Modular Synthesis (Part 2)
Creating and playing brass sounds.
The topic of this month's column is brass sounds - and how to create them using a modular synthesiser.
Let's first have a look at how brass instruments achieve their characteristic sound. When a note is played on any brass instrument the fundamental is the first element of the sound that we hear. This is quickly followed by a fast, progressive build up of both odd and even harmonics and it is this harmonic movement that gives the characteristics 'wah' effect that is at the front end of every brass sound in one way or another. Accompanying this effect, however, are all the various 'human' elements such as overblowing, inaccurate tongueing, etc., which give rise to all the blurts, squeaks and other bits of dirt that characterise brass sounds. Setting up the 'wah' effect is fairly straightforward but the other elements require a bit more thought and hardware.
As with string sounds which we have already looked at, there are a few ways in which you can produce brass sounds which are covered separately below.
The simplest way to produce a brass sound is to use the sawtooth output from one of the VCOs and to set cutoff frequency and resonance to minimum with maximum sweep from the EG. The attack should be set to about 500ms so that the filter sweep will give the 'wah' effect required. The decay and release should also be set to about 500ms with the sustain set at approximately 80%. Adjustments can be made to 'fine tune' the sound to your exact requirements but these settings should give a pretty good brass sound. To improve further the harmonic movement you could try mixing in a pulse wave whose width is being swept by the EG. This will add a touch more 'bite' to trumpet sounds in particular.
As with strings, it's possible to use two VCOs to create a richer sound. In this instance, however, we are going to use an EG to sweep the pitch of the second VCO in order to try and create the 'blurtiness' of brass instruments. A patch for this is given in Figure 2. The VCF and EG controls are set pretty much the same as in Method 1 but as you can see, another EG is routed to VCO 2. By tuning VCO 2 an octave lower than VCO 1 and then adjusting the modulation level of the EG (whose ADSR controls are set practically the same as EG 1) so that VCO 2 rises an octave to match the pitch of VCO 1 you will hear the effect I am trying to describe. Adjustment of the ADSR controls of EG 2 will yield some interesting and varied brass effects which can be used for a wide variety of purposes. If you find the sweep of an octave too much, try tuning VCO 2, say, a fifth below VCO 1 and adjusting the EG modulation amount so that VCO 2 is in tune with VCO 1.
Points to watch in this method are not to set the decay and release times too long (ie. more than about 500ms), otherwise the effect will be that of pitch bending instead of the 'squeak' you are trying to obtain. Also, don't detune the VCOs against each other but get them as close to unison during the sustained portion of the note.
This method utilises even more hardware than the previous two, but is capable of providing a very realistic range of brass sounds indeed. I suggest that to begin with you use the simpler techniques of Method 1 but to this you can add a touch of distortion into the filter to introduce the overblowing characteristics of brass. Patch up as in Figure 3 so that the sinewave output of a VCO or the output of an oscillating VCF is fed into another CV input of the VCF via the EG/VCA combination. The frequency of the sine-wave should be about 100 Hz (although there's plenty of room for experiment if you're looking for different effects) and the EG 2 should be set to give a very short envelope with absolutely no sustain. Again, experimentation with the ADSR controls will give various effects. This will give a short burst of modulation at the front end of the note which produces the 'farting' effect present on many brass instruments. With careful adjustment of the modulation level the effect can vary from subtle to harsh, though some messing around will probably be needed to get the exact effect you want. By routing the EC 2/VCA 2 via a footpedal you can introduce the effect as and when you want. Using a single triggering system might also help here as you could vary the effect with your playing technique; playing staccato will give the 'blurt' on every note while playing legato will cancel the effect, and it will only be heard when you release the keyboard and play a new note.
These three methods should give you fairly good, all-purpose brass sounds. You might like to use two EGs on the VCF (as discussed in E&MM, October '83) to create a more complete envelope shape, and this technique can be applied to any of the above methods for even more variation.
Different brass instruments have quite a different tonal characteristics, however, so it's a good idea to look at each one in turn.
Trumpet: A very bright, brassy sound. Use Method 2 with a wide filter sweep and as short an attack and decay time as possible. Be careful not to lose the all-important 'wah' effect.
French Horn: These are more muted than trumpets so decrease the filter sweep and increase the attack and decay times. Method 3 is probably the best one to use with a soft 'blurt' on the front end of the sound. A soft tremelo on a VCA is also recommended if this is available.
Trombone: Again, use Method 3 but set up a wider filter sweep and a faster attack time. The 'blurt' is harsher on trombones than on horns so increase the modulation level of VCO 3 a bit. Pitch-bend, slow vibrato and portamento (especially fingered portamento) can be used to great effect as well.
Tuba: Quite a raspy one. Play low down in the bass end and use a bright tone but a slowish attack (not too slow, though) because, as with most instruments, the bass end takes longer to 'speak' than the top range. Again, Method 3 is probably the best but any of them will do.
Those, then, are the four main types of brass instruments. I would imagine you could try and synthesise other sounds such as Flugelhorns, Cornets and muted effects. These can be obtained by 'tweaking' any of the above.
Phrasing and playing technique is also important in trying to recreate the sounds more accurately and the main points to watch are as follows.
Length of Note: Brass players only have a finite amount of breath and cannot sustain a note indefinitely as you can on a synthesiser, so try and split the musical phrases into smaller sections so that the sound can have a chance to 'take a breath'.
Intervals: Wide intervals are also difficult so try not to have too large a jump between notes.
Vibrato: This is almost impossible for all brass instruments except the trombone so avoid it if you want a truly authentic brass sound. I must admit to using it myself as it can sound effective, but it won't be totally realistic - tremolo is better. The same applies to pitch-bending but, again, this can sound very effective.
Voicing: Brass sections don't usually play the block chords so beloved of keyboard players so try and space the notes out a little. You'll find they play a lot of 5ths, 6ths and soon but rarely a straight triad. Orchestrating your brass section will give you far more effective results than just playing the block chord and I advise you to listen to 'real' brass sections to hear how they voice their parts. Try not to go overboard in the amount you play either: some brass sections only appear for an occasional 'stab' on an accent. As we discussed with strings last month, try and record each part separately and adjust the tone of each instrument slightly for more variation.
As with any synthesiser sound, effects can play an important part in the final sound but for brass you'd be better off just sticking with reverb (not echo) unless you want a special effect. Chorus and/or flanging might beef up the sound but it can sound too 'electronic'.
That just about concludes this month's workshop. You'll notice that I've not included saxophones. This is because they fall into a category of their own by virtue of being both brass and reed instruments, but we will be taking a look at them soon.
Feature by Steve Howell
mu:zines is the result of thousands of hours of effort, and will require many thousands more going forward to reach our goals of getting all this content online.
If you value this resource, you can support this project - it really helps!