Pretty much the same here.
Generally I EQ and do whatever to each source to make them sound good on their own first. Then balance the mix of the two. Then tweak that further. Sort of like mixing and mastering as separate processes.
Partial exception to the first step is when I know one source will be supplying what the other is not. Like plenty of low end from one source, so the other needn't be boosted there to compensate for thinness, but must still sound good otherwise.
Going further when the effort is worth it, sometimes I'll do that kind of thing in a more advanced way by playing one off the other as a further extension of balancing the mix of of the two in more ways than just level. For instance I might take both "good" EQ'd sources, make a specific cut to one and a corresponding boost to the other in an "energy-balanced" way - that is, not boosting or cutting the resulting sum response of the two so much as letting one contribute more in that area than the other.
I'll then probably still tweak the resulting mix somewhat in the final mastering step.