This here idea of overlaying mixes in separate projects was just brought up in another thread. The OP wanted to know if you could simply save the settings of the (mixer) console. 
A few replies wondered why you needed this. The thinking was every song is entirely different.. Uh , uh, nope somtimes it's the same..
 
Here's the main situation where this would be real handy dandy... 
 
A session where a band records a bunch of tunes which are all more or less with the same mikes and levels. Once you've treated song # one, The tweaks will more or less be applicable to the rest of the songs. This is way true for a bed track with real drums and a dozen tracks involved for just those drums.
 
Sonar really doesn't have a straight forward way to deal with this. Obviously. 
 
You spend hours perfecting the drum mix and now are faced with 12 more songs?? Same drums, same mikes. see what we is sayin'. 
Would be so cool to save the mix. Open the next project and apply the mix to the console. Done deal. Now that's workflow....
 
If you ever owned a digital mixer you would understand what we are talking about. 
Instant scene recall of a mix, including all effects, EQ and Dynamics. 

Actually I think there are two separate issues being discussed which appear similar, but would require different implementations.
 
1) Working within a project and being able to try out different mixes within that project, like different levels of tracks, different solos and mutes, vocals up or down, remixes, etc. (Of the two options, this is the feature that I would be able to use the most.)
 
2) Being able to apply one project's mixer settings to a separate project.
 
Bearing in mind that I know nothing about coding, it seems (2) would be difficult to do unless all projects had the same basic "structure" of channels, processors, etc. Otherwise a project might try to apply parameters to something that doesn't exist. I suppose the copy could just ignore a target destination it doesn't see, but the setups would still have to be quite similar for this option to be of any benefit.
 
The reason mix recalls could transfer from one project to another with hardware is that the the setup was fixed. You always had a certain number of channels, buses, etc. The processors were always in the same place on the channel strip. With a virtual mixer, the setup is anything but fixed.
 
It seems the best workaround for (2) would be to create track templates that you could load into a project. For example, if you have 12 mics set up on a drum kit that doesn't vary, once you get that tweaked up exactly as you wanted you would save it without the data as a template to bring into your next project. If you had to apply those settings after the fact to a project where 12 mics had already been set up and did their recording. then the workaround (although I guess it's not too painful) would be to export the 12 tracks, bring in the track template, then import the 12 tracks. With the "export tracks as individual files" option, that wouldn't be too difficult or time-consuming.

They are different ideas but with the same requirement of a mixer that saves a scene. 
 
For your needs this would still work the same. 
 
Once again I'll use the way a digital mixer works to demonstrate how it does this perfectly fine. 
 
You get a song very close and you save the mix and name it. Save it in scene # 30
You then make a few changes and you save this to scene #31 and so on. 
You bring the band into the control room and flip through a couple of mixes and agree on the best one. This is pretty standard fare for the last 15 years. 
 
You are correct that this does not work if the second and third songs have added tracks. It would confuse the scene recall. So this function would only be used when dealing with 12 songs that started from the same template and have not been altered yet. Later once this is a done deal you can easily start adding stuff to each song and make the changes needed to effects etc. 
 
Example: 
Make a template and pre determine most of the goodies you will need for each track and add the sub mix / output buses etc. Even determine a bunch of overdub tracks that will be used later. ( my preferred method)  Use this template to record the 12 songs at that session. 
Now if that same scene save/ recall function is applied before any new tracks are inserted it will work. 
After all we are takin' the bed tracks which I like to nail down before the fancy stuff starts to happen. 
Anyhow, worth a ponder, I need another banana... 
 

I do like that idea but the problem is mixing in a DAW is done with so much automation and so many different plugins, a new song version just makes sense. It would be pretty much impossible to decide which parts are part of a mix and which are not. Different volumes, sure, but now I need EQ. A different send level for the vocal reverb. A different reverb, which requires different automation. I want to compare a drum mix with vastly different plugin chains. Etc etc.

Read what da monkey said, We are not talking about per song type projects which is I guess what da the majority of folks here is doin'.
We are talking about the treatment of a group of songs which all share a common bed track. 
We are talking about recording real musicians, real drums. Remember! That was what we did..
And when digital mixers arrived we could now save mixes for later recall instead of taking a picture of the desk and pieces of masking tape stuck on the side with channel assignments. 
 
