The meters are not accutate!

Well let's say you have an effect which has some kind of time based modulation (delay, chorus, phaser, that sort of thing). If the modulation speed or delay time is not synced with the DAW and is specified in absolute units like Hz or ms, then the resulting audio will be slightly different depending on where you start playback from. Whereas if you've specified this modulation time in musical units like 1/8, 1/6th notes etc then it is synced with the DAW and the effect will be the same every time. 
 
But if you don't have an effect like that on the track in question then I really don't know, I mean in my non-expert mind I can imagine the possibility that something with a look-ahead algorithm could give slightly different results depending on where you start playback from, maybe a tiny difference in the way the resulting math is calculated could result in a small difference in rounding which just pushes a sample over the threshold when the calculation starts from point A as opposed to point B. But then again, I have absolutely no idea what I'm talking about and should probably leave the answer to the egg heads among us....

If none of the tracks are clipping, I'd consider just pulling down the input gain on the target bus. I don't think it makes sense to have to pull down each of the tracks a bit when the bus input gain is right there as a one-fader solution.
 
I've also seen the behaviour where the meters show something different for the same audio when I start playback from different points. I'm not entirely sure why this is, but I expect it's to do with the way that loudness is integrated over time. I expect the only safe solution is to back off a dB or two to be sure.

Bristol_Jonesey
Your starting fader positions will also vary depending on the total number of tracks in your project.
 
As a rough rule of thumb, the formula is (in Excel) -20 x Log(total number of tracks)
 
So as an example, for a 15 track project, initial fader positions should be around (on average!) -23.5dB

I was told there would be no math.

Yet another reason to work with bus input gain instead, right? Fix it in one place.

Bristol_Jonesey
They lied

:-D
 
To the OP (vlad)... as has been noted multiple times in this thread there are far too many variables to expect the meters to respond EXACTLY the same way every time. That is because the audio is unlikely to output EXACTLY the same way every time except under extremely clinical scenarios where every possible variable has been eliminated.
 
By that I mean every effect, synth, EVERYTHING has been bounced to pure audio throughout the entire project so there is no possibility of variations FROM those plugs. Even then I'd still think in a complex project with many tracks the way the computer reads/rounds/processes it all might vary slightly.
 
Now if you had a single stereo wave with NO active effects and the meters still did this then sure, that would indicate maybe the meters or Sonar was doing something funky.
 
BUT, from what I can tell the track/bus meters aren't intended to be ultra hyper accurate (we can even incur that type of resolution). Just reasonably accurate guides to show what's doing what. So if you are pushing things so far that you are only leaving 0.00000001 db of headroom and the meters only respond to a resolution of 0.00001 db maybe that could account for the clip light being triggered every so often.
 
Expensive digital mastering software provides much more accurate metering resolutions if that's what you want.
 
Thing is all of this is completely moot because trying to make everything as loud as possible at all times by just turning it all up using the faders is NOT how this stuff is done.
 
General practice is to mix all your tracks and busses so your master output only reaches at most -3db. The amount varies though depending on who you ask and/or what you are trying to do. -6db to -3db seems to be about right.
 
Why? Because again, as noted above, getting those levels up to -0.1db is generally a "Mastering" process... as in you leave that to your Mastering engineer or you yourself apply it from the perspective of mastering your project either right on the Master bus or exporting your final exported mix then mastering the stereo wave in a new project.
 
There is LOTS of stuff that is either used to make up that extra volume in the mastering stage in a complimentary way (saturation, etc) or things like EQ where you simply need the extra room to boost a band if you need to.
 
BUT the most important "effect" is the limiter. You can completely not use all that other stuff if the mix is THAT good and simply raise the volume with a limiter and COMPLETELY avoid clipping.
 
So if you think your mix is absolutely perfect in every single way and that is EXACTLY how you want to release it you can put a limiter on you main output bus (Master bus) as the absolute LAST thing in the signal chain then set the output threshold to -0.1db (or even -0.0000001db if your limiter can do that... which it probably can't). That makes it so even IF for a split second something hits 0db or higher the limiter stops it from clipping the bus. Kind of like a simple overload protection.
 
Then you can turn up the volume as loud as you want and never clip. Ideally in that situation you would turn things up until the limiter just BARELY registers that it has been triggered then turn things down just enough so it doesn't activate at all (so the limiter doesn't do ANYTHING but is still there just in case something does go over).
 
Usually though you DO want it trigger at least a little (unless you have managed to mix the most perfect mix that ever done got mixed) to bring up the volume of everything below the peaks so that the song plays back nice and loud but there are no clips.
 
Some people SMASH the limiter to bring even the quietest sounds up and "squash" the dynamics to make EVERYTHING as loud as possible (which, IMO sucks... google "Loudness Wars"). They may also use other stuff before the limiter such as saturation/compression/levelers/etc like mentioned before to bring the volume up then only limit the signal a bit (or still smash it).
 
Point is... the limiter is there to stop any clipping whatsover but still allow you turn things up as much as you want. Doing this stuff like you are (just turning up the faders as loud as it will go) is unnecessary and counterproductive. Lower your levels right from your tracks through to the busses to leave headroom on your final Master bus THEN raise the volume using a limiter OR raise it with other master effects BEFORE a limiter.
 
In all cases the limiter's output is set a fraction below 0db which completely avoids all clipping while letting you to turn things up as much as you want.
 
BTW... I send EVERYTHING through a "Premaster" bus before it hits the Master bus. That way, even though I completely try to obey "Gain Staging" principals from start to finish if I somehow go over/under -3db a little I can just adjust the Premaster fader a bit (never need to except as a project grows... then I correct it in the mix). It's also a place where I can experiment with mastering effects such as EQ and Compression.
 
The ONLY thing on my Master bus is a limiter (usually the Concrete Limiter) as clip protection and it never gets triggered unless I do something stupid or I want to do an export at broadcast volume (in which case I adjust the limiter to turn up the volume so it is just barely being triggered then export).
 
I do my fake, crappy "mastering" in a new project on the stereo export (but make sure the limiter in the mix project isn't adding ANY volume at all... just operating as a clip protector... which isn't even needed because I gain staged it all to not clip anyway... usually).
 
 
There are much smarter, more experienced/concise people around here who could explain all that better (and correct any dumb crap I just fudged up).
 
tl;dr...
 
You're not gain staging properly.
 
Use Limiters to avoid clipping.
 
Cheers.