BASIC MIXING GUIDELINE (for DAWs, "in the box"):
I like to do this occasionally so as to keep up from memory what I learned from pro-Nashville producers Kevin Ward and Jon Wright. I had mix lessons with them in Gospel, Rock, Bluegrass, Folk, Jazz, Country, and Big Band. Last I heard from Kevin was that he moved to LA and was working for Capitol Records.
Here's a mix (not mastered) I did long ago of a tune Kevin Ward recorded and taught us to mix. The singer's name was Darcy is all I'll say, and I will only have this link up for a short time, because the song is probably copyrighted, (but my mix isn't). The song was recorded in Pro Tools by Kevin Ward and mixed by me in native software Pro Tools 11 using an RME soundcard interface and UAD and Waves plugins.
Falling In Love With a Girl
MIXING:
Note: you can... mix on a good set of earphones. But you still need to check it in several listening areas, like your car, your home stereo, room monitors, etc. Most folks don't have the resources to do acoustic treatment of their DAW area, but that's really the only way to trust a mix done with just stereo monitors. I also recommend using the SoundID Reference VST Plugin. You can get it for whatever headphones you're using (like a pair of Sony MDR 7506 earphones, a long-time studio standard), or for whatever professional type monitors you have. This will help with the untreated room problem. You can mix on home stereo speakers, but you might not get a good balanced mix, simply because home stereo speakers 'color' the sound to try and make it sound good. When mixing, you instead want monitors that give you as flat a frequency spectrum as possible, so that you hear equally all the frequency range, from 20kHz down to at least 40-50Hz.
1.
Artifacts -- check every track for errors, listen all the way through each one, pops, clicks, digital distortion, artifacts, etc. Fix all those problems (Waves had tools for that, so does Ozone).
2. Setup your DAW session. Pro Tools is a bit difficult to learn with how to do this, but once you get it, it's easy. Create individual instrument buses based on instrument types. For example, all percussion instrument tracks might be sent to a Bus named 'Percussion'. All guitar tracks might go to a Guitar Bus, keyboards to a Keys Bus, orchestral strings to a Strings Bus, etc. Then create your Reverb Bus(s), and other FX's buses you might want to apply to many instruments at the same time. For Reverb, I use 3 different ones most of the time, one set for just the 'room' to back the instrument away a bit from being close miked, and then another reverb to place instruments back a little further, and then another reverb for the overall Hall for the highs. Each reverb will have been EQ'd (I recommend learning this from the Puremix website by Fab Dupont). Then for each instrument track, use its 'Send' function to point to the Reverb's Bus. This way you use less resources, because reverb takes up a lot of computing power.
2.
Phase -- Check the tracks for phase cancellation issues, mainly for instruments recorded together in the same room, as there might be mike bleed between them. For example, if a tamborine was recorded in a mike standing close to the mikes recording a drummer, there will be mike bleed between the two instruments. So you'll have 2 versions playing against each other, like the snare heard in the tamborine track a little, and the tamborine heard in the drum tracks a little. They will tend to cancel each other out if 'out-of-phase' with each other, making each of the two tracks sound weak, or in extreme cases completely unheard (cancelling each other electronically).
The manual way to correct this (which we were taught), is to look at the audio wave expanded to a straight line in your DAW, and compare the two instrument track's wave alignment. If the alignment is opposite of each other, flip one of the instrument's phase so they both align in the same direction, using either a feature provided in the DAW, or by a plugin with a phase button. Not doing this is responsible for about a third of 'muddy-sounding' mixes. Some producer/mixers only mix in MONO because of this, and then open it up to STEREO in the latter phase of the mix.
3. Get all the track's faders in a sweet spot. Set a Limiter on the Master Bus, just hitting 'occasionally', no more than 1 or 2 dBs. This is only to pick up the occasional 'over' (digital peaks above 0.0 dBFS means automatic distortion, and it sounds terrible). The ideal level for the finished mix should be around -6 dBFS and no higher than -3 dBFS on the Master Bus peak meter. Less is OK too, not too much. I used to use the RMS setting for loudness, and shot for -12 to -14 in my finished mix level. You want to leave room for the mastering stage. The mix is not supposed be at production levels. So slapping a Limiter on the Master Bus and hitting it to get post-production levels for the Mix is a no-no.
