Hi guys and girl(s) 
I just wanted to share my take on how I approach this, there's alot of different techniques on this and I will only go trough the one I most often use and the one I feel is the simplest one.
I made 2 examples by covering one of the most famus "Wall of voice" bands, Behemoth
Conquer All
http://dl.getdropbox.com/u/1520413/Conquer All-01.mp3
Here I really went all out on the voices just to make it very extreme sounding.
Normally I record 4 different vocal takes (2x L/R 2x Center)
One of the center ones should be pitchshifted -1 to -2 semitones
(You can make this even simpler, recording 3 tracks and put a doubbler which duplicates the center tracks sound source and then adjust the pitch and delay to your liking)
I Like to keep the pitched track with a focused lowend on it to take care of those crushing basstones and boost the normal pitched tracks abit on the highend, just EQ it so it sounds good ofcourse
On this specific song I actually used a doubbler with 3 extra sources of sound on the two center vocals (Also panned 100L + <C> + 100R)
Adding a 6 voices, making a total of 10 voices coming at you! (Remember to experiment with the detune/pitchshift and delay on the doubbler)
Here's the bus I used for the vocal takes.
https://dl.getdropbox.com/u/1520413/Conquer All Voice.JPG
Ov Fire and the Void
http://dl.getdropbox.com/u/1520413/Ov Fire and the Void.mp3
This one I went with an cleaner aproach but still having relatively crushing vocals.
I did as i explained above
4 takes 2x L/R (These will only be kicking in occasinally) and 2 Center (Remember one is pitched -1)
Even here I used a doubbler adding 2 sources of both center tracks panned L/R (barely audible, they're just there to make them sound abit bigger)
Same thing here, play with the doubblers settings to make it fit, it's hard to make a doubbler sound good. But It's way easier to make tight sounding rather than actually having to sing it that many times.
Almost the same bus settings I believe.
And remember I'm pretty new to guitar, and singing I began with this year so try not to flame my slopiness
Hope you found this usefull in any way, getting ideas etc.
The reason my EQ curves might seem strange is because this was recorded on an SM58
uke:
I'm not saying this is the way to go, only that this is how I've done it.
Feel free to discuss this entire topic and share your own experiences and tips on it, I sure would love to learn something different
Have a good one!
EDIT:
Ofcourse it's best to record each take individually and skip the doubbler
I liked this method for Behemoth's songs, would sound too natural without doubbler.
And since Im not a singer I cant record 10 takes super tight, so this will have to do
But the Idea is that you can replace the voices generated by the doubbler with actual takes to make it sound more natural.
I would still use pitch shifting for one center track to blend, Behemoth did it on Demigod and so do I.

I just wanted to share my take on how I approach this, there's alot of different techniques on this and I will only go trough the one I most often use and the one I feel is the simplest one.
I made 2 examples by covering one of the most famus "Wall of voice" bands, Behemoth

Conquer All
http://dl.getdropbox.com/u/1520413/Conquer All-01.mp3
Here I really went all out on the voices just to make it very extreme sounding.
Normally I record 4 different vocal takes (2x L/R 2x Center)
One of the center ones should be pitchshifted -1 to -2 semitones
(You can make this even simpler, recording 3 tracks and put a doubbler which duplicates the center tracks sound source and then adjust the pitch and delay to your liking)
I Like to keep the pitched track with a focused lowend on it to take care of those crushing basstones and boost the normal pitched tracks abit on the highend, just EQ it so it sounds good ofcourse

On this specific song I actually used a doubbler with 3 extra sources of sound on the two center vocals (Also panned 100L + <C> + 100R)
Adding a 6 voices, making a total of 10 voices coming at you! (Remember to experiment with the detune/pitchshift and delay on the doubbler)
Here's the bus I used for the vocal takes.
https://dl.getdropbox.com/u/1520413/Conquer All Voice.JPG
Ov Fire and the Void
http://dl.getdropbox.com/u/1520413/Ov Fire and the Void.mp3
This one I went with an cleaner aproach but still having relatively crushing vocals.
I did as i explained above
4 takes 2x L/R (These will only be kicking in occasinally) and 2 Center (Remember one is pitched -1)
Even here I used a doubbler adding 2 sources of both center tracks panned L/R (barely audible, they're just there to make them sound abit bigger)
Same thing here, play with the doubblers settings to make it fit, it's hard to make a doubbler sound good. But It's way easier to make tight sounding rather than actually having to sing it that many times.
Almost the same bus settings I believe.
And remember I'm pretty new to guitar, and singing I began with this year so try not to flame my slopiness

Hope you found this usefull in any way, getting ideas etc.
The reason my EQ curves might seem strange is because this was recorded on an SM58

I'm not saying this is the way to go, only that this is how I've done it.
Feel free to discuss this entire topic and share your own experiences and tips on it, I sure would love to learn something different

Have a good one!

EDIT:
Ofcourse it's best to record each take individually and skip the doubbler
I liked this method for Behemoth's songs, would sound too natural without doubbler.
And since Im not a singer I cant record 10 takes super tight, so this will have to do

But the Idea is that you can replace the voices generated by the doubbler with actual takes to make it sound more natural.
I would still use pitch shifting for one center track to blend, Behemoth did it on Demigod and so do I.