There is an old trick to help to "tune" vocals. You replicate your vocal
track twice. In one of the copies you add a LADSPA Rubberband pitchshift
with +5% and to the other you add the same plugin with -5%. Mix these
untunned channels with a low volume, like 20% of the original one.
How this workaround help? Once that you try to reach a note, you have 3
chances to it. Our earing will listen only the right one. Normally 5% of
pitchshifting is enought to correct small mistakes. Take care to not put
the replicas too loud.