First of all, there’s a difference between Mastering and Premastering: Mastering is a process, that matches premastered Audio material with the Demands of unified Pressing-Copying-Standards. Premastering is the process of finishing and polishing-up your mixdown… The tips I’ll give in this article are suggestions, as there is no general rule or recipe how to make things sound good. The first and very basic rule one should follow is, that nothing that sounded like crap before you start premastering will sound brilliant in the end. Premastering is about dealing with slight “mistakes” of your material and about smoothing or bringing out some edges, but nothing more. The “colour” itself should not be the target of this process, but the product of your arrangement. If the sound doesn’t come close to your idea by now, do an other arrangement, change something in your track – Premastering won’t help you. Nuf babbling, let’s start!
Premastering isn’t the time for sitting in a dark room. Get you some daylight, some fresh air and none of that glomming green stuff. Your ears are a very VERY complex set of instruments and easy to disturb, to numb or to fatigue.
Convert your Audio file into the bitrate, your program is working with (for Wavelab this would be 32 bit) Get rid of your DC-Offset (DC offset is a net, no-signal DC voltage measured at the amplifier output, which should be zero, but isn’t. DC is unhealthy for speakers as it produces little nasty clicks… in a nutshell: the gap between the Zero db of your file, and real silence) Normalise your track to something around –0.5db (minus! 0.5). Now there’s some headroom left, before the Peaks of the loudest signals will touch 0db – that means you’ll have some space left for further dynamic. Make a backup of your Audio file and save it as “Title-Firsttake.xxx”. You should keep one file, that remains untouched. So, in a worst-case-scenario of loosing, destroying or overwriting the file you are processing by accident, you’re save.
3. Some cosmetics:
If necessary, denoise and then declick your material (if you work with samples, denoise and declick them before you use them in your arrangement. This might save you from this step and better samples do make a big difference to the mixdown)
Then set up one first Limiter. In most cases it does make sense at this stage, to limit the peaks, to get some extra 4db! It is important, to set the threshold properly, so that just the highest peaks of your material will be affected. (Limiter? Looking at your wav-file, you’ll see Volume-peaks, that are far above all others. With the limiters threshold, you’re able to set a max-level, above which the limiter will level down all signals. So the maximum- or peak-volume of your track can be decreased and widens the headroom left for further dynamics.)
4. Cleaning the mess: EQ
EQs are THE tool for phat, creamy, transparent, detailed sound! You can never learn enough on that topic. The way to success isn’t via fancy pseudoaccoustic-pro-tool-Plugins, these normally fatten the mess, but accurate EQing.
Some basics about Frequencies:
Around 60Hz is the low end of your hearing-ability. Underneath this edge, your ears will not find any valuable information. (But if it is loud enough, you can feel it!!!!) This frequency-range should only be used sometimes, to get an effect, but not humming all the time.
The range, where you set up a proper, low Sub-bass is between 60 and 80Hz. But notice, the deeper the sound is, the less can you separate different sources – this is why your bass sounds indifferent and rumbling, as soon as it interferes with e.g. a bassdrum. This effect is called “masking” – the louder signal suppresses the more silent one – the solution is separating the signals in your arrangement, before you do a mixdown: Set your bassdrum between 80 and 120Hz. If you like some more power on the release of the drum, raise it’s bass around 80Hz and lower it’s range at 200–400Hz. Above that is where your bassline lives. If you like, let your bass slip down to the very bottom (30Hz) sometimes, to make the crowd on the dancefloor feel cosy, but don’t overdo it!
Is a very tricky area! If your whole setup sounds dull, even if you have maximum presence in the high frequencies, 200–400Hz is the area you should worry about. Lower this range, and you’ll receive more transparency and presence in the whole panorama. Much of accidental “LoFi-Charme” can be avoided by gaining or lowering this range.
Is the range, where things get this telephone-effect (don’t know, how to describe it properly, but you’ll know what I mean) kind of a resonating, flat sound. If your mix sounds like that, lower this range a bit in a wide band.
Around 2–4KHz our ears do something wicked: You always recognise, this frequencies louder than others, that have exactly the same volume. Lower signals in this range, to “widen” and harmonize your mix.
Is where snares, HiHats and hissing noises get their presence. This range makes a mix sound aggressive, hard and “in the front”. High frequencies above 6 KHz are mentioned to be extra sharp and cutting. So if you get sounds harder then mend to be, lower some narowbanded gap here, to avoid the mix to get too hard, or to interfere too much with the area around 4–5KHz.
You’ll notice a strange thing: This is the range for silky presence and nice transparent sounds. Raise this area, and you’ll hear things more in the front and fresher, lower it, and things step back into the background. If you listen to this frequency-range solo, you’ll recognise it as kind of noisy, but in the whole of a mix it makes sounds spacy and silky wide. I love stereo-effects on this part.
