Author Archive for

12
Mar
10

From How and What to When and Why – Part 2

In Part 1, we discussed about oscillators, what characteristic sounds they have, and what to select when making different sounds. In this tutorial and the next, we’ll focus on filters and volume envelopes. Both of these are fairly straightforward, and if you’ve been following along, you should have already tried out different settings for the filter and volume envelopes.

Let’s start with filters. You must know the concept by now. Generally, you’d use a high pass filter to cut out low frequencies, and increase the brightness of the sound. And low pass filters to cut out the high frequencies, to make the sound duller or bass-heavy. As long as you know this basic idea, you are well set to use filters. You’ll obviously know what type of filter to set depending on how you want the sound to be modified. And the cutoff frequency is usually adjusted by sweeping through gradually, till the sound is just right. After that, you can add resonance to taste depending on what it does to the sound at the cutoff frequency you’ve set.

All these are very dependent on the sound you are trying to make, so it’s hard to give you any rules or guidelines to follow. However, since the usage is obvious, no rules are necessary. But there are a couple of special applications of filter that you should know.

As a mixing utility

When a mixing engineer mixes a song, it is common practice to filter out some low-end from each of the tracks. The reason for this is:

a. We do not have the equipment to both reproduce (speakers) or hear (our ears) such low frequencies.
b. Having a lot of unnecessary low frequencies have the effect of “muddying” up a song.

For a song in the mixing stage, if the individual instruments have to sound clearly, there shouldn’t be any frequencies in there that isn’t required and that makes no difference to the perception of sound. High and low pass filters do a good job of elimination such unwanted frequencies from both ends. If you are making a screaming synth lead patch, there is no need of low frequencies in the sound. But some of them might come up, especially if you use an oscillator waveform that has lot of harmonics (like saw). So, it’s a good idea to high-pass the sound with a suitable cutoff (One that does not take essential frequencies away and change the sound). Just make sure you use the keyboard track feature on your synth, so that the cutoff is uniform across notes.

To add interest to the sound

We already discussed filter sweeps in the filter tutorial. The idea is to have a small amount of resonance, and to change the cutoff frequency while the sound is playing. This leads to different movements of sound.

If you sweep a low pass or high pass filter, you can get a sound that starts off too dull or too bright, but then adds on frequencies to finally get to the original sound.

If you sweep a band pass, you get different types of sound ranging from dull (when cutoff is in the low-end) to telephone like sound (when cutoff is in the middle) to sharp shrill sounds (when cutoff is in the high-end).

These effects are usually achieved by automating the cutoff frequency on your DAW. Consult your DAW manual on how to automate plugin parameters on your DAW’s track.

Note 1: I still haven’t talked about band-cut or notch filters. One way you can use them is to remove particular frequencies in your sound that you don’t need. Automating them is also possible, but the effect is not as dramatic as the other filters, and could almost go un-noticed in a busy mix. There are times when it comes to good use. We’ll see that soon.

Note 2: Be careful about the resonance knob. When you set a resonance value, what you ask the synth to do is to boost the cutoff frequency by that much, irrespective of how loud the frequency already is. If the sound from the oscillators already have predominance in that frequency, the resonance can easily over-boost, resulting in extremely loud sounds, and may even damage equipment (your speakers or ears!). So, avoid setting extreme values. Resonance is anyway to be used just like a sprinkling of salt. Too much of it rarely sounds good.

Okay… on to part 3. We’ll next learn what volume envelopes to use when.

08
Mar
10

New tutorials coming up!

It’s been quite a while since I’ve written here. Things have been very busy lately, and I haven’t had any time to spend on writing new tutorials. I hope you understand, I had to prioritize activities that earn my bread first.

Well… the good news is, I’ll soon start writing new tutorials. Among the first things I’ll cover are LFOs and filter envelopes, and then I’ll be moving on to some advanced concepts. Also, as promised, I will be writing a series of “Make a sound” tutorials where I start off with a sound, break it down in a synth and show you how to build it from the scratch.

Thanks you for all your comments and messages. I never expected that these tutorials would be so well received, and be of such value to so many! I’m overwhelmed by all the response I’ve got. This response has made me more determined to spend time on this, and contribute more. Thanks again!

13
Jan
10

Oscillators FAQ

In our previous tutorial, we talked about the characteristics of different waveforms and the type of sound that each waveform can produce. Before we move on to part 2 of the “When and Why”, here are a couple of the common questions I get, and my usual answers to them. If these questions popped up in your mind too, this should clarify!

Q. The example sounds you’ve mentioned are real world instruments. Like sawtooth for strings, and square for clarinets. But what if I want to make my own sounds that does not model any real instrument? What if I want to make a sound that hasn’t ever been made or heard yet?

