Sometimes AI models sound so bad that they cannot be used in some parts. This usually happens with high notes, but you can never guess which part this will happen to. Maybe, Karina's model is the best
@@lmaolm9353 and it turned out that the result doesnt work well, because this song is perfect for soothing voice like Ningning’s and Giselle’s voices, unlike karina and winter that got bold vocal
![[AI COVER] AESPA - ❝SOUR GRAPES❞ | Line Distribution](http://i.ytimg.com/vi/o275jOyUBQU/mqdefault.jpg)
![[AI COVER] AESPA - ❝SOUR GRAPES❞ | Line Distribution](/img/tr.png)
![[뮤뱅 원테이크 4K] 르세라핌 (LE SSERAFIM) 'Perfect Night' 4K Bonus Ver. @뮤직뱅크 (Music Bank) 231027](http://i.ytimg.com/vi/pXhjs-35jSI/mqdefault.jpg)
In some point, this song is so thirsty coded idk why
It is!!
Right? Like most of Karina’s parts could’ve been given to others.
You gave Karina AI too much lines. Giselle and Ningning suit this kind of song the most.
No one cares she’s racist
Sometimes AI models sound so bad that they cannot be used in some parts. This usually happens with high notes, but you can never guess which part this will happen to. Maybe, Karina's model is the best
@@lmaolm9353 and it turned out that the result doesnt work well, because this song is perfect for soothing voice like Ningning’s and Giselle’s voices, unlike karina and winter that got bold vocal
@@clerith If you don't like the result, you can write artificial intelligence from scratch to copy the voice.
she's literally 3rd 💀
Live version Karina = Dva Kiriko = Giselle Winter Sombra instead of Tracer/Brigitte you have a Mei for Ningning
How do you do ai covers?