Board Thread:General Discussion/@comment-53539-20180529221920/@comment-110.184.47.217-20180531082345

Perhaps I'm cherrypicking here, but I think it is highly dependant on the Vocaloid user. Look at kyaami/cillia for example. All of the difficulties that have been previously pointed out in this thread (limited emotional expression, sudden shifts in tone that don't quite natural, etc.) were no obstacle to her at all. Tons of her work is very realistic-sounding; I distinctly remember hearing her VY1 cover of "Donor Song" and being shocked to find that it wasn't a real person singing, and the same thing with her "Rolling Girl" UTAU cover.

Also, I believe I once read a comment on the official Vocaloid YT channel asking them when they thought Vocaloid technology would reach the level of sounding like actual humans. The reply was something along the lines of "we're pretty much already at that stage where it's possible to realistically emulate the human singing voice, but it's too much work and therefore not worth the effort". So, from this response, it seems to imply that as long as an individual has the sufficient skills and commitment towards a certain work, then there really isn't anything holding them back from creating an amazing, realistic song.