The most common type of mix is a simple stereo mix where the various channels of your recording are combined into two channels represented to the listener as two speakers or headphones.
Multi-track mixes are beyond the scope of this workshop but many/most films use what's referred to as a 5.1 mix: center, left front, right front, left rear, right rear and a subwoofer (the .1). Most modern home music systems, and many computer sound systems, provide built in decoding of stereo more than 2 channels (some create as many a 7.1 channels from a stereo recording). At the very least, a subwoofer channel will be extracted from the left and right channels.
Traditionally, mixing is done by feeding tracks from a multitrack recorder or live mics into a mixing board and recording the stereo output of the board onto a seperate master recorder (a seperate tape machine). These days it's much more common to mix a song using a virtual mixer in the recording software than to use an outboard mixer. Typically, the software will allow you to perform a "render" or "export" operation which will combine the tracks into a stereo mix.
In the early days of multi-track recording, all level changes were made by hand in real time as the tapes rolled. Some of the more expensive systems started adding different types of "Automation" features that allowed the mixing board to record slider movements and other mix changes allowing much more complex mixes to be accomplished. Automation was eventually even available on moderately priced mixing boards but once recording moved to the personal computer, automation became a standard tool. With computer based automation you can record slider movements in real time or draw the volume changes right onto the channel on your computer screen. Now you can actually see the mix as well as hear it.
Though it's used in all types of music, almost every track of a pop song uses some level of compression to help keep the overall balance of a song stable. Most instruments have certain notes that are much louder than the others making them jump out of a mix. Sometimes this is good but usually it's bad. You'll aways apply some compression and limiting to your final mix but compression of the stereo mix applies to all instruments which is also usually bad. Without compression, you need to set the level of an channel so that it never takes over the mix. The result is that the channel is too low through the whole song except when that one loud not plays. You could adjust the level by hand but you won't always be able to respond to changes fast enough. You can get a lot closer using automation but it's very time consuming and the results aren't always that great. Usually, you'll use a combination of compression and automation: compression does the heavy lifting while the automation puts the finishing touch (you do want that solo to come up in the mix).
Presumably, your music will be played back through all types of sound systems including ipods, computer speakers, car stereos, boom boxes, televisions, home theater systems, club PA systems, etc.. so you want to test your mixes on as many of these mediums as you can. In the studio, you'll be mixing through high quality speakers that produce a fairly accurate representation of the recorded sounds but almost everywhere else the recording will have bumps in the frequency response that can change the sound dramatically.
Your bass guitar or kick drum might sound great through your studio monitors with your subwoofer, etc.. but it will just disappear on systems with no bottom end. It's best if you can keep some midrange elements to every instrument if you want them to be heard on all systems. The mix will definitely sound different but you'll at least have something taking its place in the mix.
You'll also want to check your mix at different positions in the room. You may still want to hard pan many of the channels but you want to know what the mix sounds like if you are sitting next to the left speaker.
Ideally, you want to have more than one set of speakers in the area you use to mix your music. You'll want at least one very good sounding set of speakers but it's helpful to have a set of cheap speakers around so you can hear the difference. Keep a boom box around and have a way to listen to your mixes through it (preferably without burning a disk). The Stec Records Speaker page has our latest recommendations for speakers (go for the Genelec speakers if you can afford it but the Klipsch sound extremely good for little money).
The speakers should be evenly spaced on either side of your computer monitor at the same height.
A subwoofer is recommended if you can afford it or if your configuration allows (if you are using digital speakers, you may not be able to connect up a subwoofer). You want your mixes to sound great through a cub or other full range sound system but the only way to be sure is to have a subwoofer while you are mixing. (Remember that most listeners won't have subwoofers so also be sure it sounds good with the subwoofer off.)
If you are mixing and matching speakers and subwoofers, be sure to set the cut off correctly so you don't end up with a bump or hole in the frequency response.
A large number of listeners will only hear your music through headphones (in many cases, cheap ones) so be sure to test the mix through them. It's the same problem you have with crappy speakers: some frequencies jump out or are lost, but headphones also put the mix in the listeners face. Subtle details that are invisible through speakers can't be missed in headphones.
Tip: Be sure to shut off your speakers when you are listening through the phones. You'll get a false impression of the low end on the recording.
The physical room in which you mix your music can have a huge effect on the sound and quality. Rooms with a lot of flat, hard surfaces can make a mess of the sound. Imagine the walls of the room are like the sides of a pool and the sound waves are like the ripples. You'll notice that the initial wave/ripple is almost perfect until it hits the walls and starts bouncing around? The same thing happens with sound effectively distorting the sound so you'll want to put some sound-absorbing material on the walls and, sometimes, ceiling. Ultimately, it really doesn't need to be pretty but usually needs to be soft. Sound-absorbing foam panels can be used but even heavy curtains, blankets, carpets and even.., yes, towels. If you are on a tight budget, you can probably avoid professional foam panels but they do work well and look a lot better than your dish towels.
It's usually good to have one or two untreated flat surfaces (wall, ceiling, etc..) but any more than that and you'll have trouble. As you add acoustic treatment they say that the room becomes more "dead" sounding. A room with a lot of flat surface is usually referred to as "live" sounding. You can add "live"-ness to a recording but can't take it away so it's best to have a room that is to dead sounding than to live sounding. You'll usually be adding a reverb plug in to create a "virtual" room anyway. The sound of a small live room can make it impossible to get a good overall room sound even with a reverb plug in.
Play some music that you really like through your speakers and, if it sounds good, you'll probably be ok. Again, dead-er is better.
Virtually all professional music recorded on a computers use effects plug-ins to spice up the mix. These are some effects plug-ins you are likely to use while recording on your computer:
You'll usually want a reverb on one of the aux sends so it can be applied to any channel. Some effects you'll put in line with a channel but others can be available globally. It's sometimes nice to have a "house" slap echo that you can use on lead instruments and vocals. Remember that each effect you add uses up a little more of the computers horsepower. This is becoming less and less of a problem but it is something to be aware of. It's better to have single slap or chorus on an aux/effects send that gets shared with all tracks than a seperate one on every channel. Somethings like compressors and EQ's you'll put on every channel but amp simulators, modulators and delays should only be added if needed. You also have to be able to keep track of all the parameters/properties of these effects in you mind's eye so it can be a little overwhelming at times. Especially busy arrangements. The rule of thumb is to use less effects if you have more instruments.
Background image by Pauline Moss. at http://paulinemoss.deviantart.com/art/Stars-307119758