How To Mix

[waveheavy]

I enjoy helping others, so this basic mix method I learned from Nashville producers I would like to share. My aim is to help newbies new to mixing in a DAW that are struggling, and not to try to change how you old pros out there mix your music.

A little background: I learned this basic method from two Nashville producer/mixers (two-time Gospel Dove award winner Kevin Ward, and Jon Wright) who have recorded and mixed records for Willie Nelson, Dolly Parton, and many other Nashville recording artists, including records at Capitol studios in Los Angeles for big band Jazz and for producer/singer Wayne Haun's beautiful album of Jazz standards called "Old Soul" which went to 3rd in the Jazz charts when released in 2013.

To test this method I highly recommend doing it in the order listed here. This method especially applies to traditional mixing of singer/songwriter material using an "in-the-box" DAW method.

1. First step is to setup track groups, mix outs, and a Limiter on your Master bus. Recommend setting up a Master bus track (an Aux track) which is the final output to your audio interface. This is where you place a Limiter set to no more than -1.5 dB while you're mixing. You only want a touch of limiting while mixing to control the 'occasional' over, (an "over" means going above 0.0 dBFS on your DAW's peak meter which causes wicked distortion). If your Limiter is showing a lot peaking in the red while mixing then your track volume levels are too hot, and you need to pull them down a little.

Then setup a Mix bus which you'll send everything else to, and then from this Mix bus it will go into the Master bus. So the path should be: a) instrument tracks to: b) bus group tracks to: c) Mix bus to: d) Master bus to: e) audio interface. On this Mix bus is where you will place your final mix plugins like EQ and compression, but right now if you place them on, leave them bypassed for now.

Setup Group buses for similar sounding instruments. For example, all synths would go to a Synth bus, all drums to a Drum bus, lead vocal to its own Vocal bus, and background vocals to its Background Vocals bus, guitars to a Guitar bus, etc. With bass you might want it on its own bus, etc. The idea is to group instruments for similar treatment on the buses for final EQ etc., and to make it easier to do final mix tweaking with groups instead of going through all the individual tracks again. Another reason is so you can 'print' stems from those group buses if you want to send off to mastering engineer, etc.

FX and Sends - for effects like reverb, delay, parallel compression, etc., create separate buses for them. When you want to use one of those effects on an instrument, on the instrument track you'll find a "Send" box to select whatever effect bus you want (i.e., reverb, delay, etc.), and a fader will pop up (somewhere in the DAW) so you can adjust the effect just on that individual instrument track. This saves a lot of CPU power because you're only setting up one instance of the reverb or delay, etc., and referring to it with a Send control as needed for each individual instrument track. Even if you want to use 3 different reverbs, this is still the best way to set them up. It's how you would do with a hardware analog mixing board also, so this is an old technique, nothing new.

2. Track Cleaning and Phase Correction - if using instrument samples, you shouldn't have to do much track cleaning; maybe a bit of low end filtering only. But with recorded tracks in a studio with other instruments around, or especially a recording at a live venue, all... the tracks will need some cleaning. When recording, it's standard to record a few seconds of the room sound and also at the end of the track. Don't chop this off before doing the mix. With just one track alone, the amount of noise might be minimal and you won't think much about it. But let's say your have 50 instrument tracks with all the track noise piling up. You'll be forced to deal with it. There's some easy to use tools on the market that will clear out 60 cycle hum, tape hiss, white noise, etc. Waves makes a plugin that will subtract those things while leaving the main audio entact. Izotope also has a plugin for it.

Failing to do Phase Correction my producer instructors said is responsible for about one third of all muddy sounding mixes. Once again, this may not be much of a problem using instrument library samples, but with recorded material, especially live recordings, phase issues can be big. Two recorded tracks out-of-phase with each can 'cancel' the audio, reducing the instrument's volume as you increase the volume control. Because the two tracks are 180 degrees opposite of each other, they cancel each other out. This is why you always... do periodic mix checks listening in Mono, because in Mono you possibly may not hear the out-of-phase tracks at all! Hear it in stereo, go to Mono, and the instrument suddenly drops out. With the reverse situation, with two tracks who audio waveform is aligned in the same direction, it will cause an increase in amplitude (volume). To correct this phase situation, zoom in to both tracks where you can see their waveform lines, and simple nudge until the phase is aligned (going in the same directions). There's a plugin called Auto-Align that can help with phase issues. UAD also has a plugin that lets you change the phase of a track in degrees. There's probably some free ones on the Internet also.

