Keyboard Master Class

							Keyboard Master Class


Do It Yourself: Creating Realistic Sounds
by Richard Formidoni, Korg USA Product Specialist

A couple of decades ago, before the birth of digital synthesizers and samplers, the only type of synthesizers
you could get were analog. We all know that analog synthesizers are typically associated with fat bass
sounds, great leads and evolving pads. Lest we forget, there was a time when synth players were also called
upon to use analog instruments to create realistic sounds! How did they do it without samples? They
programmed. They flipped switches, patched cables and twisted knobs. Believe it or not, this stuff is FUN.
And even though your synth might come stocked with tons of great pre-programmed sounds, I recommend
you try your hand at the art of imitative sound creation. You may be surprised and gratified by the results!

Every synth is different, and they all have their own methods and procedures. I’ll try as best I can to use
general terms that you can apply to almost any synth.

Setting the tone

Let’s start with an attempt at creating a brass or a string sound. There are two ways we can go here – one
instrument or multiple instruments. Before we decide which we want, let’s start with a sawtooth waveform as
our basic timbre. It has the richest harmonics of all the likely choices, it’s more “brittle” than a sine wave,
not as nasal as a square wave…a perfect starting point.

Now, if you just want one horn, set the synth to monophonic mode (which most synths were limited to until
the eighties!). Now you have one horn, playing one note at a time. So, how about an ensemble? Of course,
we need to switch from monophonic to polyphonic, but that’s just the beginning. If your synth has a “unison”
or “ensemble” mode, now is the time to turn it on. Right away you’ll notice a fatness to the tone. Also, you
might want to use the “detune” function, if you’ve got one. Not too much, just a few cents. This will help
create the illusion of having more than one player when you’re playing one note.

Creating the instrument

So, we’ve got the basic tone ready to go. Now we start sculpting it so it reacts like the original instrument
would. For this, we need an envelope generator. On a typical envelope, you have four parameters: attack,
decay, sustain and release. These are what we’ll use to shape this wave into a realistic horn or string sound.

Let’s think about how a human blows into a horn. The breath doesn’t come out full-force right away; there is
always a little bit of ramp-up time as the player increases the force of the breath. This is not a slow ramp,
it’s a very quick one, and we’re going to simulate it by slightly increasing the attack time. As you adjust the
attack, you’ll notice the sound taking a longer time to “fade in.” We don’t really want a long fade, just a little
hint of breath, so keep that attack value nice and low. A little dab’ll do ya.

A string instrument, like a violin, has much more flexibility for attack. Even in a pizzicato note, however, the
bow doesn’t suddenly start moving at a certain speed, it arrives at its intended speed over a short period of
time. So, your attack should never be set at zero. It will sound artificial. Assigning a much longer attack
(meaning, a higher attack value) is quite useful for creating a mellow string pad sound, especially if you’re
playing a slow, dramatic passage. The notes gradually come into existence, and the effect is very intense.

Next up is “decay.” Decay is how long the note takes to settle into a long sustain after the initial attack. This
is a crucial parameter for replicating instruments like pianos, where the note gets quieter the longer it is
held. It’s a little different for our horn sound, though. Our imaginary horn player has to blow extra hard to
get the note to sound, but not quite as hard to get it to keep going. So, there will be a small decay time
when the initial horn blast turns into a long, sustained note. We can set a small decay, just to mimic the
“spike” of that first initial breath. For a string sound, the decay value might be a bit lower, because a bowed
performance is usually more constant.

Sustain, the third parameter in our envelope, is where the decay stops. Contrary to what you might think,
the sustain parameter doesn’t determine how long the note is. That is determined by how long you hold
down the key. In our horn example, you can think of this as the volume we want the note to have AFTER the
initial blast. As you can imagine, this is something we might want to change often! With that in mind, you
might want to consider assigning this parameter to a control that’s right in front of you, like the mod wheel.
This can be very expressive in a string patch, especially when combined with LFO vibrato (see below). With a
string sound, you also can achieve an accent by combining a small decay with a sustain value that is slightly
lower than the initial velocity.

