Hi. Wow I guess I'm new here. 
I'm using X1 with a Lynx L22 soundcard.
Overall my songs sound like they have a blanket over them.  For example, even if I record a sound completely dry, they never sound upfront when I export the same file.
I'm wondering if I'm processing my tracks correctly.  For example, I'll record a bunch of tracks down to one stereo track.  Playing that track shows peaks around -6db.  I'll then throw a limiter on the master and my peaks will hit at 0, and then I export the audio using POW3 dithering as i'm converting from 24 bit to 16 bit.  The issue is, my songs are still no where near as loud as other professional songs (even songs that aren't victims of the loudness wars).  Even if I TRY to make my song LOUD, it just becomes completely smashed before having anywhere near similar volume to other songs.  
So I'm wondering if I'm doing something wrong or maybe there is some sort of calibration I need to do in Sonar?
Or maybe someone could even point me to an outline of the process you follow for mixing down tracks to stereo then exporting those tracks, when limiting comes into play etc. I just feel like I'm missing something in this process.
Any help is appreciated. 

Having a blanket thrown over a sound I usually associate w/ losing high - muffled.  Which would keep them from being upfront, of course, but you shouldn't have that problem w/ a lynx card.
 
My first thought is that there is a gain-staging problem somewhere between your hardware to the lynx to SONAR and back.  Lynx should have its own software mixer that works in conjunction w/ SONAR's hardware in/out.  Open up SONAR > Mixer and make sure all the busses/hardware outputs are checked at the top.  Then slide open each in the console view and make sure the faders are all at 0 db.  Do the same w/ the lynx software.  Then make sure whatever preamp etc. is outputting a nice signal?  What preamp are you using if you record anything analog?  It could be a mismatch between -10 dB "prosumer" hardware and lynx, which I expect, expects to receive a +4 dB signal.
 
As far as matching pro levels - it takes more than just raw volume to do that.  Arrangement plays a bigger role than you would believe, as well as the experience to know exactly how far to push it before the sounds folds over and becomes smaller - like w/ compression.  Example, you'll never get a fuzz guitar as loud as it sounds live.  Try it, and smash the begibbers out of it and then put a bright acoustic on top and see which you hear most.
 
You will be much better off making the mix sound good, so that all the instruments can be heard if you listen for them; the leads (such as vocal to guitar to vocal again,) match in volume to the ear, not meter; the overall mix is loud but not strained and still contains some dynamics.  You want to leave enough room (peaking at -6 dB or -3) to master, which raises the overall or average volume to 0.-3 dB (you'll need a couple of dB if you are making an mp3 from the mix, and come on, you can't really hear a  point dB or three). 
 
The mastering is when you get that last burst of loudness - but you'll still probably not match a commercial release.  Don't worry.  They have things called volume knobs.  And if your song is picked up for release you can pay a mastering (and perhaps a mix) engineer to raise your level cleaner than you could.
 
@

Use your EQs. Use your compressors. Measure average loudness for the track and pick a point like -12 or -14 then make a bounce and listen.
Now double back and change the amount of base. Re-level to your standard (-12 or -14) and bounce.
 
Now use the single solo button to switch between the tracks so you can hear the differences.
 
Try to avoid compression on drums. Compress other things enough that your limiter doesn't need to limit more than 1 or 2 dB in order to get to your chosen average volume.
 
Some folks start by getting an average volume across all tracks that is roughly equal, but myself I don't find this helps as much as listening. Bleed between tracks for example and quality of musicianship varies too much to make it  super reliable. But if you are having a really hard time, it's interesting to level and give a listen.
 
Multiband compressors can be helpful too for eking out an extra dB without making things seem squashed.

Thanks @.  For a preamp, I'm using the FMR audio RNP.  It looks like the RNP automatically converts between +4 and -10 based on the what's plugged into the output.  The cables from my lynx card is TRS so I think its +4.  Right? 
Yes I agree that the song is the biggest part of the mix.  
But overall I've just felt something was still missing.  
For example, as a test, I recorded a commercially recorded song into Sonar from my mp3 player.  I had to pass the output of my recorder through my soundcard to record it.  It's a dynamic song so my input was a high as possible without clipping.
After the song was recorded, I just exported it again with everything at unity.  The resulting output file is much lower in volume than the original.  Is that to be expected?  
It seems I'm physically unable to get the recorded song to then be mixed down to the same original volume.  If I try to increase the volume during recording, it clips.  So in this example, it seems like I have to include a limiter on the recorded track to get it back to the volume of the original.  This doesn't seem right because I then lose the dynamics of the original.  I know I'm just missing a general concept here of levels.  Thanks for your help. 

