Mixing music might seem like a black art, but here's a simple approach to help the newcomer get their tracks together.
This month's column leans toward the new 'audio recording enthusiast' who is trying to figure out how it all works. I have been there, although it was back when stereo was a new gimmick.
After you unpack the boxes, connect all the gear, look at the covers of the manuals (nobody has ever actually read a manual, right?) and try to start recording and mixing, how do you figure it all out? Like learning to play a guitar, learning to paint, or tuning the fuel injection on a Ferrari, first you need to be pointed in the right direction, and then you need experience.
There are hundreds of books to point you in the right direction. There are at least 10 books covering each major DAW, including Pro Tools, Sonar, Cubase/Nuendo, Digital Performer, Adobe Audition and Logic. There are at least that many writings about recording and mixing in general, and a few books try to teach mastering. For some people, however, reading the books doesn't help much more than sleeping with the book under your pillow. Along with reading, you need to have some practical experience.
As far as equipment goes, you can learn about recording and mixing with what you already have. It doesn't take much of a computer these days to get the ball rolling. If you can't afford the software, there are dozens of free multitrack recording programs available, as well as demo versions of most of the high-end programs. You can actually start mixing before you start recording. There are multitrack versions of songs from Nine Inch Nails and other groups on-line for download. There are some books that include multitrack material to mix from Waves, and Charles Dye's Mix It Like A Record DVD.
The biggest hump to overcome when mixing is that everything sounds better when it is louder. Everything! You can have the crappiest mix on earth, but if you play it loud enough, it will sound great. The worse the mix, the louder you have to play it.
You see this all the time in the studio when the mix is almost done and the band shows up for playback. The guitar player comes over and wants the guitar up just a little. Then the bass player sneaks over and turns up the bass. The drummer is not far behind, complaining about the lack of punch in the drums, and up crawls the drum group. After everyone finishes, the mix is right back where it started, but louder. The band is happy, approves the mix and leaves. The engineer turns down the master fader to get the overall level back to normal, and prints the mix.
If you can change the level of an instrument in the mix by a tenth or two-tenths of a dB and you can hear the change that you made, the mix is getting pretty good. If you can change some other instrument by that amount and not really hear any difference, then chances are that the instrument you are playing with is not where it should be. The problem may be down to level, or EQ, or compression, or reverb, but it is not quite ready for prime time. I will discuss all four and point out things to watch for.
Major record labels are gearing up for download of your favorite multitrack material. The record labels are actually releasing these multitracks to copyright the content. A copyright is not valid in the US unless the material is distributed publicly. The multitracks in the vaults are, therefore, not copyright material. The melody and the lyrics carry a copyright, but that is for the writers and publishers. The record company has a copyright (P) on the physical CD, cassette, or vinyl album that is shipped, but that only covers the final two-track mixes as they occur on the finished product. No protection is offered toward the multitrack recordings. Interesting. So, if they offer the multitracks for download, and leave off the lead vocal or melody line, they don't have to pay anyone anything. Artist gets nothing, producer gets nothing, songwriter gets nothing... Hmmm. Anyway, there is going to be lots of material to test your mixing skills, for a few dollars more.
UK laws are slightly different, though I doubt that record companies ever send copies of the multitracks for registration with UKCS. UK law is better than the US in that all you need to do is produce the work for it to be copyright; the root of the US record company stance is that you cannot copyright chord progressions or tempo, so without the melody and lyrics, there is nothing unique about the work. An example would be Tina Turner's 'Private Dancer', written by Mark Knopfler. When you heard it, you thought it was Steely Dan's 'FM' track. The bass line, the rhythm, the tracks seem to be uncannily similar to 'FM', but the melody and words are different, so no case.
Think of any instrument as two separate components: the low end of the instrument and the high end of the instrument. As an example, let's take an acoustic guitar. The highs are important because they give the attack of the note or the raking of the strings that extra presence to cut through the track. The low end creates the body of the sound, the warmth of the instrument in the track. If the balance between these two halves is not right, then you will never get it to sit in the track correctly. There will always be places where it seems too loud or too soft.
Start with the attack of the notes, or the raking of the strings in the case of a rhythm acoustic part. Change the level of the acoustic guitar until the high end sounds right in the existing mix. Get the level to where you can hear a level change of two-tenths of a dB. Now stop. Start with a parametric EQ set to a low Q of around 2, giving a very wide, smooth curve, and the frequency set for around 250Hz. Move the gain of the EQ up or down to get the low part of the guitar into the right relationship in the mix. Try to think of the low end of the guitar as a separate instrument. Get it to sound good in the mix. When you get the low end of the guitar in a place where you can hear a two-tenths of a dB change in the level, you are doing great.
