The loudness of mixes is a highly debated topic, so I thought I'd throw in my 2 cents from my perspective as a mastering engineer that's seen a lot of different mixers workflow.
How We Used To Mix
Before we get into how to get your mixes loud, let's rewind 10 years back to 2006... almost every major label project was mixed on a big SSL, Neve, API, ect. Protools (and other DAWS) were often used as the "Tape Machine", playing back each track or group of tracks to the console. The mix engineer worked hard to achieve one goal, to make the mix sound amazing, no I didn't say loud, just amazing. Most guys used some master bus compression, but usually just to achieve the sound and energy they wanted, not to make the mix loud. If they wanted to hear the mix loud, they'd turn up the volume knob and play it loud! After the mix had been approved, they'd print (record) the stereo mix from the console back to a stereo track in Protools, to analog tape, or to another digital medium (DAT, Masterlink, ect). When they printed the mix the only thing that they worried about was to make sure they didn't "clip" the converters, so most guys would print at a lower level just to make sure they didn't clip. Sometimes the mixer would print a louder mix or "Faux Master" mix to give the client that had a little limiting so that the mix wouldn't be too drastically quiet, but for most guys that was pretty much all they did to try to get it loud, after all, they were hired to mix because they had a reputation for being great at mixing so they didn't worry too much about it being loud. Then the mastering engineer would apply dynamics processing, EQ, stereo field adjustments, and all sorts of other "black magic" and make it LOUD!!
The Game Has Changed
Now let's jump back to 2016, many people are recording and mixing entire albums with only a microphone and a macbook, and recording schools have thousands of graduates every year.. it's no longer an "elite" club of mix engineers. So how does a new mix engineer get noticed in a sea of people who are all trying to start a career in mixing? One of the ways is they often mix louder than some of the more seasoned mixers. After all, most people don't get to hear the unmastered mixes of their favorite albums, they only hear the masters, so that's what they shoot for. So a lot of mixers, especially in the EDM world, will have every master bus insert slot filled with inflators, compressors, limiters, and maximizers squeezing out every drop of volume they can get to make it loud. If they're doing a mix on spec, many times they get picked over the other guy because a louder mix is often perceived as the more exciting mix. A&R guys tend to like mixers that turn in a mix that already sounds like a finished master.
So What's Wrong With Mixing Loud?
One of my favorite things about music is that there is no right or wrong, it's only the response that we get to the music that we make. I will say however that when you print your mixes too loud, you are potentially reducing what the mastering engineer can contribute to take the mixes to the next level. Sometimes I use this analogy... Imagine taking a photo with a great camera, then running it through a ton of Instagram filters and then giving it to a graphic designer to make a billboard ad with. He wouldn't be able to get back everything that was lost when you applied all of those filters and saved it to a 640x640 square. When you saw your giant billboard ad, it would look pixelated and terrible. But, if he had the original photo, he could see what you were going for, but using his tools and expertise, he'd be able to give you a far superior final product. The same goes for mastering, often times a mix that's really loud sounds exciting but it lacks the depth or punch that you'll hear on big budget records, why? Because the tools and experience of a great mastering engineer can both enhance a mix AND get it loud.
The Catch 22
Is this all making sense? Well, it's about to get more confusing. Many times when someone mixes with the "faux master" on everything sounds great until they turn it off to send to mastering, the whole thing falls apart and doesn't even sound like the same mix. Likewise, when a mix is too dynamic the mastering engineer can only do so much without it sounding "Squashed".
As a mastering engineer, I get every type of mix that you can imagine, from mixes that sound like they raised the faders and said "That's the mix" to mixes that look and sound like a big square brick. For the most part, the best sounding mixes are the ones where the mixer either mixed on an analog console or "acted" like they did. By that I mean, they use master bus compression as a glue but focus more on the processing of the individual tracks to get the sound they want and only worry about the loudness at the end before sending it off to the client. The best way I've seen is to have 2 master busses, one with just their mix buss compression / EQ and the other with all of their "faux master" processing. This way you can mix while only listening to the main mix bus, then when the mix is nearing completion, you switch over to the "faux master" bus and work on seeing how loud you can get it, and hear if there is anything in the mix preventing you from getting it loud, like a big sub boom or a funky frequency in the kick drum. Then you can adjust those elements and make sure the mix sounds great with AND without the "faux master". Then you print both versions down, you give the loud one to your client and the lower level one to the mastering engineer. This allows your mastering engineer the room to work but you stay competitive and make your clients / A&R guys happy. Also, it never hurts to send the mastering engineer both versions so he can hear what your clients have been hearing.
