Another tutorial I found:
how to get that massive dance sound
It's quite evident from surfing around the web that many musicians are making dance music. And a common concern amongst these musicians is that their tracks don't sound as big, loud, and punchy as tracks by their favourite artists. Though many forum participants appear to believe that there is a single reason for this, such as the model of compressor they're using, the real answer is actually a combination of factors. If you're going to get your music sounding as loud and punchy as possible, you have to look at the sound sources themselves, the use of processing and effects, and the way in which the track is mastered.
Fixing It In The Arrangement
Many problems with a track's apparent size can be traced back to the arrangement, and problems such as these aren't easy to solve if you leave them until the mixdown stage. For a start, it's worth realising that creating a 'big' track is about having tonal contrast, so it's no use combining big kick and bass sounds with big snare drums, big pianos, big pads and big vocals. Each instrument you add to a mix will wrest a portion of the available headroom and bandwidth from each of the other instruments, making everything a little smaller. If you feel that you want to add another part to make your track more interesting, why not resist this temptation and concentrate on making your existing parts more interesting instead. Your track will almost certainly sound bigger for it.
It's not just how many parts you put into an arrangement that makes a difference to the perceived size of the track, it's also important where you put them, both in the time and frequency domains. For example, it's no accident that many powerful dance tracks alternate a four-on-the-floor kick drum with an off-beat bass sound. These two parts would have to compete for the low-frequency headroom if they occurred together, so programming them to always happen in different sections of the bar means that they can both be mixed much higher in level.
Similarly, it's worth choosing the individual sounds in your arrangement so that they occupy distinct, and separate, regions of the audible frequency spectrum. If you want a bass sound, a rhythm guitar sample and a lead monosynth part to co-exist within your track, it's worth making sure that they occupy different frequency ranges — if not, then experiment with transposing their parts around until you find a combination which balances better.
Finally, the size of a track is best emphasised by comparison. In other words, try to allow for some 'drops' in the arrangement, where a number of parts drop out or become more sparsely programmed — when the full arrangement returns, it'll seem much louder by comparison.
Turning The Bass Up To 11
In the context of dance music, the things most commonly associated with sonic size are the bass sounds. The bass sounds are simply those parts within your arrangement which have low fundamental frequencies, typically between 40 and 150Hz. Obviously, this will normally mean bass synth parts and kick drums within dance music. While it might be tempting to think that simply boosting the low frequencies of a bass or kick drum sound will make it sound big, this isn't necessarily the case — it will all depend on the type of playback system on which your track is heard. For example, a bass sound with a very high-level 40Hz pure tone might threaten to part you from your recently ingested curry when played over a high-power club PA system, but it will be virtually inaudible on most domestic stereo systems, as these roll off quite sharply below 80Hz or so. To make a bass sound seem loud on a system with limited low-frequency response, it will need to contain harmonics above the fundamental frequency that can be reproduced more effectively. It is for this reason that many strong dance bass sounds rely on square and sawtooth waves, both of which have considerable frequency content above the fundamental. Similarly, a kick drum will require some higher-frequency element if its attack is to be discernible on a domestic playback system. On the other hand, if you choose a bass sound that is too harmonically rich, it may take up too much space in the mix unless you keep the bass line very sparse, so filtering some of the top end out of a harmonically rich bass sound at the mixing stage can also help.
|
The waveforms of two different real kick-drum sounds are shown above. Notice how their level envelopes decay comparatively quickly.
|
Creating a more harmonically rich bass sound from a bland one can simply be a case of firmly applying some EQ, but there are also other ways to give a bass sound more high-frequency clout. Perhaps the best of these is to apply some sort of distortion to it — not only does this make the sound richer in harmonics, it also capitalises on the psychoacoustic effect whereby distorted sounds are interpreted as being louder than undistorted sounds at a similar level.
There are plenty of ways in which you can create distortion. Most multi-effects processors provide distortion algorithms, and there are also numerous software plug-ins available. However, no two types of distortion will be the same, so it's worth really hunting around for the types of distortion which suit you. For a start, don't disregard cheap guitar pedals, or even the more expensive physical modelling preamps that are there on the market. And why restrict yourself to something that's meant to distort? A little too much level fed into a recorder, preamp or converter can create unique distortion sounds which might be ideal. Sampling your bass sound at a reduced bit-depth might even do the trick, increasing the level of quantisation distortion, though the resultant grunge won't be nearly as musical as analogue distortion — not that that has ever stopped anyone!
