlieu french. I need this to spot my mistakes and make me practice correcting myself. I said "chuis something-or-other" and it correctly parsed my intention to say "je suis", but i want that pointed out to me. also my wife corrected me from the other room on "toi" vs "tu" and Giglish just glossed over the mistake.
I've been a Duolingo French learner for 11months and I'm adept at reading French aloud or at least well enough to have it recognized by speech to text. However, my response synthesis skill is severely under-developed - I have a difficult time choosing how to respond when spoken to. I can see the suggested responses being useful to someone practicing their pronunciation, but for what I want to improve upon, they act as a crutch.
Additionally, is text-to-speech on the roadmap? It would be nice to have a fully audio conversation, even to the extent of disabling the text output of the agent or instituting touch-to-reveal. Like a lot of language learners, I am quite a bit more adept at reading and writing than I am at speaking and listening.
Finally, a small bug(?). When I've selected a language to practice, speech to text shouldn't generate text which is outside of the practice language. I purposefully got sloppy a few times and Portuguese and Arabic were generated from my speech. I understand this isn't a deal-breaker, but it broke the immersivity of the experience for me.