For me, unfortunately, this track has definitely proved that the ability for Suno AI to copy the original melodies from an acapella clip of me singing the song to prompt the AI has been killed off. In fact, it seems to generate a song with the flatest melodies possible. Maybe to put off people trying to copy "top" singers. But for me this is a disaster. I can still generate songs from my lyrics but not the original melodies.
I gave up in the end trying to get a decent track from my acapella audio input, as I had been doing for 5/6 months since signing up to SUNO. In the end, I cleared and trashed everything I had been doing, and started afresh. This gave me a perfectly acceptable song from the lyrics and even had some similarity to the original melodies, but similar - not the same. This often happens with some lyrics because ot the way the voice would naturall rise and fall when speakin the words. Strange but true, no doubt the way the AI learns how to use speech?
Anyway, I did use those first alternatives to edit the prompts and quite quickly arrived at this version which, I do like, hope you like it too.