Vocal Doubling Technique Question

Hello Group--
 
I'm about to use the VocalSync feature for the first time with two passes of the same vocal part to get a doubling effect. I am planning on one vocal being more out front than the layered part. The main vocal I'll call take A, and the layered double take I'll call take B, which will be more subtle as far as level is concerned.
 
I have a question about also applying compression. In order for the lead vocal to sit well in the mix, and all the lyrics to be clear, I need to use some compression. The question is, should I apply compression on both takes of the vocals I'm going to double, or just the main "A" take?
 
Also, at what stage do you apply the compression? I'm guessing the Vocalsync will be more accurate after compression, however, if there are any artifacts added from the Vocalsync, the compression may exacerbate those.
 
How do you use Vocalsync and compression?
 
BTW, the dynamics of my vocal takes are not too dramatic. Just a little compression should do. Also, the current syncing isn't too off either. Maybe a phrase here and there may need a little bump.
 
Thanks!!

bitflipper
Do the alignment first, along with any pitch correction if you're going to do that too. Then bounce the clip(s), apply EQ and finally compression. Compress the "B" track, too. It may not need as high a compression ratio as the main vocal, or it might need more; listen to the blend between the two tracks to decide. But it should be compressed. 
 
I'd also suggest using the lightest touch on the vocal alignment you can get away with. It's possible to get them too closely in sync, in which case you lose the thickening effect and can even thin the vocal due to phase cancellation. The opposite of what you're going for!
 
Another tip: roll off the extreme highs and lows on the "B" track. That'll help with the blend.

Thanks BF

konradh
You may get different answers because the intent of doubling can be to make it sound like two people singing, or to make it sound like one person singing but with more depth.
 
For case 1 (classic doubling), I don't like to roll off too much high end because the timing differences between the high sounds give the feeling of two people.  Of course, that is ia stylistic decision on my part.  It sounds like you are after case 2.

 
Thanks. At certain points of the track I want the vocals thickened, but I don't want to use a delay effect. I've used vocal doubling before, but just haven't used the VocalSync app to line up the parts. Since my parts aren't largely off to begin with, if the VocalSync adds any noticeable artifacts, I'm not going to use it at all.

Good point by Konrad. Sometimes you don't want the double-tracking to be obvious. Sometimes you do.
 
Compare, say, classic Jethro Tull vocals (no attempt to hide the double-tracking) to classic Pink Floyd vocals (double-tracking so tight you're not even aware they did it). There are no liner notes for DSotM or Aqualung to indicate the settings they used with their digital alignment software.