My opinion is, it depends. The only songs I've had to go nuts over is top 40 stuff, where effects can change from part to part. And it's not just vocals either- it can be any instrument and snare is a big one too. But it's highly dependent on the song, arrangement, and the genera. Being able to automate effects on one track can cut down on having to split everything out too. Just splitting up vocals to multiple tracks won't stress your cpu persay, unless you already have your system up against the wall with lots of effects and tracks. It's not the extra vocal takes and tracks, it's the effects on those tracks that will eat up cpu. And that's why whenever possible, it's better to use effects as an aux rather than an insert. If you use a reverb on an aux buss you can send any number of tracks to it and then automate along the way how much will be sent or not for each track. This is better than having four vocal tracks with four separate reverbs on each track [inserts]. That sometimes is exactly what has to be done though, but when you can, use one reverb [or delay, etc] on an aux buss and send all the vocal tracks to it, each one being adjusted appropriately. It'll save tons of cpu power.

For years I recorded my vocals no less than four takes, comping one part together for the final. I've gotten away from that though. I mostly did it for a safety net with pitch. If I was pitchy on one phrase, well sure enough there was a good one between the other three. And maybe even a more inspired performance to mix and match from too. I do this by default when recording other singers. I get the great take done and then have them 'throw it away' a couple more times and see what happens. With the great take done and the pressure off, sometimes a really inspired performance can pop out. But for my vocals, I've let go of multiple takes. I've also let go of auto tune- a great tool that is VERY labor intensive to use properly. What I do is record one section of the tune and stop and listen carefully. If pitch and performance is good, I'll move on. If it's not, I'll punch the fixes or redo until it's right. When the vocal is done, I have one finished track that I no longer have to futz with in post. I don't have to comp or tune or anything. It's one track ready to be mixed. It cuts down BIG time on extra work. It's called getting it right on the way in, instead of noodling with endless desicions after the fact. It's liberating and frankly, it comes out much better. The key to doing that is the headphone mix. Next time you're doing vocal, put only a smidge of verb on it for inspiration and mute every instrument track that isn't absolutely necessary. What you'll get is a much better tuned vocal right off the bat and usually a much better performance too. After that, the business of splitting out tracks for multiple effects and comping and all is just different ways of mixing. No one way is right or wrong. It's comes down to what works at the time, and I find that the best way is usually the easiest and quickest way to get where you need to go.

So like I said....it depends!

Dan

My YT Channel- http://www.youtube.com/user/dfizzbom