A: Well actually, most of the times, that’s the idea! When programming synths, you’ll usually want to create a truly unique and unheard-of sound. Of course, you can also program synths to model a real world instrument. (There are some excellent patches in many commercial synths that model real instruments, and they sound so close to the real thing). But for me, personally, the fun is in coming up with something truly new and amazing that sounds great in the context of the track, and thus makes the track sound great. (If I want a ‘real’ and ‘authentic’ sounding piano on a track, I prefer to use a sampler, rather than programming it to make it sound authentic).

Of course, this is just my personal preference. If imitating real instruments on your synth interests you, then let me, by no means, discourage you! It’s a very challenging and interesting task, as the human ear is good at distinguishing between, say, a recorded piano sample and a synthesized piano note. So, it’s really a true test of synthesis skills!

I think I’ve digressed. To answer the question, the real instrument examples were to give you an idea about the timbre of sounds you can generate, so that you can choose the right waveform as a starting point. So, for example, since you know that saw is good for string-like sounds, that’ll be your first choice of waveforms for making that “Aliens Playing Strings on the Frozen Sea on Mars” patch.

Q: The timbre of a waveform is well and good, but we are using two waveforms here! How do I know which combination to use for a given sound?

A: The short answer is practice :). I’m kidding. (A little!). It’s usually a mixture of timbres that you go for. Say, you have a smooth sub bass that you’ve programmed with just a sine. Now you want to add a bit of a rough edge to it. You can choose a saw or a triangle, and adjust the mix knob according to your taste.

Clearsynth has 4 waveforms to choose from. There are many synths available that offer a much larger number of starting waveforms, apart from these four. Each of those will obviously have a particular timbre that, once you get familiar with, you’ll instantly recognize. But unless you play around with the sounds, it’s hard to say what waveforms need to be selected for which sound. After a little while, you establish a mental map of waveform to timbre, so, when you program a sound, you break it down to the timbres that are achievable in the waveforms your synth provides.

11
Jan
10

From How and What to When and Why – Part 1

Congratulations! We now have enough knowledge to make a variety of sounds from Clearsynth!

Just to recap what we’ve learnt so far:

  1. We know how to use the two oscillators together, and mix them with varying volumes
  2. We know what filters are, and how to select different types of filters.
  3. We know about volume envelopes, and how to control them.

For someone starting out, these present a lot of options. In this tutorial, and the next, we’ll look at when to choose, say, a sine wave versus a saw wave, when to lowpass and when to highpass, what volume envelope to use etc.

Let’s start with Oscillators.

If you’ve been following along the tutorials, you’ve already heard the four different waveforms in isolation, and know what they sound like. Sine, saw, triangle and square, each have their characteristic ‘sound’. Why is that so? It’s all well and good to say that these waveforms look different. But what is the difference between these waves sound wise?

Remember, I told you before that any sound we hear is a mix of frequencies? Play your favorite record on a spectrum analyzer, and you’ll see a whole lot of frequencies jumping up the graph, from the lower frequencies of the kick drum, to the higher frequencies of the hihat. Even the common sounds you hear every day, like the slamming of a car door, the bark of a dog, or the noise of a vacuum cleaner, are all made up of a mixture of frequencies.

Ever wondered what a single frequency will sound like? Select the sine preset on Clearsynth (or choose a single sine waveform with everything else off) and play a note. That’s how it sounds like! 🙂

The sine wave

A sine wave is called as a fundamental waveform because of this. Explaining about these waveforms involves a whole lot of scientific formulae, swinging pendulums and French mathematicians, so we won’t get into all that. Think of a sine wave as a basic single frequency waveform (also called a harmonic). It’s the ‘atom’ of the sound world: there is no waveform more basic than a sine wave. This waveform is sometimes called as a “pure tone”.

This makes a sine wave useful for generating smooth tones, and sub-bass sounds. But sine waves are usually not used in Subtractive synthesis as much as the other waveforms. The reason is simple. In Subtractive synthesis, we take waveforms from which we subtract frequencies from to get the resultant sound. If the source waveform is a sine wave, then it’s a single frequency, and subtracting the only frequency that it has will end up in silence!

Generally, sine waves are used along with another waveform to add the sub bass, or smoothness to the sound.

The sawtooth wave

If sine wave has only one frequency, the saw tooth wave, however, has all the possible frequencies. So, while the sine wave sounds simple and smooth, the saw wave sounds buzzy and rich.

Note that I said the saw wave has all the possible frequencies. What do I mean by that? For any sound to have a pitch, there has to be a fundamental frequency, and other frequencies that are a multiple of a fundamental frequency. Saw waves have every multiple of the fundamental frequency: that is, the fundamental frequency + 2 times the fundamental frequency + 3 times the fundamental frequency… and so on. All these frequencies are at different amplitudes (a ratio of the frequency multiple).

