Filter/Resonance The most useful feature of the AWE32 for anyone producing electronic dance music are the resonant low pass filters, which can be applied to any of the 32 channels of the EMU-8000. Being a low pass filter, as the name suggests, it cuts out the high frequencies, and lets the low ones pass through. Sweeping effects can be produced, by varying the point (frequency) where the filter takes effect. A resonance control can be used in conjunction with the filter, which emphasises the signal at the frequency where the filter takes effect. Click here to listen to an example of the filter/resonance on the AWE32. Both the filter and the resonance control on the AWE32 can be varied by using MIDI NRPN controllers (NRPN=non-registered parameter numbers. These are controllers that are not part of the General MIDI specification). In order to vary the filter cutoff, enter the following set of controller values into your MIDI sequencer... controller 99 (NRPN MSB) set to 127 controller 6 (Data Entry MSB) set to 64 controller 98 (NRPN LSB) set to 21 by varying the value of controller 38 (Data Entry LSB) between 0 and 127, you vary the filter cutoff point. ...and similarly, you can vary the resonance by using the following set of controller values... controller 99 (NRPN MSB) set to 127 controller 6 (Data Entry MSB) set to 64 controller 98 (NRPN LSB) set to 22 by varying the value of controller 38 (Data Entry LSB) between 0 and 127, you vary the resonance. Some MIDI sequencers give you the option to insert a stream of controller values over a time period, so you get a smooth change in cutoff point or resonance. Others allow you to draw a graph of the controller values using the mouse. I find the best way to insert these values into the sequencer is to record them in. Many keyboard controllers have a slider that you can assign to send controller 38, and by doing this, you have a physical control over the sound in real time (i.e. you can hear the effect of the filter/resonance as you play). Using the filter and resonance controllers at the same time can be very effective (the example above uses this technique). In order to use both simultaneously, you need some sort of MIDI controller with at least two assignable sliders. The Kenton Control Freak and Peavey 1600x are two such examples. Pitch Bend By default, the AWE32 will only change the pitch by +/-2 semitones when responding to a MIDI pitch bend message. However, using the following sequence of MIDI controllers you can change the pitch bend range (up to a limit of +/- 2 octaves)... controller 101 set to 0 controller 100 set to 0 controller 6 set to X (where X is the desired pitch bend range, in semitones). Programmable MIDI Controllers Using the AWE32 in conjunction with a programmable MIDI controller, gives you ultimate flexibility when modifying sound elements like volume, expression, pan, pitch and filters. These devices have sliders and buttons which can be assigned to control these aspects of the sound. As well as fairly basic models, like the Keyfax 'Phat boy', there are those where the user can program the function of the sliders and buttons. Two such examples are the Peavey 1600x, and the Kenton 'Control Freak' (the one I use). The control freak is a very versatile device. I've found it useful, not only for controlling MIDI data on the AWE32, but also for controlling the virtual knobs and buttons of soft-synths, like Rebirth. Although a little pricey (about $700aus), this sort of MIDI controller is something that you should be able to use for many years to come, with many different pieces of equipment, so I highly recommend one. Separating Tracks This technique can be utilised by anyone using a MIDI sequencer, not just AWE32 users. Most modern sequencing programs allow you to have an almost unlimited number of tracks. I find it useful to take advantage of this feature, and use separate tracks for note data, expression data, pan data, etc... In this way you can have all the notes for a musical passage separated from (for instance) filter control data, and if you want to change the filtering, all you need to do is delete the track with the filter data, and re-record. This is a simple tip, but is still a very useful technique. Placing different types of MIDI data, on separate tracks of the sequencer can make editing/correcting much easier. Using MIDI Delay One of the real weaknesses of the AWE32 are its internal effects. Whilst a delay effect is available, its parameters are not adjustable, and using it means foregoing the chorus effect completely. However, by using some clever MIDI programming, you can actually simulate a fairly realistic delay effect. 1) We start with a simple riff. The first step is to make a copy of it, and to assist with creating the delay effect, move the copy so it is in-line with the point that you want the first delay to occur. In the example below, I want the echoes to occur every crotchet (quarter note), so I move the copy a crotchet beat ahead (and also down an octave to make things simpler, but I'll move it back to the correct pitch later). 