If you're interested in pursing the subject of controlled distortion further, it's worth realising that distortion can also be applied in a frequency-selective manner by certain processors and plug-ins. One example of such an algorithm is Steinberg's Quadrafuzz plug-in. As its name suggests, this virtual gadget splits the audio spectrum into four separate frequency bands and treats each separately before recombining the results. The outcome is a far less messy distortion than you'd expect from a regular fuzz device, and it also has an additional benefit. Because it increases the average signal level, it can make loops, drum sounds, bass sounds or even complete submixes sound very big and exciting.
If the distortion you're using is extreme, such as that produced from overloading an A-D converter, the top end can get quite ugly, though this is, admittedly, often useful at an artistic level. By all means wield your high-frequency EQ to sculpt the distortion sound — I find that a variable-frequency 12dB/octave high-frequency shelving filter is usually the most useful here. Alternatively, a speaker simulator intended for guitar use will usually also be very effective for smoothing things out.
Loud, Strong & Very, Very Long...
In addition to its harmonic content, the subjective loudness of a percussive sound is closely related to its level envelope: how long the sound lasts and how the level decays during that time. In your school physics class you may have learned that energy is simply power multiplied by time. This means that a bass sound with a piano-style envelope is likely to sound subjectively weaker than one that continues at or near full volume for the duration of the note. Changing the envelope of your bass sound, by delving into the envelope menus of your synth or sampler, can therefore help beef up the sound without making the note any longer. This effect can also be achieved within a sampler by moving sample start- and end-points a little further in than might seem natural. In fact, if you edit in the middle of a waveform, the momentary burst of distortion this creates can help the sound become more punchy.
|
The level envelopes of the electronic kick-drum waveforms below, by comparison, decay far less rapidly, and this lends them a power which accounts for their popularity in much dance music.
|
Obviously, in many cases there are no envelope controls available to tweak, and this is where compressors can really help out, once again bringing up low-level detail to create better sustain and therefore a more powerful overall sound. This is especially applicable with drum sounds, the envelopes of which usually start with a very loud transient before quickly dropping away to silence. If you use a compressor to push down the level of the transient attack, you can then increase the overall level of the sound to make the subsequent decay sound louder relative to the attack. For example, if you were to compress the peak by 10dB, you could bring up the overall level by 10dB without changing the maximum signal level. The result would almost certainly seem louder, as the overall sound would have more energy than before.
Obviously, to make the above technique work effectively, the compressor needs to have a pretty fast release time, in order to allow the gain to bounce back up as soon as the sound's attack transient is over. Though using fast release times can often result in audible gain pumping, the human ear tends to associate this with loudness, so using this effect creatively shouldn't cause problems — as long as you don't overdo it! Some compressors pump more obviously than others, so it's worth experimenting with different ones if you can. However, don't worry if you don't have much choice — most fairly decent compressors should be able to get you close to the sound you're after.
Engineers often seem to swear by particular makes of compressor for bass, particularly when it comes to making things sound larger than life, and there is something to be said for this view. Different gain elements, circuit designs and amplification topologies all have different sonic signatures, some of which are more suitable for bass sounds than others. However, while choosing a 'name' compressor associated with this type of production will often get you better results, spending a lot of money on one will probably make a much smaller difference than you might imagine.
If you're not getting what you're after from a normal compressor, particularly when working with percussive sounds, you might also consider looking at a more specialised dynamics processor, such as SPL's Transient Designer or the envelope compression mode of TC Electronic's Triple•C, both of which are worth a careful look if you're seriously into dance music of any kind. The SPL unit has two control knobs, which allow the levels of the attack and release phases of percussive sounds to be adjusted relative to the overall envelope. The process works independently of level, so there's no threshold to set up, and it works on complete loops as well as on individual hits. The TC Electronic envelope compression is very similar, except that it requires a level threshold to be set and it allows attack and release times to be specified.
Such devices can also be used to add punch and sustain to bass sounds, especially if individual notes are quite short. In fact, the SPL unit can even be used to create a rather nice subtle distortion on low bass sounds, as it begins to process individual cycles of the sound's waveform, and this can also boost the perceived size of the sound.
|
|
Putting On Layers |
|
|
|
|
Acoustic kick drum (top) layered with an electronic kick (middle), compressed to boost its decay, produces the lower waveform.
|
Distortion and dynamics processing are both very powerful tools within dance music styles, but they share one limitation: they both merely respond to the qualities of the raw sound which they are processing. If you find that no amount of processing is proving successful for your mix, it's probably time you reassessed the sound itself. Fortunately, given that most dance musicians run MIDI instruments live, this is not a big problem, because sounds can be layered to produce the desired result, simply by copying the relevant MIDI track and rerouting it.
