Mixing and mastering
What to do and what better not to do...
I am frequently asked "how do I make loud tracks?", "how do I *master*?".
This little article might serve as a general brainstormer. It is just reflecting my personal opinion on that topic. Yours may be different ;)
Making powerful tracks begins with the writing of the song. You should check at least this:
- Do my instruments and their specific sound deliver what I am looking for? Usually, most sythetic stuff comes with a rich and even tonal spectrum and might sit good in the mix from the start (as long as the sound is not too 'fat' like, for example, a lot of pads are). As opposed to acoustic instruments which have a unique sonic fingerprint with lots of resonances and interact with the room acoustics. Find out about peaks & dips in the spectrum and take things like delays and envelopes into account.
- Does each one support the overall sound or is one (or more) affecting it in a way that it turns your mix into mud? Think of particular frequency regions and the masking effect. Play with your mixer's eq to find out about this. As a general recommendation: negative eq-ing is better most of the times, so whereever something sounds disturbing, set the eq right there and dig a hole. This generally sounds way better than boosting the region around it.
- Volume is not punch. Free yourself from the common preconception that any CD has to be as loud as possible. I guess you are making music that's supposed to be played on conventional stereo systems. Well, each of them has a volume knob, really :) Plus, a potential radio station has compressors and limiters hot enough to boost your track. You won't win anything if your music is already squashed to death. Less is more.
- Natural dynamics: There is a reason why instruments are actually played at different volumes. There is also a reason why our ear has such a tremendous dynamic range. Today's music electronics are good enough not to insist on constantly high volume levels. Noise shouldn't be an issue anymore. So let's consider dynamics as a friend, not an enemy.
Whenever I start mixing one of my songs I clean up the desk and unplug any inserts and fx, so that I only hear the plain instruments. That helps to decide if things are already fine. You've got to get the groove & punch at this time, otherwise think of re-arranging the song and choose different instruments. If the sound seems okay, go for dynamics and effects. Stop at the point where it starts to lose tightness and definition.
- Think about transients. A lot of our perception happens just on the starting of an envelope and the signal's peaks. Don't just flatten them. If you do so, you are not only evening things out, you are also altering the waveform, thereby approaching a square wave which sound quite fatiguing in the end (because square waves consist of odd harmonics, with pretty high level).
- Don't mix in the 'solo' mode. Those buttons should stay untouched in the mix. Try to tweak the sound with everything turned on one after the other. It does not help you that you know that this baseline sounds fat when soloed. Mixing is to make 'room' for every instrument. It's just like painting.
- Don't plug any compressor in the sum bus during mixdown. Exciters/enhancers are a no-no, too. You'll betray yourself and won't be able to discover the true dynamics and frequency response.
The metering instruments on your recording equipment can greatly teach you how to keep levels under control. I am always using the meters on my DAT which have fast peak characteristics. After some time, I've gotten a tight grip to transfer the measurements to my perceived dynamic range and vice versa. I'm sure this applies to any metering instrument if it is halfway accurate and one gets used to it. Some are using VUs only for this purpose. It's a matter of taste.
- If you work with compressors and limiters on finished mixdowns, try only to 'polish' things and make minor adjustments. Don't make it just louder, make it BETTER (noticed that there is no compressor or limiter with a knob labeled 'better'? :D)
While the industry is constantly telling us we should aim at loud levels, a lot of people are making the same mistake over and over (me included sometimes). The louder track is more convincing to our ears and sounds better in the beginning. You have to force yourself to listen closer and make comparisons to a quieter and less 'stressing' version.
Using heavy compression on the sum signal *can* sound nice or interesting, but it does not guarantee that it improves your recordings at every time. Furthermore, the better the mix and its tonal and dynamical balance, the less will it be demanding for overall compression or limiting.
When you feel that something's not properly balanced, the drums are sitting too much up front, the bass produces boominess or the singer ssssspits right in your face, consider re-mixing the song and use proper compression on the individual tracks.
Quite often a simple hard-knee compressor works good for drums, like the mda freeware stuff.
- If you're using a (look-ahead) peak limiter, do only limit up to 3dB, not more. Sometimes, even 1dB is audible enough to introduce too much artifacts. You may replace limiting with tape-like saturation (like with the output stage of endorphin) if you perceive serious pumping or a general 'grainy' sound. But take care not to dial in too much 'heating'. Burnt offerings...
- Don't mix with headphones. Use them only to monitor single sources or to detect noise, hiss, or rumble. Phones are often very misleading when it comes to the question if all instruments are well balanced. Vocals tend to be mixed too loud, while constant and 'wide' signals such as pads are often too low in volume.
Many headphones have big difficulties with transients: you might think the drums need some extra 'smack'. There's a chance that playing the mixdown back on your normal speakers will reveal too much of transient attack (which is prone to slew-rate distortion on most playback systems anyway).
- Very often, mixing and editing tracks is a recursive operation; it might take you doing things over and over again. Don't get frustrated, and take your breaks as the ear is very easy to cheat and your perception is largely dependent on the state you're currently in.
- Think about what is the weakest device in your chain. Quite often, it's the monitors, your room acoustics, cheap cables, inappropriate miking techniques, wrong microphones and that stuff. For quite a while I was thinking I couldn't mix until I found out it was my poor and noisy mixing desk and crappy monitors (which were even at the wrong place of the room). Garbage in, garbage out :)
Finally: It's always good to have a second pair of sane ears listening to your music during the process. That's one of the main reasons you should definitely think about hiring a professional mastering engineer whenever you plan to address a large amount of people (and have the money, but it should be worth it most of the times). The mastering process can't be done with just a box of software tools and the will to make things 'loud & fat'.
- years of experience
- an excellent listening environment (speakers & room acoustics have to be perfectly matched)
- high-class equipment specialized for mastering purposes (the 'creative' process was long before...). You'll also need a certain variety of gear to choose which one is best for a particular job. None can serve all
- some distance from the tracks (you've been involved in the process for too long to tell what's right or wrong)
There are a lot of sources on the net talking about mixing and mastering with a lot more tips for you at hand. For example, try out prorec.com (good articles & reviews plus a forum) or the fantastic site of mastering guru Bob Katz at digital domain. You can also take a look at John Westman's mastering pages at johnvestman.com.