Quantising — a really distinctive and powerful technique with a nerdy name — has been around since the dawn of MIDI sequencing. It's a process that takes MIDI notes or other events and 'snaps' them into line with a user-configured rhythmic grid, often resulting in inhumanly rhythmic playback that can be exhilarating or mind-numbingly lifeless, depending on the musical context and the skill with which it's applied. But quantising doesn't have to make everything you do sound like Kraftwerk, especially when you have DP as your sequencing platform. Used wisely, it can allow you to actually enhance the 'feel' aspect of your sequences, while easily taking care of more obvious errors.
As you might expect, DP's quantisation features are extensive, sophisticated, and crop up in all sorts of places — they're by no means a 'one size fits all' solution. In fact, in true DP fashion, there's a staggering degree of flexibility. So rather than exhaustively going through every single feature one by one, I thought we'd look at how you might best use quantise in a range of typical contexts, to achieve a variety of effects.
It might not be big and clever, but sometimes really strict, mathematical quantising is just what's required, especially in more 'electronic' musical styles. As you'd expect, it's simple to achieve in Digital Performer.
Probably the best way to apply this sort of quantisation is as a 'destructive edit' that actually changes the MIDI data you've recorded into a track. In the case of MIDI notes, they'll be moved to coincide with strict rhythmic sub-divisions of the bar (or 'measure', in MOTU-speak). Here's a typical course of action for this type of quantising, which assumes you already have a MIDI track that was recorded with reference to DP's metronome:
1. First, select where you want to quantise. You could do this using any normal selection technique in DP — selecting phrases in the Tracks window or dragging over notes in the Sequence Editor or Graphic Editor, for example, or even by hitting Apple-A (Select All).
2. Then hit Apple-0 (zero) or choose Quantize from the Region menu, to bring up the Quantize window (see screen above) and decide what you want to quantise. We'll assume, for now, that it's just MIDI notes, so choose Notes in the 'What to quantize' pop-up.
3. The rest of the window is dedicated to how the notes will be quantised. The first thing to decide is the basic grid value, and for that you need to consider what the shortest note value you played was. Figure out your smallest sub-division of the beat and choose a rhythmic value to match.
4. In order for your notes to be 'snapped' into time, you'll need Attacks checked in the top part of the window, but it's up to you whether you also have the ends of notes quantised. For the most natural effect, you should probably choose just Attacks, with or without the 'Don't change durations' option, which causes the note releases to move relative to the attacks. Choosing to have Releases quantised too, however, can create a very mathematical, almost step-time effect, so experiment.
5. Make sure that nothing else in the window is checked and click Apply. Your MIDI track is quantised according to the settings you made.
If you find it difficult to predict how the settings in step three and four will affect the musical outcome, use the Preview option. To see what it does, run through the same steps as I've just described, but first set up DP's Memory Cycle loop markers around the section you're quantising. When you reach step three, play your sequence and let Memory Cycle loop round and round. In the Quantize window, check the Preview box and you should instantly hear the effect of the grid value you've chosen, and of the Attacks, Releases and 'Don't change durations' options.
Control Controllers: If you can quantise Markers, why not do the same with controller data? This opens up the very real possibility of quantising controllers relating to filter cutoff and resonance in a synth MIDI track, say, and producing rhythmic, almost sample-and-hold effects. The 'What to quantize' pop-up in the Quantize window can be set to Controllers, which quantises all controllers, or Custom, which allows you to choose exactly which ones to quantise.
Customise The Grid: The quantise Grid doesn't have to be tied to musical rhythmic values. By clicking the little Time Format button near the word 'Grid', you can quantise to absolute time values, SMPTE time intervals, or even audio samples. This is potentially very useful when matching music to rhythmic cuts in film.
Plug-in To Quantising: The Quantize window's Preview function is a tremendous help in refining settings, but if you're the sort of person who can never make their mind up and you want to be able to quantise and then change your mind at some point down the line, Realtime Quantize is for you. This is a MIDI plug-in that you select in a MIDI track's plug-in slot in the Mixing Board. It offers exactly the same features as the Region menu 'offline' Quantize, but it quantises in real time, only during sequence playback, leaving you free to change its settings at any time. If you ever decide you need a destructive quantise after all, just select the data in the track the plug-in is applied to and choose 'Capture Realtime MIDI Effect' from the Region menu.
It's Not Just For MIDI: As of DP 4.5, quantising has not been just for MIDI events. You can quantise beat-detected audio too. DP gives you three options: to quantise beats within soundbites, soundbites themselves, or beats and soundbites. It's also not particularly worried about whether you recorded your audio to a click or not. Clever stuff, which you can read about at length in Performer Notes from SOS February, March and April 2005.
