SamplecrazeStretch That Note

Synthesis - Part 5

FILTERS 2

In the last tutorial I mentioned the term resonance.

So, let us start this tutorial with resonance.

Resonance

Q - Also known as ‘width of the filter response’, this is the ‘centre frequency’ of the bandwidth and is measured in Hz. Also know as bandwidth and resonance.

A high Q value denotes a narrow filter width (bandwidth). A low Q value denotes a wide filter width (bandwidth).
I decided to include the eq equivalent of bandwidth simply because some of todays software vstis have a Q value for the filter bandwidth.

Most synthesizer manufacturers, and in the case of most analogue synthesizers, the term resonance is used most commonly. Other manufacturers of synthesizers, or software synthesizers, might call it emphasis or Q.

Last month we talked about filter cut-off and slope, and what they meant. Boosting the narrow band of frequencies at the cut-off point is called resonance.

If you were to boost the resonance to the maximum, then the filter will begin to actually self oscillate. This means that it will generate an audible sine wave, more like whistling, even when receiving no input signal. A very cool way of understanding what resonance sounds like is to perform what we call a sweep. Yes, another flash and funky term we programmers use to explain something really simple. Sweeping the filter means manually turning the resonance knob, clockwise and anti-clockwise. Select a waveform, set the cut-off point and turn the resonance knob and listen to the results. As you are sweeping, the resonance goes through all the different frequency harmonics, of the waveform, and boosts them, at the cut-off point.

Altering the resonance of a filter can create incredible sounds. Using the resonance to it’s maximum will give the shrieking effect, as the frequency band is so small when the resonance is maxed out. This leads to the filter self oscillating and the resultant sound is a sine wave that simply screams at you. There are other ways of using resonance to spice up your sounds. Assigning an LFO to the resonance can give a nice undulating effect, if used subtly. In this instance, if the resonance is set higher, then you will have a more dramatic effect. Using a sine wave and assigning the LFO to modulate the resonance at a high value will give you the siren type of effect, or if you assign a lower resonance, you will get a deeper throbbing effect.

These are all dependant on where you select the cut-off point to be. Using resonance in more subtle ways (higher resonance values on a low frequency bass sound) can bring out the higher harmonics of low frequencies. This is a great way to depict, ‘presence’ and’ perception’. Although you are boosting the higher frequencies of a low frequency sound, your ears perceive the sound as more pronounced. The same can be applied in reverse. Using a low resonance on a high frequency sound will give the perception of more bass or low end. Another great trick with resonance is to attain that Moog type of squelchy bass sound. I usually draw a graph or, input values, that displays a negative filter (ie, the filter start at the negative and rises to the positive) and then assign a lower frequency and higher resonance. I then assign a velocity curve to the filter attack and this, when a note is hit, gives that open/close filter effect on the bass sound. The squelch. Remember the envelope topic in the earlier part of the tutorial? Well, this is what I am shaping, the filter envelope.

If you recall, at the start of these tutorials, I gave a huge list of useful source and destination routings, terminology, components etc. In there, you would have seen Filter Env and Filter Amt.

Filter Env is the filter envelope, much like the ADSR envelope we discussed in and earlier part of this tutorial. This should make sense to you now. The filter has an envelope as well and it can be shaped much in the same way as the Osc envelope.

Filter Amt refers to the amount of movement there is at the filter cut-off. The filter env and filter amt go hand in hand. The higher the amount, the more open the filter, the lower the amount, the less open the filter. This defines the filter envelope amount at the cut-off.

How many Trance tracks have you heard where a sound starts muffled and then gets brighter and brighter? This is the filter opening up. It starts closed and opens up over time. You can easily create this effect by having your mod wheel assigned to the filter frequency (FilFreq). As you move the mod wheel up it opens the filter by raising the frequencies. You can then close back down again. There are other nice little tricks you can use. Try assigning the mod wheel to resonance. This is great for lead sounds.

You could also use the technique we touched on earlier, by choosing white noise as the osc and having the LFO modulate the resonance of the osc, and using small amounts of resonance and slow LFO rate, you will get an ocean wave effect. You could even have a sine as the osc, use an LFO with a mid rate, assign it to the resonance (low values) and get a bubbling type of effect, like in sci-fi films. You could even create the effect of static by using the white noise osc and set high resonance values and have very high LFO rate at a higher pitch, assigning the mod wheel to the resonance means that by using the mod wheel fervently, you can create a disjointed, or modulated, static type of effect. You could even throw in a square wave osc into the equation with a very high rate and have that running with the white noise osc at the same time. Madness I tell you, madness.

There is so much you can do with just these simple parameters, imagine what you can do once you have completed this entire tutorial?

I think now is a good time to explain the distinction between passive filters and active filters.

For me to clearly define the differences, I would have to explain some basic electronics, but I do not feel that you are quite ready, at this point, to enter that world of madness. So, I will try to explain in very layman’s terms and very briefly the differences between the two. This is not crucial to know, at this stage anyway, as you will not really need to know or adopt the distinction unless you were to build your own filter. However, it does help in understanding the basic concept of how a filter works in the circuit world.

Passive filters are filters that derive their power from the input signal. Another way to look at this would be to say that passive filters have no power of their own until a signal is passed through them, then they wake up and get to the job. That is very simplistic but, to a certain degree, true. The amplitude response and phase response of a filter are crucial in determining the result of what is put into a filter and what comes out. The relationship of what goes into a filter and what comes out is called the Transfer Function. Phase is a subject that is very important and one that I will deal with in depth, at a later date, when you can fully understand what I am talking about.

Active filters. To explain this you would need to look at a very basic test case scenario. Imagine that you had to build a circuit board and had to have a band pass filter in there. Now, we know that having a low pass filter and a high pass filter, together, creates a band pass filter, because we attenuate the frequencies below and above the cut-off, and we are left with a band of frequencies. So, you think by putting a low pass filter, followed by a high pass filter, on the circuit board, you would get a band pass filter? I wish. It doesn’t actually work that way and is much more complicated than that. The reasons are simple.

We know that the filter elements of the two filters interact, that their cut-off frequencies would be different for each filter, and that there are phase shifting effects for each of the filter stages. I don’t want to go into explaining each of these elements as I will come to them at a later date. So, trust me on this. Putting a low pass filter and a high pass filter together on a circuit board requires a little more than just a bit of component soldering. What you do need between these filters, are called operationalamplifiers, (op-amps). These components separate the filter elements from each other. This is what makes a filter active, the fact that they have op-amps placed between the filters.

Most analogue synthesizers have active filters. There are exceptions but the majority have active filters. The reason that you can distinguish a filter from a Moog to an Arp is the fact that they have design elements in their circuitry, mainly the op-amps, that separates them from their counterpart and gives the make and model it’s own unique character or sound. Otherwise a low pass filter is just a low pass filter and should sound no different if it were in a Moog or Arp or any other make. This is a useful bit of information, as I am so often asked what makes a certain make and model sound different to it’s counterpart. You now know. There is , obviously, more to this than just having op-amps on the pathway. The design of analogue filters varies as well. The filter itself has it’s own character as the component parts are ‘rated’ and they are never 100% rated (technical term). This is in the hardware design, measured and tested. If one component is rated slightly differently to the next component, then it will have a slight variance in it’s function and performance. This also adds to the characteristic of the filter.

Next time we fall deeper and deeper into the synthesis abyss.

Join me for the adventure.