|
What makes a great dance track? Part 1
|
View this Thread in Original format
| Beatflux |
Part 1: Repetition vs novelty
Repetition and novelty: every track will have both, but not every single track will have the same amounts. What is the ideal amount of repetition, and what is the ideal amount of novelty? Dance music has a lot of repeated ideas. Dance music is different in this regards to other kinds of music in that it tends to repeat ideas more frequently, but also more importantly its designed for dancing. This is what sets it apart from other types of music that do not necessarily make dancing its focus. All dance tracks are not made equal. Some seem better than others. Some songs seem to bounce more. Some of them seem to make you want to dance up a storm, others just a tiny rain cloud. The quality of a dance song is intimately related with dancing. People have, over time, made accommodations to people to make music easier to dance to. That begs the question, what makes a great dance track?
Do me a favor now: put on a song and start dancing. I know for some of you, this will be a very disconcerting experience because you got into producing/DJing to avoid looking like a fool.
Notice how there is a delay between when you really start dancing, and when the music starts? There is a lag, there is a brief time window where every dancer has to feel out the beat in order to "learn" the groove. Try playing another song, and try to start dancing to the beat as soon as the music starts. It's going to feel especially weird and awkward, but luckily orientation to the beat happens pretty quickly.
Any time you hear a loop for the first time, a loop where no material repeats, your brain marvels at the novelty of the situation and studies it. The next time you hear the loop(the 1st repetition), it goes into groove mode. Your brain, depending on how well the loop is composed, will allow for a dancer to anticipate the beat and this is what makes for good dancing. At this point the repetition was a pretty good idea for helping dancers along, but when it comes to the 2nd repetition, it becomes less interesting than when we first heard the loop. This is the point where the loop over stays its welcome. While the repetition serves to help our dancers anticipate the beat, it has also created a new problem.
Suppose we have 8 identical kick drum samples, laid out so that each one hits on each beat. If you played it back, it would sound pretty bland. A snare/clap is pretty much a given, so a snare drum is put on the 2's and the 4's. Now it sounds better, but why? Our long passage of kick drums has been transformed into 4, 2 beat long loops. Our snare drum has given the listener a distinction between the first beat of the loop, and the second beat. This is immensely useful, because at any point in time, a person can tell when they are on the 2nd or 4th beat, or the 1st or 3rd. But there's a new problem. How do you tell if the snare is playing on the 1st, 2nd, 3rd, or 4th snare?
Each of our loops sounds exactly the same, and after the novelty wears off and the dancer syncs themselves with the groove, the music grows stale. The music becomes boring and also confusing. The brain always likes to know its making progress through a song, and like a long stretch of road, different landmarks help people identify where they are along the path. The brain wants to know when each beat is coming, it also wants to know when each new bar is coming, when the end of a hyper measure(8 bar segment) is coming. The brain constantly wants reassurance. The more reassurance you can give it, the more comfortable it is and the easier it is to dance. It always wants to know it is, but it also wants to know where it is going and when a large structural change is coming.
Using different timbres for different sections of a song is a really simple way to ensure that listeners will have a good understanding of when the song is making transitions between different sections. Some songs will employ repetition with subtle changes in timbre or gradual introductions into new sounds, while other songs will make distinctions between different sections by giving each section its own sonic signature. Some of the sounds may be repeated from section to section, but most of them will have at least one unique sound per section.
One great example of a song that varies up its timbres to make useful distinctions is Gigi D' Agostino's "I'll Fly With You" For the first chorus, he uses a pitched up voice, and for the repeated chorus he returns the vocal back to normal. Another big thing to listen for is high frequency energy. The 2nd chorus has more high frequency energy and the strings have more high frequency content. The chorus literally ends on a high note. Popular songs tend to end with more high frequency energy and this song no different. Shifting the spectral balance of a song is another way to create distinctions between different sections. Typically the high frequency energy will start out quite low, and towards the end it will peak, or it may even peak at the very end.