4.
Rough Mix: this is the next stage. No EQ, no compressors, no effects, no reverb, nothing but just balancing the volume of the audio of the tracks, and panning them, to get the best sounding balance between them you can get. You can try different versions of the rough mix by picking a certain instrument and balance all the other tracks around that. Panning may clear up instruments near the same range masking each other too. Experiment.
5.
EQ Only: Go to MONO listening -- start with one track. I was taught to start with the drums, usually the kick, then toms, then the snare and cymbals last, balancing them as a cohesive unit, the drums sent to a Drums Bus, and the Cymbals sent to a Cymbals Bus. Could start with the bass, or vocal. It just depends on what style you're doing. With orchestral mixing, it's usually recommended to start with the high instruments, like the Winds.
The object is not to select each track in Solo and EQ it until it sounds good. You can, but know you're probably going to have to make some changes once you play with the other tracks and try to get it to fit in the whole mix. How it sounds in the whole mix is what is important. Pick a track, listen for what spectral range where most of the power is. Most often you'll need a cut in that range. Start with no more than -3dB cuts. See a YouTube video on how to use an EQ plugin, for it's too much to cover here. The idea is with all your faders down, pick a starting track, EQ it, then add another track playing along with your first one and EQ it, adding tracks as you go, and always listening to each track within the mix, and only solo when absolutely necessary.
Do subtractive EQ first, like most all of the tracks will need some kind of Low Cut in the extreme left bass range. Find where each instrument's bass range ends by doing a sweep, then do a Low cut up to that point. The bass really piles up with all the tracks together, so this is an essential clean-up step. Then do creative EQ, which is where you highlight only the ranges of tracks that you want to bring out in the mix. It will take a little time to learn the frequency ranges of different instruments and vocals, but every mixer must pay that learning price, so don't skimp on learning that. When you have two instruments close to the same ranges, you may discover they mask each other so they aren't clear. If panning won't separate them, then you have to use EQ to do it. Each instrument will have several important frequency ranges that gives it its particular sound. You can cut some of those ranges and still keep its fundamental sound, and then boost those same areas in the other instrument track that is masking with it. Cut one, boost the other, will clear up both usually. It's simply a matter of which frequency ranges to choose, and that should be based on what you hear in the overall mix.
For frequency areas you can't quite find that are causing problems, setup one of the EQ points to a narrow 'Q' (width), and then boost it several dB, and do a sweep of the ranges. When you find an annoying frequency, do a narrow cut at that point. Then move on. With some instruments, depending on what kind of mike was used to record it, you might have to do with in several frequency areas of an instrument or vocal. With a vocal this is especially important, because you want to concentrate on what the natural sound of the singer is like, and try to reproduce that in the DAW vocal track. The microphone is a unique thing, it adds it's own set of artifacts to the recording that takes away from the natural sound of a voice or instrument. Thus there are tricks to what kind of mike to use for recording, depending on the vocal or instrument. This is a whole other learning curve for the recordist, not the mixer. Kevin Ward is one of the best at choosing the right mike setup that I've ever heard. That's why many of his recorded tracks don't need hardly any EQ or compression. I'm sure it was the same with Alan Parsons' recording ability.
The EQ step is about 60-70% of the mix process.
