Arabic course ruined by switching from real voice to TTS Engine
I understand you might be data mining the user responses and correct answer rates to build a valuable TTS engine but in doing so ruined the only thing I loved about Duo.
It was amazing to see how accurate the pronunciation was for all arabic letters and words in the beginning. Real vocal actors, possibly of Egyption origin, resulted in such high Quality.
One day I login to find all audio has been replaced with TTS engines which pronounce everything wrong.
What a shame!
Hm, I always thought the first version was also TTS. Some things do seem really wonky after they switched the voices though. There are syllables added in the middle of the words that I wouldn't suspect to be there (I'm not talking about the declension endings at the end of the words here) and را is for some reason pronounced as "rii" in the tapping exercises (yes, I'm mostly still practicing the earlier skills).
If you want to listen to real native speakers' recordings you can use forvo.com. The difference is huge. You can add your own words and phrases and wait for someone to record them.
p.s. Duolingo "officials" will never see your post in here. You miiiight get some attention if you post it to the main forum titled "Duolingo", but generally they do not care.
They never had any real recorded voices in Duolingo, neither in Arabic, nor in any other course! ALL voices are tts, the difference is that some voices are more accurate than others. They all make mistakes that Duolingo is unable to do anything to fix them, they just buy a language pack and that's it. After spending 3 years in here I consider it to be the main flaw of Duolingo. I have learned so many wrong pronunciations in Russian that I'm still trying to get rid of. And that voice was one of the better ones, it made few mistakes compared to the one they replaced it with! It seems they have a habit of using worse and worse voices...
The course has been ruined by adding a voice that does not say the words as intended or as voweled, leaving out the vowel sound on the last letter of the word, pronouncing an "a" as a "u", and adding 2 vowel sounds for y’s and w’s. This will only confuse and frustrate new learners and they will give up due to lack of standardization. For me, I have to turn down the audio to read words or sounds in the matching exercises or I can’t find the word the way the voice is pronouncing it. I don’t understand why this was done in already complex language and one that intends to teach MSA, S standing for Standardized, especially when the first roll out had such precise, clear, consistent pronunciations.