When I mix films in stereo, I mix for dynamic range at 78dB SPL. This means I have my setup calibrated, I mix the dialogue until it's comfortable, then I mix everything else around it to match. In some action films, the explosions peak pretty high (say, -2) while my dialogue sits lower (say, -10). But in more dialogue-based films, I don't have those transients, so my peak level of the whole mix might be at -6.
Question I have is, when all is said and done, should I compensate for the -6 peak by normalizing the level?
I hate the world "normalize" with a passion, but it does seem logical in this case. Having said that, I presume that this is what the "dialnorm" spec is for, right? Is this why we don't normalize our final film mixes?
Thanks for everything!