I typically do my own Mastering (and please, no lectures on why I shouldn't ~ I'm also a Junior Mastering Engineer; I know why). What I typically do, since I'm aware of the tools and methods I'll often use in Mastering later, is add one or two of the plug-in models to my Master buss in Sonar. I will tweak those in a "nominal" fashion, to emulate how the final Master might translate through my Mains (speakers). If something's not working, I disable them and go back to the mix. This
really saves time before sending a mix to another Mastering engineer, when I choose to. My own feedback loop saves money, too.
With that said, I recommend only doing this above when mixing. When you feel your mix is ready,
disable those Mastering plugs and export the Entire Mix. Either re-import them into a new stereo track project strictly for Mastering, use another program designed for Mastering, or send the tracks off to the engineer from there. No changing bit-rates, please, unless you're Mastering for yourself!