3. Static Mix - OK, got the tracks setup with groups, cleaned, phase corrected. Now it's time to do a rough mix. No effects, only adjust mix balances of volume and panning. Not ready for EQ and compression, nor effects; not yet. You want as good as an initial mix as you can get at this stage, just using volume and panning of the instrument tracks. This is where you experiment with panning, determining a best fit for spreading out and placing your instruments in the Left/Center/Right stereo field. It's common to keep the kick, snare, bass, and lead vocal panned to the center in today's music styles. In general, heavy bass instruments panned center, higher range instruments can be panned anywhere from center to wide.

Track Volume Levels, watch these. Remember, you only want the Limiter on your Master bus hitting the occasional "over", not constantly. On your DAW's peak meter, -18 dBFS is equal to 0 dBVU on an old analog mix board VU meter. That is the recommended recording level into your DAW, ie., -18 dBFS on the DAW's peak meter. It's always good to leave the final mix level anywhere from -6 to -3 dB of headroom on your DAW's peak meter, to get the later final master up to production volume level. If you find your mix hitting your Limiter too hard still when doing the initial mix, simply select ALL the tracks faders and adjust them down slightly until the Limiter isn't going over, then continue the mix.

4. Treating Vocals - if you record a lot of vocals, you might want to look into your DAW's feature for creating a composite vocal track, which means piecing several vocal takes together to create the perfect vocal track. Besides EQ and compression, the vocal track is almost always going to need a run through automation leveling. The idea is to bring out the low, undefined, areas of words, while reducing areas that are too strong, especially sibilance in the vocal (like 'esss', 'p', etc.). (A simple Pop Filter made with a metal coat hanger and nylon stocking will help a great deal in preventing a vocal's 'p' effect getting recorded.)

(Continued...)

[waveheavy]

(Continued...)

5. EQ and Compression - the idea here is only... EQ and compression. No other kind of effects yet. This is where I recommend listening in Mono while doing this, especially for the beginner. Mono will reveal any phase cancelation. It's also easier to mix volume levels in mono.

I recommend applying EQ before compression. Some producers do it the other way around (like Dave Pensado). EQ first is how I was taught, because do you really want to compress all the bad frequencies before cutting them out? The idea is to do 'subtractive' EQ first. That's where you look for conflicting frequencies of an instrument and do little EQ cuts, using a narrow Q (width). Some things, like toms, you may want to use a wide Q cut in certain areas (like the mids). One of the ways to find conflicting frequencies on a track is to create a narrow Q boost all the way up in your EQ plugin, and then sweep the whole frequency spectrum slowly. When you hear a loud squeal, similar to microphone feedback, then you might want to apply a narrow band cut to that specific frequency. Listen to the instrument's frequency "power" area(s). If you hear a specific range of an instrument that seems overpowering compared to the rest of the frequencies of the instrument, then you know you're gonna' need an EQ cut at that frequency power area, otherwise the instrument is going to stick out too much in the mix over everything else.

OK, pick your first track to start with and do the above EQ step. With Rock, Pop styles, it's usually common to begin with the drum kit, the kick specifically. Some begin with the vocal. It's up to you. When you get rid of possible conflicting frequencies, then apply "creative" EQ if needed. This I must say so there's no confusion. An instrument track may not need anything... done to it; no EQ, no compression, not anything. That's always the main desire for a track when mixing, not having to touch it, the track having been a well done recording. The object of subtractive and additive (or creative) EQ is to make the instrument sit in the mix, having it's own space. Instruments that have similar frequency ranges can overlap each other, which is called "masking". You can either cut some of the frequencies out of one while listening to the other until it becomes more defined and clear, or you can boost the frequencies of one until it sounds more defined and clear. In my experience, cutting for clarity is better than boosting.

Add Compression only as needed. With the rough or static mix, you'll discover some tracks need a little help to get up to the level of power of all the other tracks. Likewise, some tracks will be over powerful. Be careful not to over compress your instruments and/or mix. The standard axiom is "compress often, but only a little". That means prefer multiple compressors hitting just a little over just one compressor hitting hard. It's actually one of the mixing secrets to good gain staging, keeping the dynamic of the mix, and preventing the squashed sound of over compression.

With the first track done, pick the next track to work on. You're listening in Mono (right?). While listening to BOTH tracks at the same time, apply EQ and compression as needed on the second track until it sits well with the first track. Continue this process until all the tracks are blended.

