So I'm attempting to do my first mix ever. I have a couple of questions. I'm using all synths for my instruments. As far as setting the initial levels to begin balancing the mix do I set the initial levels through the volumes in the synth (Ass opposed to using the faders)? I read in an article that stated I should start by using the volume in my synths to set the levels for the faders. I've also heard that I should initially try to make sure the levels from my synths max out at -12dB? Then I should begin balancing my mix by using the fader to set individual levels for each instrument? Also since I'm using a lot of synths mainly for pads and leads) how do I determine if I should leave them mono or stereo?

My advice? Run the outputs of your synths at a reasonably healthy level. What's healthy? I'd say keep the volumes as they are on the original patch unless they are obviously distorting.
 
Next I would then adjust the levels coming into Sonar by using the trim controls at the top of each channel so that nothing peaks above about -12dB to -9dB. This assumes all of your track faders are set to 0.
 
Now you're in a good position to mix the project properly and having all your faders at 0dB as a starting point means you can fine tune the levels of individual tracks with the fader where they have their greatest resolution/sensitivity (due to the logarithmic nature of the Decibel scale)
 

If you use vst synths then with the trim knob on sonar make sure you seperate the instrument track. While in a instrument track the trim affects the midi velocity values. When separated the trim on the audio track affects the gain. If you keep the tracks together as an instrument track adjust the volume on the vst

Synths have a lot of cool things in them. Effects that change the output.

Syntorial is a cool program for learning how to get the synth sound you want.

Unless the output of the synth comes before a master limiter there is no difference between an output gain on the synth and the volume of the track. Lots of synths do have a final limiting stage.

But be careful of using velocity as a volume control. Velocity often changes the timber and sound of a patch, 
I just use the midi tracks level. Never seen a reason to do otherwise. But then I always have a track for each instrument for this reason. Not a fan of having an instrument track next to the actual soft synth. I like to change which synth I'm using so much easier to work with a track for each part. 
If your using a lot of midi tracks then I would use a sub mix buss for groups of similar sounds. 

If recording the synth's analog output, I prefer to leave the DAW's faders at unity and adjust the synthesizer's volume control to set recording level.
 
However, if the synth features s/pdif output, that's better because it's normally a straight digital-to-digital transfer, bypassing the DAC in your synth. The downside to s/pdif is twofold. First, you may not be able to set output levels at the instrument and might subsequently need to lower or boost the recorded audio in SONAR. Second, you have to switch your interface's clock source before recording via s/pdif; forgetting to do that, or forgetting to switch it back to "internal" will result in clicks in the audio.
 
Of course, all that's moot if we're talking classic analog synths. Some of the vintage machines have issues with hum, which might require setting their outputs higher than the normal line level. I'd recommend getting a DI box for recording analog synths.
 
As for mono versus stereo, it's generally a good idea to favor mono tracks. You'd intuitively assume that stereo tracks result in a better stereo mix, but it doesn't work that way. When you mix many stereo tracks, the end result is unsatisfying mush, panoramically-speaking.

bitflipper
If recording the synth's analog output, I prefer to leave the DAW's faders at unity and adjust the synthesizer's volume control to set recording level.
 
However, if the synth features s/pdif output, that's better because it's normally a straight digital-to-digital transfer, bypassing the DAC in your synth. The downside to s/pdif is twofold. First, you may not be able to set output levels at the instrument and might subsequently need to lower or boost the recorded audio in SONAR. Second, you have to switch your interface's clock source before recording via s/pdif; forgetting to do that, or forgetting to switch it back to "internal" will result in clicks in the audio.
 
Of course, all that's moot if we're talking classic analog synths. Some of the vintage machines have issues with hum, which might require setting their outputs higher than the normal line level. I'd recommend getting a DI box for recording analog synths.
 
As for mono versus stereo, it's generally a good idea to favor mono tracks. You'd intuitively assume that stereo tracks result in a better stereo mix, but it doesn't work that way. When you mix many stereo tracks, the end result is unsatisfying mush, panoramically-speaking.

Not to hijack the thread, but I posted in the hardware forum a few days ago about my new Scarlett 18i20 (used to have a Saffire Pro 40) and one of the issues I had with the old interface was exactly what you said: clicks when running my Eleven Rack through SPDIF. When I changed the interface's clock source from "Internal" to "SPDIF" everything worked like a charm. However, now that I'm on a USB interface, if I change the clock source to SPDIF, SONAR no longer recognizes the Scarlett as my audio device, only the Eleven Rack.  (I have the Eleven Rack connected via USB so I can use the Eleven Rack Editor software.) My guess is that because both devices have installed ASIO drivers and both are connected via USB, SONAR is going to "see" whichever one is set as the master clock source.
 
So right now, if I want to keep using the Eleven Rack Editor (requires USB connection), it seems the only option is to keep the Scarlett clock set to "Internal" and hope I don't get any pops/clicks when using the Eleven Rack. That being the case, thus far I haven't noticed any clicks...is it just a matter of time before I do? :)