Release is the last parameter of our envelope. Release is the amount of time the sound takes to fade out,
after you let go of the key. Let’s treat this one the same way we did the attack. Just a tiny little bit of release
will let you hear the horn or string “come down” from its tone, without sounding robotic or inhuman. Now, if
our player is doing a very fast solo with lots of staccato notes, we might want to decrease this value a bit, to
make sure every note pops like it should. If it’s a horn, the breath could be cut off very quickly. If it’s a
violin, the bow could be bouncing off the strings. Also, the desired value will be different for your
monophonic and polyphonic settings, so experiment to get it to play the way you want. Again, though, let’s
consider our options! A dramatic string ensemble passage calls for a long release. Combined with a long
attack, this makes for a great pad-like atmospheric sound. This is great for that one really sad scene in the
movie where the hero is down because his girlfriend got eaten by giant mutant lobsters.

Your instrument ought to sound pretty sharp by now. You’ve got a good basic tone, and using the envelope
generator, you’ve replicated the way the instrument is played.

The finishing touches

Now, let’s add some polish to our sounds. The following techniques can be applied to both string and brass
sounds.

Want to change the tone without digging too deep? That’s what filters are for, my friend. You’re concerned
with two controls here, cutoff and resonance. If you’ve ever used a synth of ANY kind before, you know that
these are the most useful “instant gratification” knobs you can grab. Crank up that resonance just below
halfway (results may vary), and experiment with the cutoff knob to find just the right tone. This is a quick
and dirty way to go from tuba-to-trombone-to-trumpet, or bass-to-cello-to-violin. Experiment with the
different types of filters, as they will all give you very different effects.

Suppose our instrument has some vibrato? It’s time to de-mystify more synth lingo…this time it’s the LFO, or
Low Frequency Oscillator. It’s another one of those things that are pretty universal in the synth world.
Basically, an LFO applies back-and-forth motion to the sound. The motion can be pitch-related (vibrato),
volume related (tremolo), and on high-end keyboards (like every current Korg product), you can usually
route the LFO to nearly every synthesis parameter. Most LFOs allow you to choose from a number of
waveforms, usually a saw, sine, square, or triangle. Think of them as “shapes” for the movement you want
to create. So for vibrato we want to use a sine wave, because its smooth shape closely mimics the gentle
motion of natural human vibrato. Then we need to assign the LFO to control pitch.

Now, if your synth is truly tweakable, it will have controls for LFO modulation depth (or level), frequency (or
rate), and delay. The depth control lets you specify how far that LFO will bend the pitch. You probably don’t
want to go too far with this one. Once again, we’re applying the “less is more” technique. Just a little pitch
variation is fine. But then again, you might want to get waaaay out there at the end of your Sinatra cover,
and really wail those last couple of notes. You might want to consider routing the depth control to a nearby
knob or slider, so you can adjust it freely.

Frequency determines how fast that pitch bend happens. You know that joystick (or Pitch Bend wheel) that is
sitting to the left of the keyboard? If you move it slowly back and forth, that’s a low frequency. If you move
it back and forth like you just inhaled fourteen pounds of sugar, that’s a high frequency. You get the idea –
think human! Of course, the frequency depends on the mood of the song. So again, here’s another
parameter you might want to have assigned to a knob right in front of you for a live performance!

Delay is quite useful for a realistic performance. It will cause the LFO to only come in after a specified time.
Most players only bring in the vibrato after holding the note for a second or two. This is the parameter that
allows you to do exactly that. Experiment, and see what works for your sound.

Know your roots

When all is said and done, the last thing I recommend is that you get down on your unworthy knees and
thank the almighty gods of synthesis that you can SAVE YOUR WORK on whatever modern synthesizer you’re
using. If it were before the 1980s, you’d probably be writing down the position of every single knob or slider
that you touched during this creation process!! Oh, the horror…