You would have to have your mp3 player at full volume, not just the lynx (for recording, SONAR is out of the loop since it simply captures the digital input signal and cannot process it).  Most of them crap out at high levels.  I get much better listening results to my wife's iPhone (and my win phone) at about 3/4 vol.  That is part of the problem, along w/ the lynx expecting a +4 line signal which I doubt your mp3 outputs.
 
The RNP is good gear and should be +4.  Does the Lynx 22 have a software mixer?  I know the aurora does.  That is probably where the problem is.  To record, one usually has to dicker w/ the software mixer that controls the volume in and out of the soundcard.  So mic to preamp (controls the mic vol and converts it to a line-level signal) to Lynx hardware to Lynx software (which sets the relative strength of signal sent by Lynx HW to SONAR).  The lynx should be set to unity, that is 0 dB.   Output is SONAR digital to Lynx digital in to analog output.  Then your monitors.  Nothing in SONAR during recording has an affect upon the recorded sound - only the return output signal.  What levels do you get in SONAr tracks at unity when recording?  You should have no problem getting a good, strong signal out of SONAR to the Lynx card, so if there is a gain problem it is likely in the software mixer for it from what you are describing.
 
Of course, it could be all kinds of other problems, but that is the first place I'd look.  Another thing to do is use a softsynth or audio loop, which takes recording out of the picture.  Pull it into a track, and see if the volume responds as one would expect (ie. Sonar track levels match the output on the lynx soft mixer).  If the loop track is at 0 dB in SONAR and on the mix output but is weak and muffled when you listen, you can lay the blame outside of sonar.

I presume if you just import an MP3 to a track and export the same track that it sounds the same. True?
 
 

gswitz
I presume if you just import an MP3 to a track and export the same track that it sounds the same. True?
 
 

I believe this is true but I will have to check.  If it is true, then it sounds like the issue is me right? :)

AT
You would have to have your mp3 player at full volume, not just the lynx (for recording, SONAR is out of the loop since it simply captures the digital input signal and cannot process it).  Most of them crap out at high levels.  I get much better listening results to my wife's iPhone (and my win phone) at about 3/4 vol.  That is part of the problem, along w/ the lynx expecting a +4 line signal which I doubt your mp3 outputs.
 
The RNP is good gear and should be +4.  Does the Lynx 22 have a software mixer?  I know the aurora does.  That is probably where the problem is.  To record, one usually has to dicker w/ the software mixer that controls the volume in and out of the soundcard.  So mic to preamp (controls the mic vol and converts it to a line-level signal) to Lynx hardware to Lynx software (which sets the relative strength of signal sent by Lynx HW to SONAR).  The lynx should be set to unity, that is 0 dB.   Output is SONAR digital to Lynx digital in to analog output.  Then your monitors.  Nothing in SONAR during recording has an affect upon the recorded sound - only the return output signal.  What levels do you get in SONAr tracks at unity when recording?  You should have no problem getting a good, strong signal out of SONAR to the Lynx card, so if there is a gain problem it is likely in the software mixer for it from what you are describing.
 
Of course, it could be all kinds of other problems, but that is the first place I'd look.  Another thing to do is use a softsynth or audio loop, which takes recording out of the picture.  Pull it into a track, and see if the volume responds as one would expect (ie. Sonar track levels match the output on the lynx soft mixer).  If the loop track is at 0 dB in SONAR and on the mix output but is weak and muffled when you listen, you can lay the blame outside of sonar.

I'll check this also by recording a soft synth internally and 0 and exporting. My Lynx mixer is definitely set to unity.
 

Oh are you bouncing and then exporting through the master bus? Lots make the mistake of bouncing to an internal track and then exporting that track routed to the master bus. This double applies all the FX.

Route the bounced track instead to your lynx outputs.

gswitz
Oh are you bouncing and then exporting through the master bus? Lots make the mistake of bouncing to an internal track and then exporting that track routed to the master bus. This double applies all the FX.

Route the bounced track instead to your lynx outputs.

Yes that's exactly what I'm doing.  I'll mixdown say 20 tracks to 1 stereo track.  Then I just solo that track, apply any limiting etc, and export to an mp3.  I never touch the master and yes everything routes through the master.
So you're saying I should bounce all tracks to stereo as normal.  But then when exporting, that new stereo track should be routed directly to my sound card?  Would I then have to put any "mastering" effects on that stereo track before doing so? And why would it matter since export is not really using the sound card anyway and is just creating an mp3 file for example?
Thanks for the help.  yes this is exactly what I've felt I've been missing.  Not understanding best practice for mixdown and export and maximizing levels for each.  

Export bounces out entire mix usually ... Unless you change it. This means it applies track and related bus fx. So, if you don't route to a bus, no bus fx.