Vocals -- Special Attention needed here. Most DAWs have a 'gain' function where you can edit the actual waveform of the track, selecting even a very small region of the audio wave and raising or lowering its level. Electrical mikes tend to pick up a lot of artifacts not heard in a natural vocal, like breath noises, sibulance like 'S', 'P' which pops the mike, and if singing too close to the mike you get a proximity effect, which increases the bass in your voice (which is how impersonators produce special effects noises with a mike, like bombs, trucks, etc.). Don't just rely on a plugin to fix these problems. Look at the audio wave first, and go through it and try to balance the vocal performance. You can use the Automation feature at this stage, but only if you know what you're doing, because it could ruin the vocal performance. Using automation means you can do a record of the gain, or volume riding the track's fader while the audio plays. It won't affect the original wave, but only record automation data of the volume changes. You can do automation effects with many different parameters offered by each DAW, and actually with the parameters of each plugin you use too.
Vocal Compression -- very important, since the vocal is prime focus of a song. Say you've recorded a strong vocalist at the standard -18dBFS record input level setting of the DAW. And you're having trouble getting that vocal performance under control to blend with the other tracks. Don't just slap on a compressor and hit it hard. Instead use multiple compressors, each hitting just a little bit. Using just one in that situation will almost always guarantee over-compression effects. Watch your Gain structure too between the levels of each compressor, that matters too. For example, you have an Input gain, and an Output gain on your compressor plugin. If you set the first comp's output too low, it will force you to really raise the Input level on the next compressor so you have enough signal to process it. Careful of this, called 'gain staging'. Digital plugins in a DAW actually have 'sweet spots' for processing the audio, usually explained in the manufacturer's manual.
6.
Compression: this is next. Only use it for tracks that require it, that might be much weaker than the rest of the other tracks. The goal is not have to use compression at all. And it is possible, if you're good at editing the gains of an audio track. Alan Parsons, who produced and mixed Pink Floyd's
Dark Side of the Moon, said he didn't use compression. I admit I ain't as good as he is, so I do use compression.
Again, look to YouTube on how to use a compressor. My point here is only use it as needed. Not all tracks will need it. And the Limiter you put on your Master Bus at the start of the session should take care of the occasional 'over'. If not, then either find the track(s) that are causing the over and reduce its Fader, or select all your tracks and pull down all the Faders together until the 'over' goes away. Don't just slap on another compressor because of overs on the Master Bus. The main thing to remember about compression is to 'compress often, but only a little', and only for tracks that need it.
7.
Add FX -- If you used MONO listening,
now change to STEREO. Your mix ought to sound about 80-90% done. Careful with too much reverb. It's good to run the reverb through an EQ plugin on the Reverb Bus it's on, and cut out much of the low end, while maybe cutting some of the highs and midrange (to taste). For a Hall reverb, you might want more of the high end range, a slight boost. To apply reverb, simply move the level up on whatever you're applying it to, until you can just hear it. Then back it down about 1-2dB. That's it. You might need another compressor or EQ on your mix Buses, like for each instrument Group. Don't do it unless you feel you actually need it. And remember, compress only a little.
8.
Automation, Fader Rides -- depending on the style of music, you might want to do Automation Fader Rides on certain parts of the mix, like reducing the volume of background instruments in certain sections of the song. Other sections of the song you might want to boost the level of something, or change the reverb level, or any FX you might use. Might want to change the pan of an instrument, like bringing an instrument at the sides just a bit closer to center for a solo (not direct center from the side though, making it too obvious). It's also common to bring the level up of a solo instrument at the first, and then back it down once it is noticed, just to bring the listener's attention to it coming in. These techniques are used in Pop and Rock mixes all the time. Not so much in Folk, Bluegrass, Jazz, Orchestral styles.
9.
Final Mix Output -- depending on what style you're mixing. I usually shoot for an RMS level of -12dBRMS. RMS is using for loudness, not peak. All DAWs that I know of have this setting as a meter choice. Or you can use a plugin. For Pop or Rock, it's usually a bit louder, but still leaving -3 to -6dBFS on the peak meter for later mastering.
If you stream your music to an online music website, realize they all will process your audio some, and some require your audio to be a certain type (like .wav or .mp3) and level just to accept it. Most newer DAW's recognize this, and have direct upload links processing your audio for you, just for that site, like SoundCloud, Spotify, etc.