On the more micro side of things, what happens between kicks is just as important. Sounds that help anticipate the beat are immensely useful. Things like an open hihat playing on the "and" of the beat(e.g. 1 & 2 & 3 & 4 &), provide for a "stabilization" of the beat. A reverse clap that leads into a clap of a different timbre can work extremely well. Ghost notes that accent and augment any kind of rhythm adds timbre and velocity variations that create a wave of change that help the ear. A lot of pros like to clamor about modulation. It's a key component to making things sound "interesting." The most popular trick, the low pass filter is so popular because of how effective it works. Constantly opening up the filter with an extremely stagnant synth can still sound very exciting as the energy keeps rising.
All of this change must be controlled and used wisely. It makes no difference to how wild and novel your music is without the appropriate amount of organization and repetition. A great track will lie somewhere in the middle, where repetition and novelty are constantly at odds with each other, shifting at different moments in time. |
|
|
| J.L. |
MM... how did this thread get 0 replies?
Good dance music moves people physically.
Great dance music moves people physically and emotionally.
Cutting-edge dance music moves people in a way that hasn't been done before.
I think the a problem producers face is they spend too much effort analyzing dynamics, EQing, panning, chord analysis, etc.... that they almost forget the movement aspect of dance music.
People sometimes use the term 'overproduced', and although some feel it's a meaningless term, I think it exists for a reason. Perfection does not exist in art, but too many producers approach music production from a lopsided technical standpoint.
I think it's bad for the art form how producers can use templates and select from their percussion samples and loops. Yes, it's more accessible and faster, but the novelty and humanness of the product goes down, akin to the industrialization of commodities that we buy. I'm not saying we shouldn't use any, but that these things should be used sparingly.
The beauty of a live classical performance vs. a MIDI classical is that imperfections enhance the beauty of a live classical performance. I have been trying to recreate imperfections in performances using Vienna Symphonic Library, and it is difficult and takes a lot of time, but the result sounds much more human.
Great analysis by the way, because I think sometimes producers just need to stand up from their chair and actually try dancing for once in their lives (me included).
PS
Before I get flamed, I am not completely without fault for some of the comments I have made |
|
|
| Mad for Brad |
i disagree with the term imperfection. The push and pull is very deliberate and is not a mistake. Imperfect, I suppose but it isn't imperfect in the sense of an accident.
As far as loops and quantization, that isn't the problem. There are so many groove templates that the problem has to do with your quantization grid, not the actual task of quantization. |
|
|
| Beatflux |
| quote: | Originally posted by Mad for Brad
i disagree with the term imperfection. The push and pull is very deliberate and is not a mistake. Imperfect, I suppose but it isn't imperfect in the sense of an accident.
As far as loops and quantization, that isn't the problem. There are so many groove templates that the problem has to do with your quantization grid, not the actual task of quantization. |
There was this thread on Gearslutz about the MPC groove, and someone concluded that it was because of very small timing deviations from the MPC that made the groove seem tighter. With small timting variations, you will get at least slight timbre variations. |
|
|
| Mad for Brad |
| how will the timbre change ? The mpc quantization settings are just varying amounts of swing. The things you can do now with groove templates is incredible. |
|
|
| Beatflux |
| quote: | Originally posted by J.L.
I think the a problem producers face is they spend too much effort analyzing dynamics, EQing, panning, chord analysis, etc.... that they almost forget the movement aspect of dance music.
|
If you mean groove, then yes I agree. I am the same. I am just realizing that my listening technique has been wrong all along. When it comes to dance music, the most important thing is groove above everything else.
Things like mixing can greatly affect the groove, so it's still an important part. |
|
|
| Beatflux |
| quote: | Originally posted by Mad for Brad
how will the timbre change ? The mpc quantization settings are just varying amounts of swing. The things you can do now with groove templates is incredible. |
It had nothing to do with the swing setting. It just had small timing errors.
I'm not putting down templates. |
|
|
| johncannons1 |
| quote: | Originally posted by Beatflux
Part 1: Repetition vs novelty
|
+1
interesting read mate :) |
|
|
| Mad for Brad |
| check out Huron's book sweet anticipation for an actual scientific look at this topic and not some guys opinion. |
|
|
| Mad for Brad |
| lol. not that there is not jitter but it is not something that you will notice in terms of rhythm. whether something is .1 ms off is not really something you can hear. |
|
|
|
|
