I might record some tracks (a vocalist or a choir or a string section) at a large commercial facility.  The other tracks and the final mix will come from my project studio.  Here is how I think would be best to approach this, but if you have done it a different or better way, please let me know.
 
Example using a choir:
 
1.  Mix the existing track to a wav file.  Have a known and cleanly defined starting point.
2.  Ask the studio to load the wav file and record the choir in stereo to a separate stereo track (or tracks if you want to divide the choir into sections.)
3.  Ask the studio for the additional tracks (new material) in wav format.  Ensure that these tracks start at exactly the same time as the reference track.  For example, if the choir starts singing 1 minute and 10 seconds after the reference track starts, there will me 1:10 of silence on the wav files the studio delivers.
4.  I might put an audio spike (e.g., click) in my Sonar file and on the first of the reference track, and ask the engineer to do the same on the new content.
5.  Back in my studio, I will load the new tracks into Sonar and align their starting points with the start of the track.
 
Thoughts?

You could also try using Broadcast WAV files, which have the starting time code embedded in the file. SONAR supports those.
 
Or you could even use the OMF project file format, as long as the studio supports it. SONAR does.
 
Scott

--
Scott R. Garrigus - http://garrigus.com - SONAR X3 Power! - http://garrigus.com/?SonarX3Power
* Author of the Cakewalk Sonar and Sony Sound Forge Power book series: http://garrigus.com/?PowerBooks
* Author of the Cakewalk Sonar ProAudioTutor video tutorial series: http://garrigus.com/?ProAudioTutor
* Publisher of the DigiFreq free music technology newsletter: http://digifreq.com/?DigiFreq
* Publisher of the NewTechReview free consumer technology newsletter: http://newtechreview.com/?NewTechReview

You can export OMF, but not import which may or may not be an issue. Also remember that OMF doesn't cover MIDI, virtual instruments, etc. so if you have 46 instances of Kontakt...
 
I agree the easiest approach is to record a premix that starts at the beginning of the song, have the choir sing to that, then bring the wav files (which also start at the beginning of the song) into Sonar.

Oh, that's right... you want to import back into the same project so OMF won't work, but Broadcast WAV files should work fine or regular WAVs at the beginning of the project will definitely work.
 
Scott

--
Scott R. Garrigus - http://garrigus.com - SONAR X3 Power! - http://garrigus.com/?SonarX3Power
* Author of the Cakewalk Sonar and Sony Sound Forge Power book series: http://garrigus.com/?PowerBooks
* Author of the Cakewalk Sonar ProAudioTutor video tutorial series: http://garrigus.com/?ProAudioTutor
* Publisher of the DigiFreq free music technology newsletter: http://digifreq.com/?DigiFreq
* Publisher of the NewTechReview free consumer technology newsletter: http://newtechreview.com/?NewTechReview

Thank you, gentlemen.

The only problem with broadcast wavs is if the studio uses something like 1:00:00:00 as their default "0." (This can happen, I've had it happen to me). So you give them a file that starts at 00:00:00:00 and they wonder why your file doesn't show up on the timeline, or yours does and theirs don't. "Beginning of Project" is always safest.