6. A/B Comparison - You've done your rough static mix of using only volume and panning, and that's the natural balance of the instruments you found in the beginning. You really should want to keep that balance. But when you apply EQ and compression it's easy to lose that static mix balance. The way to overcome this is by comparing the track levels before EQ and compression with levels after changing it. This also applies with plugin effects too. Magic A/B is plugin that helps with this; it keeps the volume levels constant before the effect was applied and after it is applied. This way you hear what the plugin effect actually did to the audio. You hear the change clearly. The reason this is important is because of a thing called the Fletcher-Munson curve, which basically says something will always sound better to our ears when the volume is increased. The reality is that our ears can be fooled with volume boosts, which is what happens most often when applying plugin effects, EQ, and compression. So a same-volume comparison is an accurate way to know just what the applied effect did to your audio. This is what the A/B switch on your audio plugins is for. This will also make sure your original static mix volume level you worked hard to achieve earlier is maintained.

7. Add FX - At this point you're ready to go back to Stereo listening mode. This is where you apply reverb, delay, parallel comps, special FX, etc. This will require adjusting levels again, but if you kept your gain staging in the previous steps, your mix still should be close to your initial static mix.

8. A Mix is not complete without some Automation - this is where you ride the faders of your tracks to bring in instruments, reduce them back into the background, level vocals, or whatever you need to do to create a living mix that moves with the song's build and groove. A Fader Port controller is a good tool to have here.

9. Final Mix Tweaks - if the mix is done well, there really shouldn't be much left to do for the mastering stage. I recommend starting a separate project for the mastering stage. You could just slap on your mastering stage plugins on the Master bus and master within this mix project, but there's some drawbacks. One is that your computer's CPU is already taxed pretty hard by this stage of the mix, so resources is something to think about, as a separate mastering project will have those resources free at the start. Another reason is about focus. With the mastering stage you're focusing on the final balance of frequencies and the final level. No more effects or reverb to add in the mastering stage, just final EQ and bus compression should be used (maybe a small touch of enhancing too, and a DeEsser). I also recommend waiting a day to let your ears rest after a Mix before going back to verify. Same if you're going to do the mastering.

[waveheavy]

Good book by Mike Senior. I used to have a copy.

Of course there's a whole lot more to mixing than what I covered regarding specific techniques. I only wanted to cover the basics for newbies. Thanks for checking it out.

I probably should have added a section about the mastering stage and general mastering philosophy.

[waveheavy]

Here's a short essay on Mastering your mix. The MAIN thing is 'use your ears'. Mastering technique is only as good as when you 'use your ears'.

1. Mastering Philosophy - originally back in 1948 in the vinyl record days, and when Ampex introduced its tape recorder, mastering was about getting the audio from tape (Ampex) to vinyl with the audio balanced so as to obtain a good vinyl master. Too low of an audio signal and there would be too much noise transferred to the vinyl, too high of an audio signal and it could destroy the vinyl and the lathe's stylus that did the cutting. These engineers who cut these masters were called transfer or cutting engineers. In 1957, the mastering engineers learned that applying EQ and compression could make the apparent loudness of the record sound louder on the radio. Producers considered the effect this loudness would have upon the buyer, and the world of the mastering engineer then took off with the art of trying to get the best sound with a final mastering of the audio.

2. Method - analog or digital? By analog I mean outboard non-digital gear.

Most essential mastering-chain plugins - linear phase EQ, a DeEsser, and a compressor. These 3 tools will be needed most often.

EQ:
analog EQ or digital EQ?, that is the question. Analog EQ can introduce phase-shifts around the target frequency. A digital linear phase EQ won't do this and is more transparent. Using an analog EQ in mastering can introduce these phase shifts and color the audio in a way that can be desirable if used judiciously. But for aggressive narrow cuts or boosts it can be a bad thing. Producers like Fab Dupont prefers an analog EQ, like Great River, and he uses it aggressively. In many cases, analog EQ allows a greater amount of boost than a digital EQ, just as it can also be with analog compressors. Digital linear phase EQs are more transparent though. Not all digital EQs are the linear phase type (these are: Waves Linear Phase EQ; Algorithmix LinearPhase PEQ; Flux Epure; Melda MEqualizerLinearPhase; FabFilter Pro-Q, Izotope, etc.).