Yeah… I know I said I won’t get into the math of it all. It’s perfectly okay if you don’t understand the math. What’s more important is to get the feel of the sound. Just realize that a saw wave has the most harmonics among the four waveforms we’ve seen. This results in the buzzy sound. Which is what makes it ideal for subtractive synthesis, as it is a rich source of frequencies, and there is so much to subtract from.

Because of this richness, saw waves are an ideal starting point to synthesize brass instruments like trumpets, and string instruments.

Note: It’s not a random choice why I asked you to choose the saw waveform when I was explaining filters. Try the same filter exercises on a sine wave and hear the result. Since there is only one frequency there, you cannot hear the effect of a gradual filter cut. The sound is either on or off.

The square wave

This wave has only the odd harmonics. The even harmonics are not present at all. So, it’s slightly duller than the saw, but it still has more richness compared to a sine wave. Square waves can be used for brass-instrument-like sound too, but for those which have slightly less character (like the clarinet or sax, for instance).

The triangle wave

This wave too has odd harmonics, but the volume of the harmonics dampens out quickly. This makes it duller than a square wave. Triangle waves can also be used along with other waveforms to add extra character to the sound.

I know this is a whole of theory for a single tutorial. But trust me, this is really important. If you’ve put the effort and read this through, then go and listen to the waveforms now. It’ll all fall into place! Not only will you have a mental picture of why the different waveforms sound different, you’ll also instinctively know what waveform to use when you are looking for a particular sound.

On to part 2. The next tutorial won’t have this much theory, I promise. All the hard work is over. 🙂

08
Jan
10

Controlling the volume envelope

Now that we know how envelopes of different sounds look like, it’s time to tweak Clearsynth and find out how different envelopes sound like!

Set Clearsynth to the basic saw preset. We’ve seen the envelope of this sound in our previous tutorial to be something like this.

Image 1: Default saw preset

Observe the panel at the bottom right corner of Clearsynth called Envelopes. There are many knobs here, but some of them are already familiar: attack, decay, sustain and release. However, there are two knobs of each! That’s because this panel can generate two separate envelopes. Each of these two rows of knobs allows us to control these two envelopes separately. Now, why would we need two envelopes?

What we learnt in our previous tutorial is just one type of envelope: the volume envelope (also commonly called the amplitude envelope). There are other types of envelopes possible, and more things we can do with them. But that’s for later. 🙂

Now, focus on the bottom row of ADSR knobs. That’s the one that helps us adjust the volume envelope.

Attack

In the saw preset, the attack knob is turned to zero. See what happens when you increase the value of the knob. The sound builds up in volume more and more slowly as you increase the value. What you are controlling with this knob is the time of the attack phase. The greater the value of this knob, the longer it takes for the sound to build up to the maximum value. At about 60%, the s(M)exoscope shows the waveform like this.

Image 2: Waveform with increased attack

Here is a small sound clip that has a single note played, with a gradually increasing attack value.

Decay

Let’s look at the Decay knob now. Just like the attack knob controls the time duration of the attack phase, the decay knob controls the time duration of the decay phase. Set the attack knob to 50% and play around with the decay knob, to hear how it sounds.

Here is the waveform with the decay knob set close to the middle.

Image 3: Waveform with an increased decay value

And here is a sound clip with the decay value starting from zero and gradually increasing.

Sustain

This knob is slightly tricky. Remember that the Sustain phase lasts for as long as you press and hold the note. How can you control the duration then?

Well, you cannot. It is entirely dependent on the note played. Play a short note, and the sustain phase is short. Play a longer note, and the sustain phase is longer. What is this knob doing then? It is actually controlling how loud the sound has to play when the sustain phase is reached. So, while the other knobs control the time duration, the sustain knob is the exception. It actually controls the volume, or the amplitude of the note played during the sustain phase.

Try reducing the value of the sustain knob as you play a note. As the value reaches zero, the sound becomes short notes, no matter how long you press and hold the key.

This is how the sound looks with zero sustain. Notice that only the attack and decay phases are visible.

Image 4: Waveform with zero sustain

Here is a sound clip with the sustain value gradually reduced towards zero. The length of the note I play isn’t changing. But as the sustain value decreases, the complete duration of the note sustain becomes less and less audible, until it finally disappears, and you hear just the attack and decay.

Release

This knob controls the duration of the release phase. Using this, we can make the synth play the note for sometime even after the key is released. If the note has to stop immediately after the key release, the knob has to be set to a very low value, or even zero. Having a good release time makes the end of the note less abrupt.

Here is a sound with a long release.

Image 5: Waveform with an increased release value

This sound clip plays notes with gradually increasing release time starting with the release time of zero.

Now that we understand how to control the volume envelope, in the next tutorials I’ll be taking about something which is probably more important when it comes to learning synths. The standard “when to use what” question. I’ll also try to answer a few common questions asked about volume envelopes. Keep reading!

