I just spent more than 100 (!) credits to produce just a first verse and a short musical theme after it. During all these attempts, I was not even able to produce the chorus that would sound natural and as desired.
I absolutely require ability to specify emotions and intonations for the vocals. FOr example:
{loosing hope}What are we do it all for?
{sad}{despair}I can not stand anymore...{pause}
{sad}{accepting}I'm just a part of the Void,
{intonation down}With own self destroyed.
Or, even better: it would be absolutely perfect to be able to hum/sing the melody+intonation+emotions as a reference and then to generate a voice basing on this kind of template. And, surely, an ability to be able to correct some of the intonations/emotions of already generated vocals - just to correct, without regenerating the entire vocals from scratch - is an absolute must-have.
The similar ability to be able to hum/sing the melody would be perfect for music as well - e.g. to be able to hum/sing something, specifying: this will be a lead guitar melody.
Finally, but still extremely important: I did not find ability to explicitly specify that chorus (refrain) must contain a background melody/solo. I mean, Udio tends to generate simple rhythm/basic melody during the chorus/refrain, while I want a solo-like melody during the chorus/refrain to explicitly distinguish the chorus/refrain from verses.
Without these changes, I'm afraid I will not be able to continue to generate songs with Udio and thus will have to cancel my paid subscription.