I'm sick of doing those basic topics, so I'm taking a break from that. I want to write about binaural mixing. First of all, I wanna say that I have learned jack shit about binaural mixing at school. They've talked about binaural recording (which is nothing more than sticking a fucking dummy head with built-in mics into a room), and they've mentioned BRS (binaural room scanning) and the fact that nobody uses it.
So we've learned a lot about binaural in general, but not making binaural mixes from regular sessions. They haven't even mentioned the binaural panner in Logic, which I find odd, because IT'S FUCKING SWEET! It can really take a mix and just open it up, or sometimes, make it total mush.
I use it in different ways. If I'm doing something that's already kinda spacey, has a lot of cool noises, or is unbelievably simple (having only several tracks), then I'll do all of most of the mix with binaural panners. If something is really complex (having a fuck load of tracks), is really heavy, or is too traditional, then I'll only use the binaural panner on a few, if any, channels.
Everything, of course, is up to personal taste, so there's no need to go into when to use it and for what instruments and where to put them and all that good shat. I will say though, that I usually always leave heavy guitars (or anything that needs to be extremely "solid") just as they are. I don't even bother popping them into binaural mode, because it kind of takes them and softens them up. Notice I said usually, so if I'm feeling squirrelly or if the tone doesn't need to be uber rock solid, I'll fuck with them.
So here's what I really want to talk about: Time-based effects + binaural panners
First of all, if you are wanting your reverbs and delays to follow their sources (e.g. a left-panned guitar to have its reverb also come from the left), then you need to make sure your aux send is post pan. If you don't want this, just leave it post fader.
My real issue is whether or not to make my aux's binaural... I almost always do, though. And when I do, I give the L and R a 180 degree spread. It just sounds so much better. If I'm using space designer in a normal mix and I load an impulse response called "bigass mountains," then I'll be like, "oh man, it feels like I'm standing in the middle of some bigass mountains." But when I load that same impulse response in a binaural mix, it just doesn't cut it. It take my 3 dimensional, super realistic, super deep mix and applies a nice matte finish to it. It takes the whole thing and puts a yellow cigarette-smoke-stained window in front of it... sort of. Basically, it takes a little life out of it, and I don't really feel like I'm in the mountains.
But if I put a binaural panner on my fx return and spread it out, it's like "FUCK ME SIDEWAYS I'M IN THE GODDAMN MOUNTAINS!" At least to me, anyways.
So having said all that, even though I'm stubborn and won't give up my ways, here are something that I still wonder about, and things you should consider when deciding which way you want to do it:
-If the plug-in is getting a full stereo input from the source, then it's taking a straight up copy of that binaural signal, which is already fucking binaural, and applying the effect to it. So this seems like the correct and most accurate way to do it would be to leave that aux channel with a normal panner, because if you use a binaural panner, then you are technically (and literally) rebinauralizating it (... >_> look it up, it's totally a real word).
-If all the other channels are binaural, and are made so to create a false space, then is the time-based effect already a good enough false space as is, or should it be doubly false? Wouldn't it make sense that if a normal reverb is the standard for a normal mix, that it should have extra dimension added to it to fit in properly with a binaural mix?
-If half the tracks are normal and half the tracks are binaural, then should there be separate time-based effects for each kind?
-Should the size of each binaural panner's space be the same (I'd say yes).
-If the size of a panner's space is enlarged, should reverb time or pre-delay be changed accordingly (I'd say no)
So that's some shit you should consider. It's all really depends on how you want to do it, so ignore my little answers up there... I just felt the need to throw in my opinion.
For anyone who hasn't messed with binaural panners yet, and hasn't read the info in the manual about it, here's something you should know: if you use more than one binaural panner, you should disable that "diffuse-field" shit on every single one of them, then buss all their outputs to a single aux, then apply the "Binaural Post-Processing" plugin to that aux. This does all the processing for the mix as a whole (or just for that submix as a whole) instead of having each individual panner do it for each individual track, which obviously uses more CPU. Plus, according to the manual, it sounds better that way too. I've repeatedly compared the two, and I can't hear a difference, so I guess it's one of those things where the quality is better, but nobody could ever tell you which is which. I just do it that way because of the CPU issue, plus you can choose what type of processing you want (which gives you the awesome "speakers" option)
Fuck, my posts are long as hell. If anyone actually reads my entire posts, then thank you!!! But if not, I don't blame you.
One final word of advice: if you don't have any mixes to work with or if your mixes don't really seem like they'd be suitable for binaural, then check out NIN.com and go to the remix section. You can download entire sessions, then load them into Logic and fuck with them. Even if you don't like their music, it's still really great stuff to use for experimenting with binaural.