Preview can help you zone in on useful settings, but when you're already certain what kind of musical effect you want, there's a way to cut out the entire note-selection and quantise-application rigmarole and just quantise as you record. This is called Input Quantize, and because it relates to the actual recording process, rather than a pre-recorded region, it's found in the Studio menu, or can be called up with the Ctrl-Shift-I shortcut.
All the same quantisation options are offered in the Input Quantize window, but there's an additional 'Enable...' check-box, which allows you to keep the window open during extended recording sessions and simply enable the function as and when you need it.
I find that Input Quantize comes into its own when I'm building up drum parts using the Memory Cycle function (allied with Overdub record mode). It means that every new 'pass' goes in perfectly in time. The alternative — having to select every new drum you record and apply the Region menu quantise — is particularly tedious and best avoided! For more about this approach to recording, see the Memory Cycle section in July 2005's Performer Notes.
Many musical styles make use of compound rhythm — that is, a fundamental rhythmic 'feel' whereby each beat is split into threes or sixes rather than twos and fours, as in a typical rock shuffle. Some DP users might deal with this by entering a compound time signature (6/8, for example) in the Conductor Track, which would result in your basic beat being a dotted crotchet (quarter-note). That's a perfectly good approach, since Quantize's grid has a dotted note option, far right. But if you're happier just sticking to so-called 'simple' time signatures like 4/4 you can still easily work with — and quantise — compound beats.
In the Quantize (or Input Quantize) window, take a look at the Tuplet section below the grid section (see screen, left). You have to enable it by clicking its check-box, and then configure it by typing numbers into its two boxes (or by clicking and vertically dragging in the boxes). If you'd recorded a shuffle-rhythm track using a 4/4 time signature and crotchet (quarter-note) metronome beat, you could quantise the track by selecting a quaver (eighth-note) in the grid section and '3 in the time of 2' in the Tuplet section. You're just telling it that you need three notes in the space of two, and since two quavers add up to a crotchet (quarter-note), you end up with a quaver triplet grid. But the flexible nature of the Tuplet section means that you could just as easily quantize quintuplets or septuplets, or tuplets so complicated that no-one's thought up a name for them yet!
Get Smart With Quantising That Scores
The sharp-eyed menu-watchers amongst you may have spotted Smart Quantize in the Region menu. If you thought this might be some sort of 'clever' quantise option that could vary its settings dynamically, depending on the rhythmic content in your tracks, you'd be right, but it's not quite as useful as you might imagine. It actually performs a specific function — namely, preparing MIDI data for export to third-party scoring packages such as Sibelius. These packages sometimes transcribe what might seem like perfectly simple MIDI sequencer files in the most horrendously complicated (and unreadable) way, so Smart Quantize is there solely to make DP sequences more notation-friendly. It can be a time saver for general use too, but its lack of controls and configurability make its effects unpredictable.
Musically speaking, Swing describes a rhythm based on a division of the beat into long-short pairs — think of the most clichéd jazz ride-cymbal parts — and this, again, results in a compound beat.
However, any decent drummer or percussionist knows that swing is a very flexible thing. For example, the rhythmic feel of a quaver or semi-quaver (eighth- or sixteenth-note) hi-hat part can change dramatically with the introduction of subtle long-short variations in each pair of hits. And for some musical styles — such as garage and two-step — the precise degree of swing used can make or break a groove.
If you'd like to swing in the safety of your own home, and without any danger of long-term psychological damage, try this:
1. Record a 'straight' quaver (eighth-note) hi-hat pattern, with your favourite drum sound source, into a MIDI track, using the metronome and a 4/4 time signature.
2. Set up a Memory Cycle loop around this and play back your sequence at a reasonably fast tempo — say 130bpm.
3. Select the notes you just recorded and hit Apple-0 [zero] to open the Quantize window. Check the Preview box so that you can immediately hear the changes your settings will make.
4. Start off by choosing a quaver (eighth-note) in the Grid, making sure that nothing else in the Quantize window is enabled. You should hear your hi-hats snapped perfectly into time.
5. Click the Swing checkbox. If the percentage value is set to the default 100 percent you should hear your hi-hats snap into a swung compound-beat rhythm (see screens, right).
6. Now you can try different Swing amounts. The easiest way to dial these in is by clicking and dragging up or down in the percentage value box.
For many typical 'straight' musical styles, the introduction of, say, 20-50 percent of swing can really help to loosen up hi-hat and other continuous rhythmic parts, especially if you're able to complement it with a little dynamic variation, such as making the 'swung' hits or notes (ie. those 'between' beats) a little quieter. You could do this during the recording stage, or with subsequent editing of note velocity in the Sequence or Graphic editors.
Digital Performer News
Just when I thought Performer Notes was wrapped up for another month, MOTU announced the imminent availability of audio and MIDI interface drivers for the new Intel-based Macs revealed at Macworld San Francisco. As for Digital Performer, Mach Five, MX4 and Symphonic Instrument, MOTU say they are "currently qualifying... software products for use with the new Intel iMac and Macbook Pro". Watch this space...
