Getting a Rough Mix
When I was just starting out as an engineer, I hated the idea of rough mixes. I was never happy with them, and I always felt like someone would hear them and have second thoughts about my work. As my career continued, I began learning how to put together roughs that were flattering, sometimes to the point that they even were chosen for the record. Rough mixes can be innocent, less-thought-out, and as a result can have more energy and honesty. The flip side is that they can also be terrible. Time usually doesn''t allow all the normal tweaking, and often you end up just throwing the tracks together to get it done and off to email or CD.
If you''re just starting out, or maybe you''re the tracking engineer but are trying to land the job of mixing on a particular project, sometimes the rough mixes can help give your client or artist a barometer of how well you mix. I''ve gotten several very good mixing gigs by doing just that. I was hired to do some overdubs, then I made sure my roughs sounded fantastic so that the client would consider me for the mix. It has worked for me many times over. In the next few pages, I''m going to share some of the things I do to make my rough mixes sound as good as possible.
The examples I''ll be giving in this story revolve around Avid Pro Tools, but most other DAWs will share similar qualities so the translation should be straightforward. For the purposes of this article, I''ll abstain from the use of a console or outboard hardware. I''ll be working strictly in the box.
The first thing I do on a rough mix is to “Save Session As” and rename it with the date and nomenclature that clearly states it''s a rough mix (which helps if you need to revisit it down the line). After that, I start setting up stereo subgroups so that I have the ability to make quick overall volume rides on, say, a drum kit or a group of rhythm guitars. The most important thing is to set up what I call a mix bus insert.
The mix bus insert is a stereo aux track that has its own dedicated input path (for example, bus 1-2). Its output is set to the main
monitoring path. I prefer this to a master fader because any plug-ins I put on it will be pre-fader, and thus not affected by any fader moves such as a long fade. Also, I can send the output of this mix bus insert to a separate stereo track via sends to print the rough mix onto a new stereo track in the session itself.
Assuming I''m using bus 1-2 as the input path to the mix bus insert, I''ll then set up a handful of stereo aux tracks for the subgroups. Typically for a rough mix session, I''ll create between six and eight stereo aux tracks. Then I''ll name them: Drums, Bass, Guitars, Keys, Vocals, and so forth. For each of these subgroup masters, I''ll assign a separate input path. For example, Drums bus will use bus 3-4 as its input, Bass bus will use bus 5-6, etc.
FIG. 1: Assign the drums, keys, guitars, and vocals to subgroups, and route all the subs through the mix bus insert track.
Sometimes it helps to go into the session setup preferences and rename the bus paths to match the inputs. So instead of bus 3-4, you could rename it Drums Bus. This will help keep things nice and organized. When you go through your tracks, simply route their output to their appropriate subgroup bus. In other words, your drum tracks should all have their outputs set to bus 3-4, or Drums Bus, depending on what name you use. Finally, remember that each of these subgroups should be routed to the mix bus insert. Think of it as your final destination (see Fig. 1 ).
Setting up subgroups may seem like a lot of work, but you''ll only have to do it once and it takes just a few minutes. You can import the settings into your next session so you don''t have to go through this for every song, and it''s the best way for you to do general volume moves without making it an all-night affair.
FIG. 2: Start the mix with a compressor, a brickwall limiter, and an EQ plug-in on the mix bus insert track.
FIRST THINGS FIRST
After I''ve set up my subgroups and mix bus insert, I start adding plug-ins. In the real world, I typically use hardware inserts for a lot of my processing, but in this scenario, I want to be quick. Because I work on mostly pop and rock music, my main concern with a rough mix is that it has impact and clarity. I''m not overly worried about nuance. That''s for the final mix. That said, I usually start by inserting three things on my mix bus insert: a compressor, an equalizer, and a brickwall limiter (see Fig. 2 ).
The mix bus compressor I choose most often is the Waves API 2500 plug-in. I use the hardware version daily, and I know it well. For my taste, I usually set up the plug-in in Old mode, with a 3:1 ratio, medium attack, medium-to-fast release, and with a high threshold setting to catch only the loudest peaks. I engage this compressor right away because I find that mixing into the bus compressor is the best way to make sure your mix holds up well. I like to hear its effect from the beginning.
Next, I insert a UAD Precision Equalizer. Any mastering-style EQ with high/low shelving and fairly wide bells in the mids will be fine. I''ll keep it set flat for the time being. I only put this insert on in the event that I want to just shape the overall mix slightly at the end. I usually find that most roughs suffer from being a bit too dark as you typically don''t sculpt individual tracks as much as in a final mix. A mix bus EQ can help with clarity in the low mids and up top where some “air” is needed, without requiring that you apply surgical EQ to individual tracks.
