Here's what you have to be wary of: falling into the trap of assuming any one of your reference playback systems and spaces is the one that sounds "right", and attempting to correct your mix to suit it.
When I got my first set of proper studio monitors, I expected them to fix everything. I was unhappy with how my mixes sounded in the car, which was equipped with what I'd been led to believe was a very good system. I also had a decent stereo in my TV room, but didn't like its sound, either. I'd also tested on friends' stereos and in one semi-pro studio - none of which sounded like my home studio.
My epiphany came when I realized that each reference system sounded bad
in a different way. The car was boomy, my rec room system was thin, and my friends' stereos all exhibited varying degrees of too much bass.
That's when it hit me that no alternate playback system could be trusted to establish a baseline reference. I was going to have to try and achieve that in my studio, to somehow create an environment that was honest and middle-of-the-road, and then EQ my mixes for the statistical center. And to accept the reality that my mixes would never sound as good anywhere else.
Around that time I read Bob Katz's book,
Mastering Audio, which argued for exactly such an approach. I also read the
Master Handbook of Acoustics, which explained why having good speakers doesn't guarantee accuracy. I learned how to analyze my room so that I'd at least have a picture of how it was lying to me.
Slowly, the puzzle pieces fell into place. The first step was rearranging my room to mitigate resonances, followed by the addition of acoustical absorbers. Later, I moved my operation out to the garage where I could leverage the larger space and install more extensive acoustic treatments. The bigger room meant fewer problem resonances, and my speakers could be positioned far from walls.
But the single most effective action was to assemble a collection of well-made commercial recordings to use as audio references. I analyzed them for spectral and dynamic characteristics, and - most important - sat and listened to them for hours on my monitors. Voxengo SPAN was an indispensable tool, as was Adobe Audition. You can't do it by visual aids alone, but you can at least establish an acceptable range and know when you've gone outside it.
I'd read an excellent book called
Sound Reproduction by Floyd Toole, which I highly recommend. It talks about the link between acoustics and perception, how our ears naturally compensate for aural flaws, and how your brain can train itself to recognize what a good mix sounds like regardless of the environment. Trippy stuff, but grounded in science and it really works.
Nowadays, I rarely look at spectral graphs. Not because they aren't still useful, but simply because I no longer need to rely on them. Ten years listening to the same speakers in the same room has trained my brain to know when it's right or not right. SPAN still resides on every master bus, but it's mainly there to show me the ultra-low frequencies that I can neither hear nor reproduce.