I can’t tell for sure if a particular vocal has been processed, no. Not from within a semitone anyway. The tone of voices varies so much anyway.
Usually, autotune even used quite carefully gives itself away by flattening the little rises and swoops of an unprocessed voice, vibrato tends to sounds more like tremolo (volume not pitch variable), but since these are individual traits, I couldn’t say with any certainty. I have heard people mimic extreme autotuning quite cleverly.
Variaudio used carefully is very good. I abide by a few simple rules though.
1 never shift any ‘s’ elements as they become ‘shushy’ - they don’t contain pitch info to any degree anyway
2 always shift individual notes by hand, don’t highlight loads of notes and bring closer to mean tuning it’s too crude
3 experiment with holding SHIFT moving a block - it defeats the snap
3a while holding SHIFT try pulling at the corner of the note - it ‘warps’ the note - that’s wild!
4 leave notes that ‘warp’ alone where possible or at least find a mean tuning which sounds natural
5 split audio where two notes are contained within one box and treat the notes separately
6 split audio of a song into shortish sections and bounce - variaudio becomes sluggish on my system when processing long files. Doesn’t affect quality, but it’s frustrating
7 rely on ears as final judge - never the screen. Turn the screen off and listen - I think it changes the way I perceive what I am doing
I could waffle on, but you get the idea? Hope this helps.