If your bass sound is lacking something, a good first course of action is to double its part with a synth patch which is very close to being a pure sine wave. While this is playing, experiment with your sequencer's transpose control and with the level of this sine wave part, to see if there are any frequency-response 'holes' which might benefit from being filled in. Don't automatically use such a sine wave as a sub-bass sound, because it can be really useful elsewhere as well, but if you do then make sure also to experiment with the exact timing of the sine wave part for the best results — most sequencers will allow you to do this with a non-destructive delay parameter. If you're just wanting to add sub-bass, you might also consider using one of the sub-harmonic processes available, such as the Aphex Aural Exciter's Big Bottom function.
Another sound which you might consider layering with you bass parts is a higher-pitched distorted guitar patch — a GM one will often do — but mixed low or reduced to a short duration. In this way you can often beef up the sound without making it seem like it's comprised of two separate layers. Picked guitar sounds can also be useful for increasing the punch of basses, particularly if the guitar sound is quickly muted — in general it is often worth making any high-frequency layer slightly less sustained than the bass sound itself. Why not also experiment with putting such a sound at the end of every bass note, as well as at the beginning? This makes the envelope's release appear more abrupt and can therefore make the sound appear more powerful.
Kick drums can also benefit from a bit of creative layering, particularly if you want to use samples of real drum kits within a dance context. While real drum sounds often have just the character you're after, they rarely have the sheer power that is expected of dance sounds. For this reason, layering more powerful electronic sounds below acoustic ones can provide the best of both worlds. Mixing in the attack phase of, say, a TR808 kick drum sample with your acoustic kick drum can really help it cut through, for example. On the other hand, if you heavily compress the electronic kick sound before mixing it in, this will boost the decay portion of the combined sound's level envelope, and could therefore make it seem more powerful. |
|
|
Going Large With Effects
Obviously, when adding effects in dance music, anything goes on an artistic level. However, a number of effects also have the potential to increase the apparent size of a track. For example, stereo modulation effects such as chorus, flanging and phasing can help increase the stereo width of sounds, making them appear somehow larger. However, there is also a tendency for such effects to rob a sound of its impact if they are overused. What's more, heavy modulation-effect treatments also introduce pronounced frequency-response notches which are capable of emasculating otherwise beefy bass sounds — if you want to use a modulation effect on your bass sounds, it is therefore a good move to experiment with filtering out the low frequencies from the effect return. Stereo-width treatments, such as those offered by the SPL Stereo Vitalizer series and the Behringer Edison, can also be useful in a similar way for increasing the apparent expanse of sounds.
Static flanging and phasing (effectively just stereo delays with times below a couple of milliseconds) can also be useful for increasing stereo width, particularly if these delays can also be pitch-shifted. Similar stereo width enhancement can also be added using a multitap delay, early reflections, or ambience program, as long as you bear in mind that this might also start to push the sound towards the back of the mix.
Delay lines are extremely popular for making tracks appear more impressive, giving the impression that the sound is filling some cavernous venue or massive arena. However, the problem with using delays is that they also appear to distance the track from the listener, which can make individual instruments seem less immediate. This distancing can be made less pronounced, however, if the delays are significantly distinct from the original sound. Band-limiting of the effect return can be particularly useful here, though you could also try feeding the delay line's output to a modulation or reverb effect, to distance the echoes further with relation to the source sound. Naturally, you could also turn down either the delay's feedback level or the fader on the effects return channel — however, heavy delays are characteristic of many dance styles, so this may not be an option for you.
Another potentially problematic side-effect of using delay is that it increases the overall level of the track, and that this overall level will therefore need to be lowered to remain within the available headroom of the recording medium to which you're mixing down. This can be especially difficult to deal with if the delays are tempo-sync'ed, as the delays are often then masked behind the beat — if they are faded up enough to be clearly audible, they can really eat into your available headroom. One way around this problem is to experiment with the more unusual tempo-sync'ed delays. If, for example, you use triplet divisions (rather than straight quarters, eighths and sixteenths), this will not only cause the delays to become more audible by avoiding the main beats, it might also make the track more rhythmically interesting — just ask William Orbit...
When working within dance styles, you should be careful with how you use reverb, as using too much of it can really reduce the impact of a track — outside the trance style, at least, where vocals and pads are often treated to cavernous reverbs! If you feel that your mix isn't really knitting together well enough or that the high-frequency elements seem a little dead, then try to solve these problems using delays, early reflections programs and ambience if you can. In fact, you can often shove a whole mix through a good ambience program if it's not working, though this is much less likely to work with a cheapo reverb plug-in.
