When recording vocals, consider recording verses and choruses on seperate tracks.

Choruses tend to have different dynamics to verses, and singers automatically give a "lift" as the music intensity/tempo increases at choruses. These days, I either record the choruses seperately, or cut them out and paste them onto a different track so that I can apply different processng during the mix. Doing it the latter way is often better than seperate recording, because you don't lose the edge that the singer naturally takes into the chorus.

Always put a bit of reverb onto the headphone mix for the singer.

This is an old chestnut, but it does give confidence to the singer. However, I have also quite often not done this and used a different strategy with great effect, so:

Consider not sending a monitor feed of the singer back into her headphone mix.

It may seem wierd at first, but it has worked for me very well.

It is very important to LISTEN to what you are doing, but it is just as important, if not more so, to SEE what you are doing.

What I mean by this is that it is vital that you have good metering which displays the volume levels on each track, but absolutely on the Main Output Bus. Watch also any indicators on your mixers etc - if they go red, you are going to clip and distort. Most modern DAWs allow you to work at 32-bit float accuracy which means that you will never hear clipping. However, once you dither it down to 16-bit for output to a CD, it will sound horrendous if you have driven it to hard. Listening to your track does absolutely no good in this regard. Is it too loud, too quiet? - how do you know because the volume is subjective i.e. you can control it with the volume control on your output device. The thing that tells you is by LOOKING at the level meters.

Make sure that you leave headroom for Mastering

Everthing that you do to your track has the potential for increasing overall volume. The same is true of the Mastering process where it is likely that any or all of the following might take place: Multiband Compression EQ'ing Excitation Volume Maximising Mastering Reverb Limiting Assuming that the final product will be mastered to 0db (which it should be, generally), you have to give room for the mastering engineer (even if it is you that is going to do it), to apply processes which will boost volume. Accordingly, the volume of your track, after mixing, should be somewhere BELOW 0db. How much? It depends on what your track sounds like, and how much needs to be done in the mastering process. Personally, I aim for between -6db to -3db (although my acoustic stuff is nearer -3db). However, if I was to apply one golden rule it would be:
Never let the volume on the output bus of the final mix exceed -3db

Never use Reverb as an INSERT effect, use it as a SEND effect instead.

This one might be slightly controversial, but there are a number of reasons that I suggest this: 1. Instruments and vocals are generally monophonic, and generally appear at a fixed point in the listening space. Reverb, however, is stereo in that it appears right across the panorama, with reflections coming from all sorts of places. It makes sense, then, to create a stereo track just for the reverb, and send mono tracks to that. 2. By creating a unique Stereo Bus just for the reverb, you can save loads of processing power by having a single reverb on that bus (which is then routed onto the output bus), and route each instrument/vocal track to that bus. This could mean the opportunity to use a high-class convolution reverb rather than an algorithmic reverb on every track. 3. Using a common reverb is more natural because the reverb space is effectively common to all instruments and singers (unless you are using a lot of instruments in a small space). 4. You can control the amount of reverb added to each track without having to worry about changing the reverb settings by adjusting the SEND level. When using reverb in this way, the level/mix should always be set to 100% WET. If you have to use reverb as an INSERT effect, make sure that you adjust the mix level accordingly.

Always mix DOWNWARDS rather than upwards

What I mean by this is that you should avoid setting levels and then boosting the parts that don't seem "loud enough". It is always better in my opinion to set everything to unity gain (0db for audio tracks, 100 or equivalent for instrument tracks), and then adjust the ones that are too loud downwards. What I do on my stuff is to ensure that the vocals start at 0db, and then do a quick and dirty level mix by adjusting the instrument tracks down, always keeping an eye on the final output on the master bus. Once I am reasonably happy, I then apply effects to the vocal tracks, usually starting with EQ and then Compression. I then adjust the levels downwards/upwards as desired.

Set up a panorama using the stereo width

There are a number of really good guides on where to put stuff in the stereo mix (lead vocals in the centre, for a start, along with drums and bass), but it is more complicated than that. The human ear actually imagines a 3D space somewhere in front of the ears. As well as being able to determine whether a sound is to the left, the right, or in the centre, it is also able to imagine depth, and frequency. Taking the latter first, instruments with a high frequency such as a piccollo will appear to be at the top of the 3D space, whereas low frequency instruments like kick drums and bass will appear at the bottom of the space. What about depth? That's also easy, because the human ear determines depth by the relative volume of the sound, and we can use this to our advantage. Say, for instance, that we are arranging a traditional orchestra where the cellos and violas both appear to the conductors right, but the violas are further away, we could pan both moderately right, and then reduce the volume of the violas relative to the cellos. Getting the frequency space right is more difficult, and is a task that is generally better applied to the mix as a whole, but it generally means getting a reasonable balance between high and low frequencies, and that is done with EQ. But, that's another story for later.........................

Let Someone Else Do It!

Seriously, mastering is the final thing thathappens to your precious music before it is fit to be listened to by others, or before it is sent to potential publishers for review. At this stage your music needs to sound as good as it possibly can do, and if you choose to master yourself, you may be hindered by your own pre-conceptions. At the very least, allow some time before the final mix and the mastering process. Of course, there is the other issue that mastering is traditionally seen as a black art that can only be accomplished by those few who know how to do it properly.