SamplecrazeStretch That Note

Quantise - what is it and how do you apply it?


Quantisation is the process of aligning a set of musical notes to conform to a grid. When you want to quantize a certain group of MIDI notes in a song, the program moves each note to the closest point on the grid. Invariably, the quantise value determines where on the grid the notes are moved to. 
Let us look at this as an image so you can understand exactly how quantise works.

As you can see, I have opened up the key editor in Cubase (you can use the key or grid editor in your own software application), and have chosen 2 bars (1-3) and have selected a quantise value of 16 (evident in the top section of the editor below). 

Looking at the key editor image, it is clear that each bar has been ‘cut’ into 16 segments (count the empty boxes).   In terms of quantising Midi notes to the grid, you can see that there are 16 available grid points that will be used to determine where the input notes will be moved to. If you were to play some notes into your sequencing application and view it in the editor, you would see where the notes lie. If the notes are not timed well, then by using the quantise function you can move the mistimed notes to the nearest quantise value.   Look at the two images below, one is a ‘before’ quantise and the other is an ‘after’ quantise.    




In the ‘before’ image, you can see that I have deliberately input mistimed notes and you can see how these notes are either too early or too late.   Now, if you look at the ‘after’ image, I have used the quantise function and quantised the notes to their nearest grid positions. Because I am using 16 values (resolution), I have enough segments to use to move the notes to. This corrects my mistimed notes.   Sometimes, even using 16 quantise can still make the inputted notes sound too syncopated (robotic) and this can take away from the natural feel of ‘playing in’ the notes. So, to try to minimise the rigidity of the moved notes, we can use higher quantise values.   If we go back to the mistimed (before) image prior to me using the quantise function, and change the quantise value to 32, I will get a more detailed and finer resolution to play with.   Look at the images below and you will see what I have done. 


By selecting a quantise value of 32 I now have more segments to quantise the notes to. This gives me finer and smaller move points. You can see that there are now 32 little boxes for each bar, as opposed to 16.


By using the quantise function, set at 32, I have moved my mistimed notes to the nearest 32nd point, as opposed to the nearest 16th point. This allows for finer corrections and makes the music sound far less syncopated.   Most software will allow for even finer quantise values (up to 128th) and afford the user with all sorts of options to further define how the quantise function behaves. Most of these options include:  

Swing: It lets you offset every second position in the grid, creating a swing or shuffle feel. Swing is actually a great quantise weapon. It is most commonly used by the Hip Hop fraternity to compensate for the lack of a ‘shuffle’ feel to the beat. The amount of swing applied to the quantise is determined in percentages. The more swing, the higher the percentage applied.  

It is important to remember that the slower the tempo of your track, the more syncopated the music will sound if low value quantise is used. This has caused problems for many songwriters and they usually compensate by using higher quantise values, or working in double time (ie using a tempo of 140bpm for a song that is meant to be in 70bpm). Working in double time is the equivalent of using half the quantise value. For example, a song in 70bpm written in 140bpm can use a quantise value of 16, which would equate to using a quantise value of 32 when using the original 70bpm (beats per minute) tempo.  

 The swing function allows for a more ‘offset’ feel when quantising and makes the music sound more human as opposed to robotic. In fact, swing is such a potent tool that the Dance heads are now using it to give a little life to the hi hat fills etc.  

Grid and type:

Grid allows you to pick a note length (for example: 1/4, 1/8, and so on) to use for the resolution, while Type sets a modifier for the note length: Straight, Triplet or Dotted.  I will not go into this as you would need to understand about note lengths etc, but what I will say is that the triplet is extremely handy when programming drums and particularly hi hat patterns that require fast moving fills.

Random Quantise:

Another feature that can be useful to make your performances sound more in time without being completely mechanical is Random Quantise. Here you specify a value in ticks (120ths of sixteenth notes) so that when a note is quantised to the nearest beat specified by the other parameters in the quantise template, it is offset by a random amount from zero to the value specified by the Random Quantise setting. Basically, this takes away the rigidity of syncopated rhythms, particularly when dealing with hi hats. It allows for a ‘random’ element to be used, much akin to a drummer’s human timing.  

Most software will come with many additional tools to refine the quantise function and it’s settings. Humanise, iterative, freeze etc all go to giving the user more detailed editing power. For the sake of this e-book I am keeping it simple and only using the functions that most will adopt.