EQ Bands - mastering engineers use multiple frequency band areas to gauge the audio balance. I like mastering engineer Ian Shepherd's ranges:

1. 50-60Hz - kick thump; bass boom
2. 100-200Hz - snare punch
3. 200-500Hz - warmth area for vocals, guitars, piano
4. 500-1000Hz - body and tone, can be harsh range
5. 2 kHz - guitar bite, clarity and aggressive (careful in this range)
6. 5 - 10 kHz - vocal clarity, snare clarity
7. 16 kHz - air, space, sparkle

http://www.independentrecording.net/irn ... isplay.htm

Setup a Linear Phase EQ with these frequency range bands. Ideally you want wide Q for boosts, narrow for cuts. For example, for the bottom two ranges I'll use a Q-width of 1.00. On the mids I'll use a Q-width of .70, and on the highs a Q-width of .50. Ideally, boosts should be subtle in the mastering stage. If you find yourself having to boost too much, then it's a sign to go back to the mix and do some more tweaking.

Fletcher-Munson curve again: a curve that shows how volume changes can fool our ears. A volume boost almost always sounds good. So in the mastering stage, it's very important to match volume levels of before a treatment and after the treatment, comparing them. Steinberg WaveLab has a tool for this. Ian Shepherd offers a plugin to do this matching also. You can do it simply by noting the volume level prior to the treatment, and then setting the same volume with the plugin output control, and then switch back and forth to compare. This is the ONLY way to accurately compare what the treatment did to the audio. It is one of the main tools of the mastering engineer.

Because of the effect the Fletcher-Munson curve reveals about how we hear volume levels, it's important to get into a habit of listening only for a few seconds to compare a treatment of the audio. Play the audio only for a few seconds just long enough to hear what changes the treatment did, and then stop play. Do another tweak if needed, listen for a short few seconds, then stop play. Don't get used to listening to long sections at a time. Notice this method trains your ears because when you stop the audio you will have to think about how it sounded, and then consider what further treatment might be needed. Listen-compare, stop the audio, apply the EQ treatment, listen to the change and compare, stop the audio, apply treatment, etc., in that order. None of the constantly playing the audio while you fiddle around with the EQ controls.

Avoid boosting cycles. For example, you want more clarity around 2kHz so you boost it until eventually the mix sounds thin and harsh. So you go to 100 Hz and boost to add warmth. You then wind up with the mid range being dipped out with the mix sounding thin. It's easy to get into an endless cycle like this with trying to overcome one big boost with another. Make small changes, and reduce areas that are overpowering while doing level matching.

DeEssing - a de-esser works similar to a gate/compressor. It reduces certain frequencies according to a threshold. If the mix has vocals or cymbals, or other high range instruments that you applied EQ sparkle to, they will most likely need some de-essing in the highs. On a vocal track this plugin should actually be the first plugin you use. As with small EQ changes, that's the only usage of DeEssers in mastering also, small changes.

Compression - the ideal is 'light' amount of compression. It can vary based on the style of music you're doing. Jazz and orchestra requires very little compression, maybe 1 -2 dB at most with a 2:1 ratio. A low ratio is recommended for most styles. You are not trying to boost the loudness of the audio here, you're trying to balance and gel the mix together more. If that was done well in the mix stage, very little or no compression could be the case.

Limiting - RMS level is your gauge. Most all DAWs have an RMS meter in addition to the Peak meter. A mastering facility will almost always use a master quality gauge of some sort, either an analog meter or a digital meter. Especially use your ears when setting compression and a limiter; most often you do not want to hear the audio 'pumping'. Might be different with some EDM.

How to adjust a Waves Limiter:
On a Waves limiter, there is a small up/down arrow box in the center. You pull this down and it brings both Threshold and Output down together. When you begin to see red in the Attenuation bar at the right, that is the start of distortion being added to the audio. That means that is as far down as you can go with the Threshold and not add distortion to your audio. Move both sliders back up to 0 and set -0.03 dB on the Output Ceiling, and then at the loudest point in the audio, bring down the Threshold to set the desired loudness on the RMS meter. That's all the loudness you get for that mix without adding distortion. Some treat a Limiter like it's a guarantee safety net. It's not, it can distort if pushed too hard just like a compressor can. Some simply disregard the red showing in the Attenuation bar because they can't hear the distortion happening.

A general RMS standard for most music library tracks I've measured is -10 dBRMS. EDM is usually a bit higher like -7 to -8 dBRMS. Orchestral and Jazz somewhat lower, around -14 dBRMS. To show how the loudness wars has changed things, a Jazz record mastered in the late '70's & '80's would be around -20 to -18 dBRMS (around 0 dBVU on an analog meter). Some Rock LP's in that era would be around -16 dBRMS.