Oh! and we could also use an Atari to automate the mix. OOk. 
 
Example of a common bed track template from a basic real musician type session: 
 
1- Kick
2- Snare
3- Hats
4--5--6 Toms 
7-8 overheads 
9- Midi track for later kick drum replacement 
10- Bass Di 
11- Bass overdub for later or quick fixes during session. 
12- Guitar from session- may or may not be kept. 
13- Vox scratch track
14-Vox overdubs for later. 
15-MIDI for keyboards 
16-20 Blank audio tracks for guitar overdubs. 
 
 you start with having your common go to effects and soft synths inserted in the template. 
You get the first song tweaked with eq, dynamics, panning and balance, save the mix and then be able to overlay this mix to the other songs some how another. It would be like a global track template I guess..
 This just saved you a lot of fussy time, that's all. We can do this same thing using track templates and saving pre sets in all effects, but with 12 songs this take forever and you could miss something.
Mr Anderton's idea of deleting song one tracks and dropping in the tracks from song two would be faster, but that sort of scares me for some reason timing wise..
 

I understand but again, this really wouldn't work. You can already easily drag audio clips from one project to another, if you want to keep the exact same order of tracks and the exact same mix settings. Just grab the audio from the "unmixed" project and drop it into the "mixed" project later in the time line.

The problem described here I think is not so much having the same mix for multiple songs but having multiple mixes for one song, and I maintain that too much goes into that to expect a simple "A/B" procedure from Sonar. Just save different versions of the song, and have them open side by side if you must.

FWIW, I mostly do per song stuff but I've done a bunch of live band recordings and while you might expect to be able to use just one mix, but really that's just never the case. Inevitably, the drummer will have hit his kick harder near the end of the session, or the mic moved a little, or the guitars and keyboard sounds are completely different, etc etc. I do usually stick with the "single project" layout but am always staggered by the amount of automation and track splitting necessary to get a good sound on every song.

EDIT: actually, you're right, I'm not responding to the OP so much as to where the discussion went after that. For the OP I would say the easiest way is to drag the audio into an existing mixed project.

sock monkey
Mr Anderton's idea of deleting song one tracks and dropping in the tracks from song two would be faster, but that sort of scares me for some reason timing wise..

 
If you export audio as individual complete tracks, they will all start at the same time (i.e., the beginning of the song). In addition to project backups, I also back up all my tracks as audio (with automation, effects, etc.) using this method. I can then drag them into an empty project and re-create the song exactly how it played back from within the original project. In fact I can drag them into any DAW, and they play back exactly how they played back in Sonar.
 
Also, you can export broadcast wav files that are time-stamped and know where to go. Just make sure that the BWF import and export functions are checked under Preferences > File > Audio Data.

Anderton

sock monkey
Mr Anderton's idea of deleting song one tracks and dropping in the tracks from song two would be faster, but that sort of scares me for some reason timing wise..

 
If you export tracks individually, they will all start at the same time.
 
Also, you can create broadcast wav files that are time-stamped and know where to go.

This has been part of my work flow for years. I track to an Alesis HD24 XR. At mix time I open a project template and import the files. The files line up fine provided I import them at the same start time. I like the idea of mix scenes. It would come in handy if you track a live show and want to mix each song in a separate project. Lots of good ideas and feature requests in this thread.
 
Rocky

 "Just save different versions of the song, and have them open side by side if you must."
 
No this would not be what we are after. The idea is to switch mixes while the song is playing. I guess you've never used a digital mixer. You would see what I'm getting at. 
 
Actually I think a full on multi channel pro tools rig works this way. Sonar is a great DAW for songwritting but lacks to many features for full on studio tracking. Like internal wave editing. 

sock monkey
Sonar is a great DAW for songwritting but lacks to many features for full on studio tracking. Like internal wave editing. 

I'm curious what you find you can't do. I do most of my wave editing in Sonar, although to do restoration like crackle fixes I need to use third-party plug-ins. But for gain changes, normalization, editing out clicks by zooming down to single samples, format conversion, etc. Sonar's DSP does the job. 
 
With 8.5 and its customizable menus and such, I did a version of Sonar that was just like Sound Forge - even had the menus in the same places and the same layout. I kind of did it for a lark, but then I just ended up using it