Our church choir has shrunk to basically 4 parts SAB and I sing tenor as much as I can while having to play and direct - frankly this is frustrating and I'm at the end of my rope trying to find music that "fits".
 
I'm toying with the idea of having the singers record their parts to a recorded multi sound track since that would have to be the same.  I figured I could take care of pitch correction, etc.  and I'm not sure when a melodyne upgrade would do in the future if I had them sing alto, soprano, etc. at a comfortable "key range" and then transposed the parts back and used melodyne to correct pitch, etc.
 
as of now i only have the stock version that came with Producer and have not really had the need to use that (yet)
 
Would I need to record each voice separately?  I do have enough SM 58's and to cover 4 singers and I have a 4 track fostex digital recorder so i could get 4 wave files.  
Whats the best way to have them follow the accompaniment?  and how far does isolation play into this.
 
I want to be very careful as these are "older singers" and you can imagine how far out of their comfort zone I risk putting them.
 
As a group,  they sing and blend well and diction and vowel production is good. but there are no "solo voices"  ( which is almost a good thing with a volunteer choir)
 
I've used sonar to make sound tracks so they are comfortable with that and we have used mics and "reverb" with some more contemporary songs - even though the room is "live" and choral and organ friendly,   with using 4 mics "not too close"  and only two floor monitors at a level just enough to hear themselves....  having a good reverb mix really made things sound bigger than life.
 
Now I want to do a couple Handel pieces and an amateur quartet of 4 lone voices is pushing it.
If i could get them with an accompaniment track,  they might feel a lot more confident.
 
Any thoughts?
 
WG

So this is a quartet (four voices) and you have four inputs (mics and interface inputs)?
 
How many vocalists would you prefer?
 
Are you able to set up a headphone mixer (and would the vocalists be able to sing with playback from headphones) instead of monitors? I'm assuming that this probably is not possible but I'm asking anyway.
 
First off... the version of Melodyne included with Sonar is not capabably of polyphony detection so if you just recorded the room you could not isolate each voice and correct/pitch adjust to acheive virtual harmonies. The full version can though and you qualify for an upgrade discount.
 
Let's assume you can't upgrade for now.
 
If the vocalists are able to perform more than one part (like all the harmonies) then I'd try close micing them all singing at the same time with the singers and mics spaced far apart enough to avoid bleed but that they could still hear each other (so they can still harmonize).
 
Then play back that track (preferably through headphones but if not at LOW volume in the monitors pointed AWAY from the mics... just enough so they can hook into the pitch and timing or set up maybe ONE headphone to one vocalist who acts as the guideline to the rest).
 
In that playback the vocalists then sing the missing harmonies. So now instead of 4 vocals you have 8 all close mic'd.
 
Do that as many times as you need to get all the appropriate voices in your harmony.
 
Since it is all close mic'd you can then use the basic version of Melodyne to pitch/timing correct each individual mic'd performance as needed.
 
If using monitors the mic bleed is an issue though but that may be able to be dealt with with some creative EQing/gating/editing.
If they perform in time/on pitch, every time though the monitor sound may just blend in nicely as ambience.
 
Then using a reverb VST you can create a "virtual room mic" setup to make up for the lack of actual room mics.
 
/not a pro... just some ideas based on my minimal knowledge

Oh and of course alternatively you could just take your close mic'd tracks and use the basic version of melodyne to adjust the performances to create your desired harmonies.
 
I meant to say that.
 
You'd want to scatter/delay those harmonies a bit though so it doesn't sound all process and mechanical.

Hmmmm....    Something to think about.   you have good food for thought.    I have until Wednesday to get some things together.   I do have a headphone amp and may be able to gather a few extra headphones.
 
Will also need to lay down their parts at a set tempo to have something to build on later.l
 
Thanks for the suggestions.

I'd use one Mic close and above the singers and encourage them to sing with heads up.

I'd put another further from them to get more of the church verb. I'd mix the two together.

When I record a small choir I usually take a 4-6 mic approach.  Stereo main pair, 2 or 3 spots depending on the width of the group and if it is in a good church, a distant pair so I don't need to use reverb.  Doing this 3 tier approach can make 4 musicians sound like 12.  
 
The problem in this situation is SM58s will not cut it.  Those are live voice mics and rarely suitable to record acoustic music with any quality.  
 
If you are recording Handel or any other Baroque, classical etc music there are a few general rules to follow, regardless of the type of ensemble.
1)Work with stereo pairs of the best possible mics you can get your hands on such as condensers or good ribbon, no mono tracks except for spotting.
2)Musicians should all perform together, never at different times or isolated.
3)Record in the natural performance environment.  If you need a headphone mix of the accompaniment which was recorded separately, you're gonna need enough cans for everyone, including a mix of their voices.  
 
I know you probably want to do it yourself, but it might be a good idea to seek out an engineer or student engineer in your area who might have some good condensers and a few more tracks to work with.

Rumley's approach sounds really smart. :-)
 
I recorded a 20 member choir with 5 mics.
1 stereo pair.
2 condensors above and on opposing sides.
1 condenser in figure eight pattern the furthest away.
 
These are a couple of the recordings from that session...
http://stabilitynetwork.blob.core.windows.net/g-tunes/20141130_David_18.mp3
http://stabilitynetwork.blob.core.windows.net/g-tunes/20141130_David_20.mp3
 

I'd be very wary about recording too much room reverb, or adding any reverb at all.

If the recordings are to provide backing for live performances rather than release as recordings, then any reverb on those recordings is going to be present when they are played back, obviously. So any reverb on the recording will then be subjected to the natural reverb of the performance room in the same way that the live voices are. With the result that the reverb will build up on the recorded voices in a different way to the live voices, unless a similar reverb effect can be added to the live to that already present on the recording. In a very live room things could get a bit out of control and intelligibility suffer. Probably wouldn't be a very natural sound either.

In my opinion/experience reverb on tracks used as live backing is best avoided if you want an natural sound. Reverb applied as a deliberate effect is a different matter of course. Dry recordings can always have reverb added, removing recorded reverb can be impossible.

As well as pitch-shifting the backing tracks, it might be worth simply trying eqing them so they don't match the voice that recorded them. Or trying a wet-only chorus with low depth and a slow lfo speed, or something like Waves Reel ADT which as well as the Beatle's fake double tracking can also do subtle very well and is a true "through zero" process that can shift the "doubled" sound ahead of as well as behind the original. Waves Doubler is cheaper and can do a similar trick but is far less subtle for some reason.

Finally, if possible keep the tracks as individual ones rather than mixing to stereo/mono and play them back from a DAW in performance. Otherwise you might well find different performance spaces need different treatment of each voice, but your backing is fixed and can't easily be eq'd or balanced to suit.

Rumleymusic and GsWitz,    I appreciate your time and knowledge for recording the choir.
 
I wish i had a choir to actually record "live".   The purpose of this "experiment" is.. ( since there are only 4 voices) to try and record them singing their parts ( perhaps women sing the alto parts and then soprano parts  - even if i have to transpose the parts - and use melodyne to transpose the parts back up ( or down ) ( assuming that's possible -- so I have gathered ).
 
Then we would sing to the recorded sound track just to get the job done.  Frankly,  it's that or nothing at this point.
 
I know it's a long shot and now's the time to find out if it will work or not. 

This may be off topic, but when my daughter learns choir parts, I make a midi version of the song and bounce it with her part pronounced and loud so she can practice singing the right tones. You can easily bounce a different version for each member and record them singing along to the midi parts. Then drop out the midi stuff and you're left with a fun recording.