TP Audio's Mixing Primer

Yikes!

In order to do his job effectively the engineer has to be reasonably capable of feeling musical rhythms, and of assessing proper pitch and the relationships of multiple melodies. Thankfully this does not require musical ability, just an ear for it. Unfortunately, many mixes give one the impression that the soundperson does not have even the slightest ear or feeling for music. (There's a couple of bands out there who think that about me: Sorry - some rooms just can't be made to work with any band that has something as loud as a drum kit. And yes, your cheap wireless did sound poor, just as I said it would.) In most mixes we hear some drums, but not a drum kit. The drummer may be laying down a great beat, but in the audience the rhythms of the kick, snare and hi-hat are usually so disjointed that it sounds as though they are being played by three different individuals, none of whom are paying particularly close attention to one another. Other instruments, the melodies of which we cannot discern, accompany these drums. We can usually hear that someone is singing, but we cannot tell what or how well. Along with the drums we hear a guitar (until the guitarist solos, at which time the instrument becomes inexplicably buried in the mix - this isn't always the soundperson's fault however - see my BandRant1 page). We also hear a lot of semi-low frequency stuff that we assume is the bass. We usually cannot hear the keyboards at all (except during the guitar solo) (level those patch volumes out, Liberace! - again, see my BandRant1 page). And, it often sounds as though the bassist and the other musicians are not paying any more attention to "the drummers" than they are paying to each other.

Why?

During sound check, the drums are invariably done first. One at a time the drums are cranked up and adjusted for tone. Then they are balanced against each other sonically (but not necessarily balanced such that the kit is rhythmically coherent). The drums sound great (no they don't - they might if they were tuned down a bunch) individually, but even if the beat is not a mess, the drums are at such a level that the system can handle little else without running out of headroom and making a disaster of the entire ensemble. The other instruments are done in turn, again being tweaked at solo levels or above. Each is OK on it's own, but once you get two or more going, the system is out of steam. There's no room in the system for the vocals - and even if there were, no one who wants to retain their hearing would want to stay in the room.

A Slighly Different Approach

To dig oneself out of the aforementioned mess, when you are done with each instrument, turn it down all the way down. Now have the band play a song. Bring up the lead vocal and make any adjustments to the channel EQ that are required to optimize the vocal*. After all, the vocals are completely dependent upon the PA (if we ignore the blasting monitor wash, but we'll skip that gripe for now), and they carry the lyrical and primary melodic messages. Take the vocal to a level that's comfortable, but is rather too much to be balanced with the sound coming off of the stage from the instruments. Bring in one additional voice at a time, adjusting the level for a proper blend on a typical harmony. (If the vocalists are all using the same model of microphone, the EQ settings you used on the lead vocal should be an excellent starting point.)

* Most popular vocal mics have a LOT of proximity effect - they boost the bass and low mids tremendously when used up close. Close is of course a necessity when used with a live band. Pictured below is the response graph of a common vocal mic. At a distance of 1", 200Hz is up about 8dB. At a distance of 1/8", when the singer is really on the mic as we want, 200Hz is up 12dB! Since this also occurs with the instrument mics, it is common to compensate with the system graphic EQ. This is likely why most systems sound lousy when playing CDs and a bit peculiar with instruments that are run direct. I recommend that you optimize the system for CDs, and make the compensation for proximity effect on an EQ inserted into the vocal subgroup (or the channel EQ if you have to).

Now bring up one instrument at a time. Start with the one playing the secondary melody (the one that most complements the vocal melody). It should be sufficiently loud that the melodic relationship between it and the vocal is apparent and complementary. (Channel EQ may be required to get the instrument to fit in with the vocals**.) Then bring up the instrument playing the tertiary melody, again so that the melodic relationships properly complement each other. Keep adding instruments until all of the melody instruments are in the mix and properly interacting. (Remember that we do not necessarily want to make the instruments equal in volume, we are interested in the way that the melodies complement and counter each other.)

** When using the channel EQ, many tend to go to boost as the first adjustment. Owing to mic proximity effect, instrument resonances, speaker resonances, and room resonances, it is much more likely that you have too much of something, as opposed to too little of something else. (And it's usually between 80Hz and 250Hz.) Use the EQ to cut something instead. A case in point: at one gig, the acoustic guitar sounded great on its own, but was either buried in the mix or made a mess of things when turned up. So, we gave it something to make it "cut through" Boosted 3k - nope. 2k - nope. 1k - brash. 800 - awful. We switched gears and cut 200: hmmm. 300: what the guitarist was playing now fit right in with the rest of the ensemble.

Bring up the bass instrument. Although the bass is second in rhythmic importance to the drums, the bass has a melodic content and so is brought up first to interact with the other melodies, and also give the music "weight", "drive", "power", (insert your favorite and preferably non-anatomical adjective here).

Bring up the kick. The kick must punctuate the rhythm but not cover up the bass - or anything else. EQ accordingly. Unlike what one often hears, the kick drum is NOT the lead instrument in the ensemble! The snare (which is likely too loud already) and hi-hat must blend with the kick to form a kit that gives a coherent rhythm that sets the physical feeling of the song. The kit must not overpower the rest of the band. With the drums mixed in correctly, all of the instrumental and vocal lines will seem properly timed. (The less the audience has to work at feeling the rhythm, the longer they stay and the more they drink, which means more money in the bar owner's pocket, which means you get asked back, maybe for more money :-)

You are pretty much done now, except that the vocals may be getting a bit buried. The system should have sufficient headroom for you to raise the vocals up to where they need to be. Make any touch-ups that may be necessary. As the set progresses, keep in mind that when it seems that something needs to come up, it's just as likely that something else needs to come down instead. Don't forget that over the course of an evening of overly loud music, our ears "mush out": they become desensitized, especially to high frequencies. Booze and nicotine increase this tendency. This is why the band members turn up their already too-loud amps. If you don't remember this, your mix, or at least your frequency balance, will most likely be destroyed as the night progresses: "It needs a little more of this" then "a little more of that". By the third set you'll have added a little more of everything six times and will have a loud mess on your hands. My personal rule is no booze until halfway through the second set, and even if I'm just drinking water, basic levels and EQ are left alone after that point, unless a band member makes a change that necessitates a compensating alteration in level or EQ. (In other words, once it's working, don't fix it.)