I did some live recording this week and have one Sonar project file that has 4 songs on it.  How do I separate them into individual files so I can work on them?  Each song is 7 tracks.
 
I can easily do it, if it was only one track but with 7?

Hi. If you want to have them in separate projects, I would probably make splits between the songs.
Then I would delete all but the 1st song, and do a save as "song name", ctrl+z (undo).
Delete all but the 2nd song, move 2nd song to beginning, and save as "song name". Undo until all songs are back.
Repeat for songs 3 and 4.
 
If you want to have them in a single project, simply split, and drag them to new tracks.
Hope it helps.
All the best.

Or, save as, 4 times to 4 differing file names.
Then cut the songs out of the new files that don't belong there using ripple editing to auto move the
song desired to the front of each file.

Done.  Cut three of the 4 songs, saved that as a separate song project.  Reloaded the original file, cut 3 different songs and saved that.  Did the same two more times.  I now have 4 separate songs (projects).   Now I can edit each one individually and mixdown.   I wanted separate songs for the CD (actually 2 CD's to get all the songs)  that I'll make of the live recordings.  
 
The rest of the live recording was done with separate projects for each song.  

One thing to keep in mind with these projects is you will probably treat each song the same for processing. 
As in you get Bass track on song 1 sounding perfect so now you'll apply this to all 4 songs. If you seperate the songs all you can do now Is save pre sets or use track templates and have to drag the audio around. So there's a benifit in keeping them as one project as far as that aspect goes.  Once song 1 sounds good, the rest will be good too. 
It's why some of us long for the ability to overlay a "mixer" snap shot like when you use an outboard one. 
 
I have done a lot of live band recordings and find I will record a whole set as one project and then treat it globally. 
Only differances from song to song might be swapped out guitars and the keyboard patches ( I record this to midi) 
The guitars I will cut paste to a different track if it needs something different in the bin. So I might end up with making the original one guitar track into a few and use automation to sort that out. 
But drums, bass and vocals pretty much stay static. 
So I'll export as one big long song and then top and tail into the 12 or so songs in Wave Lab. This for me speeds things up. 

I originally did the "global" Volume automation, EQ for the bass and some EQ for the vocal.  I then mixed down to one big track.   But after I did that I found I had to go back and fix the vocal at one point in one song and the bass on another track needed different EQ.  Getting them down to individual songs allowed processing (doing whatever was needed) individually to the songs. e.g. I used POD Farm 2 Bass modeling for the bass but used different models for different songs.