Pay for mastering - if you really want a quality production sound, I suggest sending your mix to a professional mastering engineer who has the best equipment and facility for mastering, and years of experience to know how your mix should be mastered to compete in the music industry.

[LamarPecorino]

Thanks for sharing those tips. It is always beneficial to see what steps others use to get a quality mix.

[Len911]

De-essers and p popper software is difficult. It's usually easier to lower the volume on the s's and pops. pronounce the s's as z's and p's as b's, or use a flat mic without a presence boost for the s's. I have a Sonnox Supressor, "An award-winning Linear Phase Dynamic EQ ", lol, but I've still never managed to get it to work right. Maybe it's the pre-ringing problems with linear phase eq? The pros and cons of each type of equalizer https://www.meldaproduction.com/text-tu ... equalizers
The other kinds of equalizers work on harmonics, I believe Melda and SurferEq are a couple. So sometimes you can avoid collisions and masking by emphasizing or de-emphasizing certain harmonics.

I use an API compressor with unity gain and a choice of old and new style compressor and it seems to sound very natural no matter the setting for my vocals, no matter how aggressive.

I use the Sonnox Oxford Limiter, and audition the presets until I find one that sounds right. I also use one of the dithers onboard. I don't fiddle or care about rms values,lol!

I learned something about reverb using a mid-side technique from a Brainworx plugin and my reference cd. I could never get reverb to sound right before.

It would be difficult imo to write a book on mixing because there is just so much gear, and some of it works, and some doesn't work as well. Keeping up is almost impossible. The old books I have are so outdated they're not really useful.

The analog going in, the arrangement, and the as little as possible of the algorithms after it's in, is probably the best mixing philosophy imo.

[mojobone]

with recorded material, especially live recordings, phase issues can be big. )

That much is dead on. Last live big band recording I did, I counted 17 open microphones on the stage, not counting the boundary mics and whatever was on the Decca tree. Personally, I'm not such a fan of recipes, when it comes to artistic endeavors or rather, I love to play around with recipes in the studio or at mixdown, and I sometimes find 'one size fits all' advice to be sorely lacking. Nonetheless, I read it all, cuz you can pick up a useful nugget or two anywhere, and even a broken clock is right, twice a day, unless it displays military time, which tells us sump'm 'bout the military. [and before U flame, I'm a decorated Navy vet]

On my first mastering gigs, all I had was an Aphex Exciter with Big Bottom, two-band EQ from a Tascam mixer, and whatever else I could get from an Alesis MidiVerb IV. This might seem meager, but in the early days of ADATs, this and a few multitimbral MIDI instruments amounted to a pretty capable rig.

Applying processing at the mastering stage is fraught with perils, but a tight mix makes all the difference. I theenk having control of how the sounds are recorded helps with the mixing; I'm willful about getting sounds off the studio floor that won't give me trouble, later. Mastering ideally involves another set of ears, but not everyone can afford all that; it's worth learning how to get the best from two channels of audio; that's mastering, in a nutshell.

[waveheavy]

Thanks for taking the time to read it mojobone. I have to disagree with you about what mastering is. If it were as simple as you stated then there would be no need for professional mastering houses and mastering engineers who's sole job is to do just that, make a mix better, and with using the audio tools that I mentioned.

I had a TASCAM mixing board, a TASCAM 38 8-track reel, Alesis Midiverb, Dolby noise reduction system, back in the '80s, but I would never claim that substituted for a real mastering engineer's equipment and setup, even in that era.

In case you missed my main introductory statement, I offered this mix method for the newbie who doesn't have a clue where to start, not to you old pros who through trial and error have figured out what works best for yourself.

And I am a U.S.veteran also; USAF, served as a crew chief on B-52Ds during the tail end of the Vietnam war.

[artturner]

Just wanted to say thanks for your time and info! You definitely have some ideas on organization I want to try to incorporate.

[waveheavy]

Just wanted to say thanks for your time and info! You definitely have some ideas on organization I want to try to incorporate.

You're very welcome Art.

That is the basic method I still use today, although my ears are not as good today since I've gotten older. That method may seem a bit disciplined, but that's for those starting out. After a while one learns where to make it shorter. For example, I probably should, but I don't do the match level after applying a treatment to the audio as much.