In the meantime, users of those quaint, old-fashioned IBM and Motorola Macs can take advantage of a MOTU software update that has already been released. Symphonic Instrument has gone to version 1.1, gaining new features that include stand-alone operation, disk streaming, 64 parts per instance, and individual part outputs. This is a major update, which owners of version 1.0 of the instrument can download for free from www.motu.com.
Very often, when using quantise, the problem facing you is tracks that are a bit too loose when unquantised, but which lose all their feel when quantised strictly. Something that can come to the rescue in this situation is the Strength option. This is found towards the bottom of the Quantize window, and by default is unchecked, so you need to click it on and dial in a value to start working with it.
The basic idea of Strength is this: if the option is unchecked, or turned on with a value of 100 percent, any events (such as MIDI notes) moved when quantise is applied are yanked from their original positions and repositioned precisely, according to the Grid, Tuplet and Swing settings. But when Strength is set to a value of less than 100 percent, any events moved end up between their original position and their fully quantised position. So if Strength is set to 50 percent, for example, quantising moves out-of-time notes only halfway towards their precise, mathematically correct positions. This opens up the possibility of much more gentle and less destructive quantising, eminently suitable for more soulful tracks that nevertheless need some tidying up. You may need to experiment with Strength settings a little every time you use it, as the results can vary depending on the material you're quantising (see screens overleaf).
Another subtle quantise concept is Sensitivity. This is possibly an 'experts only' quantise option, but it can do some clever things. To understand Sensitivity, it helps to first think about how quantise actually works. Let's say you choose a grid setting of a crotchet (quarter-note). Assuming that you're not using any other quantise features such as Swing or Strength, DP first finds any events which lie up to a quaver (eighth-note) before or after each beat, and then 'snaps' them directly on to the beat. To look at it another way, quantise casts a crotchet-sized 'net' around each beat, to catch those events that don't align exactly with it.
The Sensitivity option, then, allows you to shrink the size of this net, so that events that lie 'in the gaps' of the quantise grid are never included in the quantise action, while events that lie nearer to the rhythmic divisions of the grid are quantised as normal.
Why on earth would you need this? Well, for one thing, it allows you to have rhythmic accuracy where you need it, while allowing 'free' rhythm in other places. As an example, imagine you have a synth solo track. For 16 bars there's a strong downbeat at the beginning of each 4/4 bar (that needs to coincide with a precisely placed drum hit on another track) and a loose, free-flowing solo for the rest of the bar. You could do with a quick solution that quantised all your downbeats but left the rest of the bar untouched. As you might guess, this is possible using the Sensitivity option — and here's how you'd do it. First, select the section, then in the Quantize window choose a semibreve (whole-note) grid, turn on the Sensitivity option, and dial in a sensitivity of, say, 10 percent.
By doing this, you're narrowing the quantise 'net' to just 10 percent of the length of each semibreve — five percent before it and five percent after — which, of course, aligns perfectly with your 4/4 bars on their downbeats. Now only the notes very close to your downbeats are quantised, and DP ignores the rest of each bar.
It's also possible to set the Sensitivity value to a negative number, in which case the quantise 'net' is aligned not around the divisions of the grid, but around a point halfway between them. This is actually very useful, because you can use it to make quantise catch only notes that are really out of time and some way from the rhythmic grid divisions you've chosen, while ignoring ones that are already aligning with the grid or are close to it. You might have deliberately played a snare drum track, for example, with each hit fashionably behind the beat. By quantising this track with negative Sensitivity you'd leave most hits where they were, quantising only those that had strayed too far out.
Playing with Sensitivity might not be something you'll do every day, but getting under its skin and knowing how to use it to your advantage can be very handy every now and then.
As the possibilities afforded by the Sensitivity parameter show, quantising doesn't have to be about making music soulless. Two extra parameters ensure that it doesn't have to be 'square' at all.
The first is the 'Offset grid' option, which simply moves quantise's rhythmic grid ahead of or behind the rhythmic grid used by your sequence as a whole. Clicking the little button next to the value field selects positive or negative, and the amount of offset is entered in beats and ticks. Likely candidates for offset quantising include individual drums, to modify the 'feel' and groove of a pattern, or perhaps one or more of several arpeggiated synth tracks, to achieve precise delay-like effects.
Finally, there's Randomize. This is an impostor in the quantise camp, as it can only detract from the rhythmic precision promised by the other features. However, a few percent of Randomize can loosen up an otherwise rigid quantise, and by using some Emphasis you can decide whether you want your Randomized placement mostly ahead of the beat (negative emphasis) or behind it (positive emphasis). Higher values of Randomize, which cause notes to be placed all over the shop — even changing their order — are the DP equivalent of opening a picture of a friend in Photoshop and using Liquify, Pinch and Spherize on it. We've all done it, and it's funny at times, but not particularly useful all that often!