Kevin sings very clearly and articulates every sound. It has a musical singing quality. This is great for some applications. But it may sound very clean for some genres. Is it possible to make it less clear with some mumbling?
This isn’t something I’ve done before, but there are likely four main things you can do:
- Reduce the tension parameter
- Reduce phoneme strength for consonant sounds (or just any sounds that stand out too much) (phoneme strength sliders are only in the Pro edition)
- Change the lyrics/phonemes to use less clear enunciation. In many cases this will mean removing trailing consonants, for example how the word “something” can be spoken with varying levels of clarity while still being understood in everyday speech (ie “somethin”, “sumthin”, “summin”) or how many people will shorten “and” to “an’” when speaking quickly.
- As an extension of point 3, when people speak English lazily their vowel pronunciation tends to lean towards schwa, regardless of which vowel sound is actually in the word. Of course you can take this too far and your lyrics will become unintelligible, but you can try changing vowels that sound too clear to the
axphoneme to get this “lazy”-sounding pronunciation.
「いいね！」 2
It can seem like he’s reading the lyrics for the fist time as does Eleanor. I’ve likened her to a english teacher from middle school.
I spell things wrong to get Kevin to “relax dude” such as.
Whiffle Balls I changed to whiffo bows and that sounds more rock and roll western lazy. Also missing sounds almost like, y’all need to listen to the radio bro, so I spell it missin.
And is pronounced oddly so I use anne and that works a treat.
Someone covered More than a feeling with Kevin and left it feeling which sound positively cringe worthy! 
Contrast this to Solaria who “gets it” and doesn’t sing words like a dork who has no clue, I thought it was the new AI stuff but bought Kevin and oh here we go again.
「いいね！」 1