Don't worry too much about normalisation, it turns up the noise but also turns up everything else by exactly the same amount! The signal to noise ratio is untiouched and THAT is what matters.
That said, there is no need to normalise either before or after limiting/compression, it will do nothing useful. In fact normalisation has very little use in mixing as well, its a simple yet specialised process that is only of use in the occasional situation, if you've done enough mixes you'll know when
I'd agree with get all the waveforms in the same project, listen to and even look at them to see whats going on and adjust accordingly. Compression is noce because you normally have a bit more control over what it does, followed by maybe a little limiting, or maybe not, or maybe neither etc etc. It all depends on context!
limiting is just compression with a high ratio, it is a big part of mastering but should be treated with respect, it can squash the life out of a mix real quick. Compression as well but it will move around with the dynamics a bit more, or at least allows you to tailor them specific to the song more with the extra control's they afford.
If you are sending it to mastering then leave off ALL bus processing, no compression etc. Its worth noting that in a good studio a good engineer will very often compress the master bus BEFORE sending to mastering to get the track dynamics how THEY want it, leaving the mastering engineer with less to do. However, this is only to be done if you REALLY know what you are doing, and a good mastering engineer will soon start asking questions about what you were thinking if you do it badly!
In future, it can help to have a good meter plug in on the pro tools master bus (or use the desk meters if mixing outboard). During the mix you can have a target range, and keep referenceing it throughout the mix to make sure you are not getting carried away. The technique takes time to learn, many of your first mixes will soon end up going way above the meters and you'll have to restructure everything halfway into the mix! But get it down and you will soon learn healthy, reserved mixing practice.
Some fancy meters like the waves Dorrough show average and peak levels at the same time, so you really know what you are aiming for in a mix. Mix like that and all your tracks will come out sounding at the same level, safely away from clipping and compressed the right amount as well (if you REALLY get the hang of it). The K-Sytem expands on this, but is a little more in depth to set up...
Finally, to the OP, a track has an rms rating, which is kind of like average level. Its not one value for the whole track, it will change on quiet sections etc but you can tell a lot by what sort of RMS your track is hitting at the loudest point. -14Db RMS is healthy but loud. -18 Db RMS is nicely dynamic but quieter, -10 RMS is a smashed rock mix and -6 Db is a completely obliterated pop mix, and to my ears sounds HORRIFIC but some pop music goes this far or further in a vain attempt to get the listeners attention! many good metering plugins will display the rms.
+1 for referenceing other tracks and eqing. Its an easy way to get masters sounding better BUT you want to aim to learn what you have to do to make your tracks balance with your favourite recordings and next time acheive that in the mix instead. Thats the way to a great sounding recording. Its the best compliment for a mix engineer for a good mastering engineer to say 'I didn't have to do ANYTHING with the mixes except get it to the right level!'
EDIT - damit, ended up with another essay post! Oh well, hope some of its useful to someone!