2) The copied part needs to be reduced in volume. The best way to do this is to reduce the velocity of all the notes. Many sequencers have a subtraction or scaling function, that you can use to reduce the velocity of the copied part. In this example I reduce each copy by 50% (cumulatively), but obviously this will vary depending on how pronounced you want the delay effect to be. Also, we need to look at which notes will overlap. The second note in the copied part occurs at the same time as the last note of the original part, and since the original part should take priority, we need to delete this second note. Also if there were any notes in the copied part that trail into notes of the original, they need to be shortened, so they end before the original notes. So after deleting this second note, the two parts look like this... 3) Now we need to move the copied part up to the right pitch. Before doing this, make a copy of the copied part (either to the sequencer's clipboard, or to another part of the sequence). This will be used to create further iterations of the echo. Now move the copied part up to the right pitch. 4) Further echoes can be created by following steps 1 to 3 successively. Once you have a satisfactory number of iterations, the sequence should look somewhat like this... Once you've mastered the basic technique it can be applied in more complex ways. For example you can put the echoes on a separate MIDI channel (good for more complex riffs), and this will allow for all of the notes in the echoed part to sound, without cutting off the original part. It's also possible to create a panning delay, by putting the echoes on two separate MIDI channels, and panning each channel hard left, and hard right. Click here to listen to the original riff. Click here to listen to the riff with MIDI delay applied. Using Velocity and Timing Effectively When starting out with MIDI, many people tend to always use the same note velocity when creating percussive and melodic parts. A simple but often overlooked technique is to use slightly different velocities for the different notes in a particular passage. This tip is particularly aimed at those producing electronic styles of music, which often contain little note velocity variation. By simply making the velocities of certain notes slightly higher or lower, you can really add some life to an otherwise boring, robotic passage. If you own a controller keyboard, often the easiest way to do this is to record passages using the keyboard, rather than entering the notes by step. Even those who aren't good keyboard players will inadvertently accent the correct notes to some degree. The sequencer's quantise function can always be used to fix the timing if necessary. Whilst on the subject of quantisation, another often overlooked technique is to use some variety with MIDI timing - both the start times of notes and their lengths. Human drummers rarely (if ever) play with the precise accuracy that MIDI offers, so by introducing a random element into the timing (particularly of percussion parts) you can give a robotic passage a more human feel. Again, a good option is to use a controller keyboard to play the parts in, and then quantise to about 90% accuracy. Some sequencers also offer a 'humanise' function, that can be used to liven up a passage after it's been entered into the sequencer by step. To demonstrate the velocity technique described, I've included two audio samples of hi-hats played in straight 16ths. In the first example, all the hi-hats are played at a velocity of 100. However in the second example, accents are placed on the first and third hi-hats, whilst the velocities of the second and fourth are reduced (so the velocities for each group of four hi-hats are 90, 60, 120, 50). Notice how this very simple change dramatically changes the rhythm and feel of the pattern, and makes it more interesting. Setting Reverb and Chorus to 255 Normally, the amount of reverb and chorus applied to a particular MIDI channel is controlled using MIDI controllers 91 and 93. This allows the amount of either effect to be set between the standard MIDI values of 0 and 127. However the AWE32 actually allows the amount of reverb and chorus to be set as high as 255, allowing for much thicker and deeper effects. This can be achieved (as with the Filter/Resonance control described above) using NRPNs, and although it is covered in the official AWE32 FAQ, it is not explained particularly simply. Because one standard MIDI controller only offers values ranging from 0 to 127, to create values ranging up to 255 we need to use two MIDI controllers. In this case the two controllers in question are Controller 6 (Data Entry MSB) and Controller 38 (Data Entry LSB). By setting controller 6 to a value of 64, setting controller 38 to values between 0 and 127 produces an equal amount of the chosen effect. However, by setting controller 6 to a value of 65, values of controller 38 actually produce an amount of effect equal to the value of controller 38 plus 128. To differentiate between setting the amount of reverb or chorus, controllers 98 (NRPN LSB) and 99 (NRPN MSB) are used. Controller 99 is always set to 127, but by setting controller 98 to 26 you control the amount of reverb, and by setting it to 25 you control the amount of chorus. For example, to set the amount of reverb to 55 you would use the following set of controller values (equivalent to setting controller 91 to a value of 55)... controller 99 set to 127 controller 98 set to 26 (to select reverb effect level) controller 6 set to 64 controller 38 set to 55 ...however if you wanted to set the amount of reverb to 200, you would use the following set of controller values... controller 99 set to 127 controller 98 set to 26 controller 6 set to 65 controller 38 set to 72 (since 72 + 128 = 200) Similarly if you wanted to set the amount of chorus to 165, you would use the following set of controller values... controller 99 set to 127 controller 98 set to 25 (to select chorus effect level) controller 6 set to 65 controller 38 set to 37 (since 37 + 128 = 165)