Processing The Mix Buss
Though the above technique can sometimes work wonders, as a general rule, it's worth avoiding any processing of your complete mix if you can avoid it — such processing is best left until the mastering stage. One notable exception to this is using a full-band compressor on the mix buss to get a particular sound — the subjective balance of a mix can sometimes change quite noticeably, so this is best done during the mixdown, so you hear the effect it produces. However, even in this case, it's best to tread with caution.
|
|
EQ'ing For Size |
|
|
|
A carefully constructed arrangement will usually avoid conflicts between sounds, so there can often be very little left to do at mixdown in terms of equalisation. However, if there are any remaining problems, then it can often be useful to bracket any troublesome sounds with EQ to prevent them trampling on each other.
Even though so many samples and synth sounds seem to sound great straight out of the box, you will still sometimes need to fine tune the tone of individual sounds. Bear in mind while doing this, though, that it's safer to cut than to boost, especially where narrow frequency bands need treating, and too much boost using a less-than-excellent EQ can make sounds appear unfocused, phasey or disaccociated with a mix. While maintaining a natural sound isn't usually a priority with dance music, the last thing you want is to lose impact. If you are not sure which frequency needs treatment, try applying a fair degree of boost and then sweeping the frequency control — once you've found the relevant area, you can return to the cut/boost control and set it to a more suitable position. But above all, whenever you decide to use equalisation, try to do so within the context of your mix, rather than with the sound in question soloed. |
|
|
Mix buss equalisation in particular is easy to abuse, and enhancement processes (such as those offered in stand-alone units by Aphex, SPL, BBE and Behringer), while often very useful, are easy to get used to and overdo. Just remember that you want to avoid doing anything that will get in the way of the mastering process, expecially if you're having your mix mastered commercially.
If you prefer to master your own music, then be sure to use the best monitors you can, as well as constantly comparing your mixes with commercial dance tracks in the same style. I would advise keeping tonal changes to a minimum, and when mastering dance music I usually find myself confining these to gentle changes at 80Hz and 15kHz, for balancing the bass end and for adding detail respectively.
Multi-band compression is an excellent tool for increasing the energy of a mix without the side-effects of full-band compression, and it can be particularly useful for applying extra compression to the all-important bass end if you feel you're still short of energy there. When I'm mastering dance music for clients, I'll often use a multi-band compressor to apply very low-ratio compression to all three frequency bands (no more than 1.25:1 usually), but with a threshold of around -30dB. I normally aim for no more than about 6dB of overall gain reduction, and this tightens up the sound considerably without destroying the impression of dynamic range.
Dynamic equalisation can also be useful to add weight to bass sounds without upsetting the rest of the balance. To make this work, I apply a little 80Hz boost (around one octave wide) only when the input signal exceeds the threshold, which I set to trigger only on kick drum beats and bass synth sounds. A boost of two or three decibels is often quite sufficient.
Another process that is worth trying at the mastering stage is valve simulation, which can be used to add even more energy to a mix without making it seem obviously distorted. Not only is this available in some of the digital mastering processors, but there are also numerous plug-ins available offering this. Such processing can be particularly powerful if distortion can be added to individual frequency bands in a similar manner to the way the Quadrafuzz plug-in operates — particularly for increasing low-frequency 'bigness'.
Most dance tracks should really have their bass sounds panned to the centre of the mix in order to share the load of reproducing the low frequency energy between the two speakers in a stereo system. However, it can still be worth experimenting at the mastering stage with the stereo spread of different frequency-bands in the mix, if you have the facilities to do this. Not only can the bass be narrowed down for extra power, but the high-end can be widened for a more panoramic and impressive effect.
As a final stage in the mastering signal chain, I generally use a peak limiter to trim off between 3dB and 6dB of peak level, increasing the average mix level accordingly. All that remains then is to listen to your potential master CD on a number of different music systems on which you can imagine your track being played, and at a number of different volumes. For dance music which you want to play in a club, try to persuade someone at your local venue to let you play the track over their PA so that you can hear how it sounds — you could well be very surprised. If you envisage your track appearing both on the radio and in the club, then you might consider doing different versions of your track for these different purposes. You might want masses of sub-bass on a club mix, but this will just be an inaudible waste of headroom on smaller systems. If you filtered out the lowest bass frequencies, you'd be able to increase the perceived volume of the entire track for a radio mix.
Larger Than Life
A lot of the activities involved in music production, and particularly in dance music production, are about creating an illusion. If you've put the above advice into practice, you ought to be well on the way to achieving one element of this subterfuge: the impression of size.