Note: If you are having trouble playing any of the audio clips, drop me a comment and let me know.

06
Jan
10

Understanding envelopes

In this tutorial, I’ll introduce you to the concept of envelopes. We’ll see how different envelopes look like, and how they sound.

Start with a fresh project (or track) on your DAW. Load Clearsynth and s(M)exoscope on the track. Load the “saw” preset in Clearsynth. Remember, in the previous tutorial, we modified the s(M)exoscope’s parameter to see the detailed waveform? This time we are going to leave it at the defaults. Play a short note and observe the waveform. It’ll look something like this.

Image 1: A short note with the saw preset

Try this again. This time, press and hold the note for a longer time. You’ll see something like this.

Image 2: A longer note with the saw preset

There are a few observations that are common in these two waveforms.

1. When you press a key, the volume increases from zero to a particular value.

2. From this value, the volume then immediately decays down to another value.

3. This value will now continue to sustain for the duration of time that you hold the key down.

4. When you release the key, the volume fades down to zero.

Every sound that you hear has some volume pattern like this, and it draws up such shapes when you view it on s(M)exoscope. Some are similar to what you just saw. Some are drastically different.

See how a kick drum looks.

Image 3: A kick drum

Notice how it goes from zero to maximum in an instant. However, from maximum, it immediately starts decaying its volume to zero. Also note how the volume does not sustain. A kick is supposed to be a one-shot sound. So, no matter how long you hold the key, the length of the sound does not increase.

Let’s look at a slow string sample now.

Image 4: Slow strings

As you can guess from the nature of the sound, this one starts very slowly. But it doesn’t seem to decay to a lower value. The sustain part is prominent, as you’d expect. You’d want the string sounds to be heard distinctly as you press and hold the key. There are lots of changes in volume during the sustain period. But essentially, the average volume is the same. Once you release the key, the fade out to zero is also quite slow.

In all these cases, what you’ve been observing is called the “volume envelope”. It’s the behavior of volume, and its changes over the lifetime of a single note of sound. As I said before, every sound has such an envelope. In the synthesis world, this is broken down to four parts in order to understand and control it.

The parts are the same 4 observations we made above.

1. The part where the volume goes up from zero to maximum as soon as you press the key. This is called the attack phase.

2. The part where the volume decays down from the maximum to a new volume level. This part is called the decay phase.

3. The part where a constant volume sustains throughout the duration of the key press. This part is called the sustain phase.

4. The part when the key is released, and the volume goes back to zero. This is called the release phase.

We can control the volume envelope of the sounds that we generate in synths by controlling these four parameters. Note that this is just one way of controlling volume envelopes. Some synths provide more finer controls, but these four controls will most probably be there anyway. This type of envelope is called the ADSR envelope (The letters stand for the names of each phase of the envelope).

Image 5: Standard ADSR envelope

This is the envelope control used in Clearsynth. See if you can identify these phases in the sound Clearsynth makes (or in Image 1 of this tutorial).

In our next tutorial, we’ll see how to control these four parameters to shape our sound.

05
Jan
10

s(M)exoscope

Before we begin to learn about the envelopes, I highly recommend you download and install this freeware plugin. s(M)exoscope is a free real time signal analyser plugin that shows a nice graphical display of the signal that is fed to it. No matter whether you laugh or cringe at the name, it’s a very valuable tool in learning synth programming.

In my previous post, I wrote about the importance of the ability to visualize sounds. This is one of the many useful tools that helps you do that. This was a tool that helped me a lot when I started learning. (Well… I’m no expert yet, and I still use it. 🙂 ) Spending time on this will establish a mental connection between sounds and images, and that’s very useful. The more you use it, the more you don’t have to use it!

Let’s see this plugin in action. Load it as a VST effect, following Clearsynth. Play some notes on your keyboard, and notice the waveform that appears on the screen.

The default settings shows how the waveform behaves over a larger period of time. While this is useful in what we’ll be learning about in subsequent tutorials, for now, we need to decrease the display time to see what we’ve learnt so far. Notice the Time knob on the upper left panel.

Set the value of the Time knob to around 0.200. Now, if you play notes, you’ll be able to see the nature of the waveform. In Clearsynth, select only Osc 1 and turn off the filter. Now choose different waveforms and see the waveforms being rendered in s(M)exoscope. Add Osc 2 to the mix and see how the waveform appears. Compare the saw waveforms with the waveform diagrams in this post. Now switch on the filters and see how it affects the sound. Play different waveform and filter combinations, and watch them on s(M)exoscope, till you get an idea of how these waveforms look.




Enter your email to subscribe and receive email notifications when new tutorials are available.

Join 84 other followers

Categories

Blog Stats

  • 81,408 hits