Synthesizer V has been gaining acceptance among media composers for a few years, and if you really can't sing it'll do usable guide tracks you can send to human singers. It seems well ahead of Vocaloid at this point. If you release tracks with it doing the lead vocals, though, you'll likely be appealing only to the singing synth fanbase, though there are exceptions.
If you can sing (at least with help from Melodyne etc.), there's also a whole "you sing it and the software replaces your voice timbre with somebody else's" thing. DiffSinger seems to be currently popular with people making voices for it, but I haven't looked at it closely.
There's even a free female death metal vocal out there, using older tech. That particular one is largely my fault. And, if you need robot voices that actually emulates old speech synthesizers and is not just a human throuh a vocoder, Chipspeech emulates ones from 1938 to the 90s. Tangent: 1938 was not the beginning of speech synthesis. Speech synthesis goes back to the late eighteenth century.
If you can sing (at least with help from Melodyne etc.), there's also a whole "you sing it and the software replaces your voice timbre with somebody else's" thing. DiffSinger seems to be currently popular with people making voices for it, but I haven't looked at it closely.
There's even a free female death metal vocal out there, using older tech. That particular one is largely my fault. And, if you need robot voices that actually emulates old speech synthesizers and is not just a human throuh a vocoder, Chipspeech emulates ones from 1938 to the 90s. Tangent: 1938 was not the beginning of speech synthesis. Speech synthesis goes back to the late eighteenth century.
