One of the hardest issues I find with using Suno, is that there are times when I generate lots of iterations, sometimes changing the prompt slightly sometimes not. I usually try to keep down the number of versions I actually keep on file rather than trashing. Keeping a file system as the number of songs, versions and iterations increases can be a real drag. Yes, I know that downloaded songs can be filed on my lappie in folders, etc, but it is a drag.
This was one of those songs. I started off, loading an audio clip singing the song, acapella. But after a few iterations I could not understand why the AI was having trouble with the melodies. Until I listened carefully to the audio clip. NOW, I had originally prompted for "traditional chinese style music". and It was so long since I had played/sung this song, that I had miss-pitched notes and considering chinese music sounds chinese is because it is baasically similar to the western pentatonic. I think this was causing problems I think. Any way I loved the way my voice had been cloned so far, I picked the "best track" and decided to use it for this Soundcloud Upload. I can always practice the song and try again and hopefully, when a better version comes along I can replace the track.
This is getting to be a long story, but I will try and finish quickly. I now have another four versions which I find it impossible to "trash". I reconsidered and prompted for a female vocalist and simplified the prompt. I will be posting one of these as tomorrows upload. Just need to decide which one!
Sorry this sounds a little long, but couldn't delete any of it, I need to write more lyrics before I try another version with my own cloned voice. SUNO has tried to keep it intresting however. How you like it anyway.