Now you are going to do one more thing. It is usually the low end of all the instruments adding up that makes the mix sound muddy. My rule is, if you can't hear it, you don't need it. Turn on the lowest band of the EQ and make it a high-pass filter, or use the high-pass filter if it operates independently. Start with the frequency all the way down. Leave all of the tracks playing, with the acoustic guitar in the mix. Start raising the frequency of the high-pass filter slowly until you can hear a change in the sound of the low end of the acoustic guitar. Look at the frequency on the high-pass filter. Now reduce the frequency by about 15 percent. That means if the frequency reads 200Hz, then move it down 30Hz to about 170Hz. If you do this to each of the mid-range instruments, it will clean up the muddiness problem in your mix.
If some notes just stick out or get lost and EQ by itself doesn't help, then maybe you should use a little compression. Notice the key word: a little compression. Of course the best way to set a compressor is to listen to the results, make an intelligent judgment, and twist all the knobs until it sounds right. Or, you can start with an easy rule of thumb that can get you most of the way there.
Set the ratio to 2:1. Depending on the compressor, use either the input control or threshold control while the music is playing, and lower the threshold (or raise the input gain) until you see the gain-reduction meter jump into action. Keep adjusting until the gain reduction peaks at -3dB to -4dB. For now we will just leave the attack and release values at whatever the defaults were when you loaded the compressor.
This is now a pretty good starting point for compression. It is not even going to really sound compressed. You now have a little more control over the dynamic range of the instrument without hurting the sound of the instrument. Now re-adjust the level of the instrument to get it settled in the track. When you can start hearing small changes in level, you are in the right place.
Sometimes EQ, sometimes compression, and sometimes a combination of the two may be necessary to get the desired results. Try the EQ first, as this will maintain the greatest amount of dynamic range in the instrument. If you can't quite get it, try a little compression.
We already have a pretty good mix going, but how do you know if it is the best possible mix? Think of it this way. It is always easier to tell when something is wrong than when something is right. This is true of instrument levels in a mix. While listening to the mix, one at a time change the levels of each instrument in varying combinations. Turn the bass up 1dB. Does the mix sound worse? If the answer is yes, then turn it back down. Turn the bass down 1dB. Does this sound worse? If the answer is yes, then the original level was right for the bass. Now try one of the guitars. Then the piano. If an instrument sounded better at a new level then leave it there and start back through the cycle.
This can be a time-consuming task, and you should take lots of notes so you can get back to previous settings, but by the time you get through the process and have eliminated everything that sounds worse, then you have the final mix. Cool, huh?
Reverb is the last thing to add, because it depends on all of the other levels being close to the final ones. Normally, when you add reverb, your initial impression is that it sounds more three-dimensional and lush. This is true in a way, but it also starts sounding like the whole band is playing with you in the shower. I don't mean playing with you, I mean playing along with you on their instruments. I mean... Never mind.
Solo each instrument that has reverb added, along with the reverb. Turn the reverb down until it no longer sounds like a big reverb, but more like real-world ambience. 'Yup, sounds like the acoustic guitar was recorded in a nice-sounding room' should be your goal. There is a place for special effects if you want, but make sure there is room in the track by backing off on the other reverbs.
One more reverb trick I always use, both with reverb plug-ins and external reverb hardware, is to insert a one-band EQ after the reverb return. Set the EQ to high-pass filter and set the frequency to 250Hz. Reverbs usually generate some of their own low-frequency artifacts that are not part of the music. Also, low-end reverb just makes the mix rumble. Limiting this unnecessary sound makes the reverb accentuate the music in the middle of the spectrum, and the nice splash on the snare and the vocal, without adding mud.
This is by no means a complete course on mixing, or the definitive answer to all of your troublesome tracks. This is meant to help someone new head in the right direction. Use these tips and expand on them. See what works for you.
If you do things differently and make great-sounding records, then perfect. I want to know how you do it. Share your methods with others who are trying to figure it out. I like buying CDs and listening to new bands. What I do not like is bad-sounding CDs. Put together someone who knows how to make great-sounding records with a band who know how to write and play great songs and you've got something!
Roger Nichols has been professionally involved in the music business since 1968, working as a staff recording/mixing engineer at ABC Records and Warner Bros before becoming an independent engineer/producer in 1978. His work with Steely Dan in particular has led to a string of Grammy Awards and nominations, including a Best Engineered Album award for Two Against Nature. An advocate of digital recording since 1977, Roger designed and built the first digital audio percussion replacement device and has lectured on digital audio around the world.