Finally, I insert a brickwall limiter, purely for level. While I personally dislike the “volume wars” and what they''ve done to audio quality in the past few years, it has become necessary to add brickwall limiting to even a rough mix. My favorite for a few years now is the UAD Precision Maximizer. It sounds great, gets plenty of level, and doesn''t destroy a mix like some others can. If you don''t have access to it, I recommend using the most transparent brickwall limiter you have. As with the EQ, I keep the brickwall in bypass until near the end.
Before I start pulling up faders, I like to set up a reverb send and a delay/echo send. Back when we mixed on consoles, there was usually a plate reverb and an echo normalled to the console, so it was always available. This is my way of keeping up that tradition. I start by creating two new aux inputs, and calling them Reverb and Delay. I set the input accordingly, and the output to the mix bus insert. Lastly, I disable the solos on the aux tracks so that they''re in Pro Tools'' Solo Safe mode. This makes it so that when I solo something, the aux input is still active, making it possible to adjust the reverb or delay in isolation. The reverb and delay on the aux inserts are set at 100-percent wet.
I almost always use the UAD Plate 140 plug-in for reverb. It''s uncanny how close it sounds to a real EMT 140, and it''s every bit as warm and lush. Any decent reverb will do here, but I prefer to use “plate” presets simply because of subtlety, habit, and familiarity.
For an echo, I''ll choose something quick and easy, such as the Sound Toys EchoBoy plug-in. It lets me change tempo, sync the tempo to the track, or choose whether I want eighth-notes, quarter-notes, or any other subdivision—all on the fly. If I sense that the track will not need a long delay, I will sometimes set up a slapback-type echo, mostly for vocals. If this is the case, I will use it in mono because I prefer mono slapback to stereo.
THE RHYTHM SECTION
It would be impossible for me to give you specifics on treating balances, mostly because every track is different and each track requires its own set of fixes or complements. However, I''m going to give you some general tips on treating drums and bass quickly so that you can put together a respectable rhythm section mix within minutes. First, I start by getting a decent balance on the drums and bass, with no EQ or compression. Then I''ll just try to deal with things that stick out or sound bad.
I generally start by putting a compressor on my drum subgroup. Often, this will be something like the Waves API 2500. I''ll usually start fairly aggressive with about a 4:1 ratio, with a medium attack to keep the transients alive but a fast release to make sure the drums have impact and a little bit of pumping. This breathes some excitement into the drum group. I don''t want to squash them; I just want to get a little life and sparkle out of the kit (see Web Clips 1a and 1b).
If your kick drum lacks definition, try pulling out a few dBs around 220 to 240Hz. This will get rid of some of that mud. I don''t do very much in the top end on a kick drum. I prefer to increase definition and punch by subtracting rather than adding. I also rarely compress a kick drum, unless it absolutely needs it. But don''t be afraid to EQ the heck out of it if necessary. If it sounds right, it is right.
Snare drums in rock music tend to lack depth and low end to my ears, and they''re often too “honky” or have too much midrange boxiness. That being said, I usually will pull out a little around 700 to 800Hz to pronounce the low end a bit. I may also just do a slight boost, somewhere around 5kHz. Also, check the phase with the overheads. I can''t tell you how often I''ve gotten tracks to work on and the snare is out of phase with the overhead mics, making it lose impact. I''ll also usually insert a compressor after the EQ. I set it to be pretty light, mostly grabbing the highest peaks, but I want to keep the transient alive, so my attack is set medium/slow and my release is set fast/medium. Most times, I''ll also do a similar treatment with the tom tracks. Usually, they have a bit too much boxiness and not enough depth. So, again, I''ll pull out some 400 to 600Hz, and give a slight boost to the upper mids for the attack. To be safe, I''ll check the phase, too.
I''ll usually do very little to the rest of the kit at this point. If I''ve tracked the project, I''m pretty confident that my sounds are there so I don''t need to do much. It''s all subjective, and you may find that you need to put a highpass filter on the overheads, or perhaps some compression on the room mics. These are common things. Don''t be afraid to play around, but in keeping with the principles of a rough mix, don''t labor over it. Just get it to sound good quickly and move on.
The whole time I''m balancing drums, I''m also keeping an ear on the bass to make sure it''s working with the drums, specifically the kick drum. Look for the usual suspects here: Too much pick noise? Shelf the top end so it''s not too bright. Not enough impact? Try pulling out some 300 to 400Hz to make the bottom seem more powerful. And, of course, try doing a slight boost around 90Hz for some thunder if you need it. If you''re feeling good about the way the rhythm section sits and your levels are at a good spot, with no clipping on your subgroups, it''s time to move on.
THE REST OF THE BAND
Next, I start pulling up faders on everything else except for vocals. This means guitars, synths, and anything that may be in the track. Because every song is different, my best tip is to maintain clarity by using your panning effectively. In other words, keep similar textures in different parts of the stereo field. For instance, if you have a synth and a guitar in the same general octave range, try panning them apart to maximize clarity and width. The goal of a rough mix is to have everything sit in a place where it can be heard. A lot of critical decisions will be made on the basis of a rough mix, so unless you know otherwise, try to make sure that you have clarity on the entire arrangement. Try to only spend a few minutes getting this balance together. It should be instinctual and not overworked. The biggest mistake people make is overworking the instruments.
