The [R] of Tsurumaki Maki English AI

think AI voices in English are taking the [R] as a vowel. By writing words that start with a consonant together with an [R], like for example words like Crazy, brother, fracking, etc, SynthV lengthens the duration of the sound of the [R].
I have seen that in the AI voices in Japanese, when one writes complete words, the first vowels tend to last much longer than the others, so I think that is why in the English version the [R] is being taken as a vowel.

Hopefully this is something they can fix, I’ve encountered it a couple times myself as well.

For the time-being it seems we can work around the issue by adjusting the individual phoneme timings. Thanks for providing a clear explanation of the issue!