Just turn it up! Bad samples have lost their presence long ago. So there you’ll need to turn things up. The upper edge is at 15KHz. Get a dog for reference listening above that range hihihi…
EQing is a bottomless science. But as far as I know, avoid to raise frequencies, if you can lower the rest. Don’t worry about getting to silent, as you can gain later. So as a general rule, use EQs for lowering some carefully selected frequencies, but don’t chop your mix into slices. Find the critical areas, the resonating ranges and work precisely on them. You should set a Lowcut at 30Hz. To all the bass-addicts: There is no relevant information beneath 30 Hz. Gutquirling Sub-bass happens somewhere else! But if you can’t resist the voices in your head, push up (~+3db) a narrow curve around 65–70Hz. Then lower a second gap narrowbanded around 130Hz. Now you should have separated very low basses from punchy drum- or synth-basses. Now try to raise a Hi-Shelf around 10KHz to freshen up the top-end. In 90% of all cases you’ll now be able to forget about Enhancers or Exciters.
Save and backup your material!
5. Reference listening:
To get an Idea, where you’ll want to go from here, listen to one of your favourite CDs on your monitors and try to get the gist out of that sound (do not use mp3 for reference-listening, as most of the mp3 are ruined with bad compression-algorithms). Then listen to the Audio material, that you’ve just worked on – got it? By now, the “colour” of your material should be close to your idea of it (don’t worry about lacking stereo-panorama jet – this will be added later on. If this disturbs you, listen to both in mono-mode). If still think, you’re far away from your idea, then go and make a new arrangement!
Watch out, that all the sources for reference-listening have the same
The sense of reference-listening is to avoid getting used to a specific sound. If you listen to your track for an hour, you’ll never remind how other tracks sound like. The second thing is, our brain gets used to specific sounds. If you had a good night of sleep and “fresh” ears, you’re OK. But if you listen to the same thing over and over again, you’ll notice the following: You feel the need to add some bass. Louder parts, or parts, where you added more boost, sharpness or volume start to sound better. Why? The more tired your ears get, the less details you’re able to notice. But as details make things more interesting, your tired ears tend to like more of something that gives them a difference. So if you notice that, take a walk! Get some rest! (try this – turn up the high frequencies and listen to your music for 5 minutes. Then turn them down again, everything will sound dull and low.)
First rule: Know the Traps and weaknesses of your ears and
Second rule: Slowly, gently and careful!!!!
Now that you’ve set the sound and the equalisation to the mixdown, it’s time to care about the dynamics. The No. 1 reason for a boring result is a lack of dynamics. If the Hats’ or the snares’ or whatever sound does not change slightly during your song, it will stay boring and flat. O.K. – Listen to your track several times, play it loud and play it silent, soon you’ll notice several things, that could be optimised… All in all, your mix still has many parts, where too much sound interferes with too many signals? Search a part of your mix, that does sound overloaded, where you cannot clearly say, which sound is caused by which instrument (bassdrum or bass, is it the snares reverb, or part of the synthi?).
As EQing is for separating and pointing out frequencies, for a more detailed panorama, compression is for smoother, more elegant dynamics.
What a compressor does? No one knows, but you’ll notice the difference… in a nutshell: A compressor alters the signal in it’s dynamic. If you imagine the wave of e.g. a solo bassdrum, it’s curve raises fast to it’s max volume and makes this punchy slap-sound. Then it decreases again to zero with a low bass-thud-sound. The speed of this curve defines the dynamic, and with a compressor you can modify this curve. Normally you use Compressors to lower the louder parts of a signal, to “flatten” it’s curve and then – by gaining it again – make it’s whole sound louder.
I would recommend multiband compressors for premastering. They allow you to set different compressions on different frequencies.
Now follow a simple pattern and trust your ears: select a frequency-range with interfering, competing signals on your multiband compressor and play this range solo. Now set the threshold, till both signals are affected. How long is the decay of the signal you are working on? This is important for the attack- and release-values of your compressor. I would recommend the “finding by sliding” thing… If you’ve managed to match the position of the signals interference with the attack-time of your compressor, you’ll be able to set the Ratio. There’s lots of theory about finding the right ratio, but all I can advise is, listen to it over and over again, and just try, where your sound starts to get tight and powerful. There are some good reviews on compression on this board already – as a hint…
7. Final steps
We haven’t done much jet, but you see, it takes quite some time to get
Does your track sound better? But does it lacks some wideness, some hall or whatever? If you need to, now you can play around with stereo effects and stuff like exciters or enhancers or reverbs or whatever, watch out, as these may ruin your mix quite fast!
Set a final Limiter, best would be kind of a maximiser, to regulate the dynamics.
Normalize your track again, but again not above –0,4 to –0,2 dB (often you don’t need to normalize your material again, after all this dynamic-procession)
Convert your track back to 44,1 Kbps/ 16Bit, using a good dithering-tool.
Save your thing, lean back and enjoy…