Tricks To Try On Your "Faux Master" Bus
Filter Out Unwanted Low Frequencies
Low frequencies take up more space in a mix than high frequencies, so try using a linear phase EQ to filter out everything below what you're hearing in the mix (example everything under 35hz) Put this filter first in line so that your dynamics inserts aren't compressing the unwanted stuff on the bottom
Use a Mid / Side EQ to Filter Out More on the "Sides"
Usually the kick and bass in most songs is mono, so often times it's helpful to filter out unnecessary low end on the sides to clean up the low end using a good Mid / Side EQ. (I'll probably talk more about mid/side in the future if you're not familiar with it). An example would be in a rock song where the kick and bass guitar are panned right up the middle, and the floor tom is the lowest frequency instrument that's panned off center, try cutting the side a little below where the floor tom resonates. Sometimes the different elements like drum room mics or acoustics guitars have a lot of "sloppy" low end on the sides that can eat up real estate that you can use for another db or 2 in overall gain. Although if you find that this helps on your "faux master" bus, you may want to look at applying filter to those specific elements that it's helping with.
A good multi-band compressor used properly can work wonders, you can tighten up the low end while leaving the top end totally open, or you can add energy to the midrange without it feeling too compressed. I personally love the FabFilter Pro-MB because it has so many features and it sounds amazing. It's also has a lot of great presets that you can start with if tweaking mulit-band compression sounds intimidating.
You usually don't have to compress a lot to get a mix loud. If you do, try a gentle compressor like the UAD vari-mu, or the compressor section of the Slate FG-X. (Also, I'm not a fan of the "level" section of the FG-X, it can add all sorts of artifacts in weird places.)
Loudness Maximizers / Limiters
Please don't use the waves L2, trust me, there are many other better options. Again, kuddos to FabFilter for the Pro-L, it's easy to use and sounds way better than the L2. Also, the Ozone 7 maximizer section has a "transient" mode and "Clipping" mode that can sound good on more aggressive mixes.
One of the tricks I see mixers use a lot is to add a transient designer or the Waves TransX Multi before their limiter/maximizer if they dig what the limiter is doing but feel like the transients are just a little too soft after limiting. This way they can compensate for what the limiter is softening by exaggerating the transients. This should be used very gently though, if you find your self cranking on the transient designer then you're probably limiting too hard.
Turning up The Gain
Sometimes at the end of your chain, turning up the master fader a couple of db and clipping the output sounds great, don't be scared to clip it a little if it sounds good. Just make sure that you attenuate the final file at least -.5db
Final Output Level
Make sure that you're final output is attenuated a little. I can do a whole other post going into the details of inter-sample clipping, but for now just attenuate either the last thing in your chain or attenuate the file before converting to MP3, AAC, or a different sample rate. Basically if you don't you may hear clipping once the file is converted to the other format or sample rate.
Again, this is a topic for another post, but if you choose to dither or not, it won't matter for your "faux master" unless your working on very dynamic classical music, in which case you shouldn't even be concerned with making it loud.
Ideally you want your mixes to sound good with and without your "faux master", that way your clients can hear them loud and your mastering engineer has room to work their magic. And I get it, you want your mixes loud, but please don't let the volume be your total focus. The loudness wars are starting to finally calm down and it defeats the purpose of a great sounding mix if you destroy it to make it loud. Also, if you're mastering your own mixes, keep in mind that Apple music, Spotify, youtube, pandora, ect, all level match everything now, so if you make it really really loud they'll just turn it down.