You may want to apply an overall EQ to a guitar or keyboard subgroup. I do this to save time so that I don''t have to go into individual tracks and apply treatment. In some cases, I''ll insert an EQ on the guitar subgroup and engage a highpass filter to get rid of any rumble or mud in the guitars. You don''t want to lose any power; you only want to clean it up so nothing unnecessary gets to the mix bus. I may also add a little boost in the upper midrange because that''s where I can get some extra bite out of the guitars if needed. Because distorted guitars are inherently compressed, I typically opt not to use a compressor on them. However, I will sometimes put a compressor on the keyboard subgroup. Once again, my goal here is to simply catch any stray peaks and perhaps breathe a little excitement into overall group, if needed.
Finally, I''ll get the vocals into the mix and do my best to get them sitting right without anything on them. Of course, that''s easier said than done, and in pop and rock music, more often than not you''ll be reaching for an EQ and a compressor. Here, I''ll typically reach for a channel strip plug-in, most often the Waves SSL Channel. It lets me quickly set a highpass filter for removing rumble, EQ as necessary, and compress. Usually, I end up with a slight upper midrange boost and a lower midrange cut, unless the voice is nasally and thin.
With vocals, I always prefer very wide bell curves so that the voice doesn''t sound overly processed. Then I''ll set my compressor so that it grabs peaks, with a medium attack and a medium release. I may also insert another subtle compressor on the vocal subgroup to help it sit better in the mix (see Web Clips 2a and 2b). I''ll usually set this compressor to have a slightly faster attack and faster release, but I''ll set the threshold much higher so it''s just slightly hitting the signal.
There may be times when the use of Auto-Tune is necessary. In the case of a rough mix, however, the artist may prefer the vocals untreated to get an honest sense of the performance, especially if you''re doing this for comping purposes. Otherwise, it''s usually simple to insert Antares Auto-Tune, or something similar, and put it on the most conservative setting, using the Chromatic C scale so that all notes are active. If you set it right, you should barely see it doing anything, only occasionally grabbing the most egregious of notes. If it takes you more than a couple minutes to get it to sound transparent, then don''t bother. This is a rough mix after all, and you likely have 12 more to do!
Through all this, remember that you have a reverb and a delay set up. I usually use them for vocals and maybe a guitar lead or synth lead. Either way, they''re there for your
disposal, and don''t forget to consider them when you''re focusing on the vocals.
PRINT AND FORGET
At this point, I usually have a serviceable rough mix together. In many cases, it shouldn''t take more than 15 to 20 minutes to get to this stage. You''re now ready to check your brickwall limiter and perhaps a little mix bus EQ to make it shimmer and sparkle a bit more.
The whole point of the brickwall is to increase your level without clipping. Depending on what plug-in you use, be sure that the mix doesn''t go over 0dB and that your input to the limiter doesn''t clip. If your mix bus insert is clipping the limiter, just bring all of your subgroups down equally until you''re out of the red. Sometimes when you engage the brickwall, it can seem as though the drums get quieter or the guitars get louder. If so, it''s handy to use your subgroups to readjust your overall balances to compensate.
FIG 3: Printing to a track in the session keeps your mix inside the session, making later retrieval easier.
When I''m satisfied, I create a new stereo audio track at the top of the session, and using a send route the mix bus insert to the new track, and set the send level to be at 0dB, or unity gain. I mute the new track while it''s in record (to avoid feedback) and name it accordingly. This track will be for a print of the mix. The reason I print it in the session is so that I have a copy at all times inside the session, thus avoiding the need to hunt down a lost rough mix later on. It also lets me export different formats quickly without having to bounce in real time. I then arm the track, put Pro Tools into record, and print the mix (see Fig. 3).
When it''s finished, I recommend opening up the session for the next song, saving it under a new name, and importing all of the subgroups, aux inputs, and the mix bus insert from the last session. If you have the same audio tracks for multiple sessions, you can also match tracks and import your track settings from the last session to save even more time. In most cases, after the importing you''ll simply make sure your audio tracks are routed properly and you''ll be in good shape. Adjust a few balances, and you''re off to the races.
Rough mixing can be frustrating. Hopefully these tips will help you get better sounds more quickly, using broad strokes to sweeten things, and not getting caught up on details. As always, your mileage may vary, but use your imagination, and if you have the time, please experiment. Have fun with it!
D. James Goodwin is a producer and mixing engineer, operating out of his own studio, The Isokon, in Woodstock, N.Y. His discography includes The Bravery, Norah Jones, Devo, and Natasha Bedingfield.