I work with Melodyne Editor (ME) in much the same way Danny does.
I apply the effect to the track in a destructive way. Since I rarely do outside client work, I don't worry about saving the original track. But when I do outside work or have a singer on my tunes, yes.... absolutely preserve that original track in a safe place.
Most times, I will let ME copy the entire track in and then I will apply the corrections I want MANUALLY after I have listened several times. No auto-correct for me. I work through it phrase by phrase and click on SAVE every few minutes. Once I'm done, I will do a playback start to finish listening for anything I might have missed. Once satisfied, I will "apply" ME to the track.
I used to try to fix everything. But now, I have changed my approach a bit and only fix the things that are obviously bad and need fixing.... well, most of the time. In some songs, I simply do "Spot fixes" to catch a stray note here or there. All fixes are tested by playback and then applied destructively.
One other thing I do, is, at the beginning, I set up vocal busses and place my reverb and compressors in the buss and NOT the track FX bin. That leaves the track FX bin open so I can drop in ME to fix something, and when I print, I don't have to remove the reverb and then put it back after. Nothing sucks like printing a reverb and realizing it much later...... it's hard to go back and fix the next day.
I don't clone or copy or use ME to do track doubling. ALWAYS take the time to record a different take and do that as many times, on as many tracks as you need. I will often do 2 good takes, for left/right panning at lower levels and similar for the harmony/BGV tracks. 5 tracks is about average in a finished project. Cloning or copying a track results in a "bit perfect" track. That can and will introduce other serious issues that you need to be aware of. Recording a different take, even though it sounds like a perfect match.... isn't. Problems can still occur but the end result is a better sounding mix due to the slight differences between the tracks. That's the magic of a choir. The different voices.
The big question you need to ask and answer is..... do those secondary & BGV/Harmony tracks even need to be pitch corrected? If they are low level tracks, panned, and the singing is as close to syllable, pitch, timing, and phrase perfect as you can get them, and they are going to be used to thicken the much higher volume lead vocal track, the answer I have come up with is often "No". The inconsistencies will add character and if a note is slightly sharp or flat, 99% of the listeners won't even hear it.
You can, and I have, run ME on each track one at a time. Fix the timing, the pitch, the vibrato issues. And of course you would certainly want to do this for tracks where the vox will be clearly audible. Such as a duet singer or a vocal group where each voice is expected to be heard. If you do take the time to pitch correct everything, be careful about getting it "too perfect".... there's a downside to going to that extreme. It starts to sound "fake" to one degree or another because we all know people don't tend to sing perfectly.
The point is, keep it sounding real. Hope this helps some.
post edited by Guitarhacker - 2015/12/24 10:16:50