klystrack tutorial - fx section

In this third tutorial, we'll take a quick look at the FX section. I say quick, because it is a relatively small section and it shouldn't take much time to explore and describe what each setting does. The FX section has one global effect called the Multiplex and four separate inserts (hereafter called simply unit)  that each have the same three effects, in the same static signal path

What's an insert? The name simply means that the effect unit is inserted between the sound source and the output. Think of a distortion pedal for a guitar. You insert it between the source (guitar) and the output (amplifier). There are other ways to apply effects to a sound (send/returns, mid/side, dark sorcery, etc) but Klystrack only uses inserts, meaning that when you send an instrument into an effect unit, you will have very little control over the dry (original sound) and wet (effected sound) balance.

What's a signal path? As the name says, it's the path that the sound will take through multiple effects. Let's make another guitar analogy. Let's say you have three pedals, distortion, chorus, and delay. The order in which you plug in these pedals together is your signal path. If you plug them Guitar > Delay > Chorus > Distortion, then your effects will be applied to the sound of everything prior to it. The delay would process the guitar sound as you'd expect it. Then the chorus will process all of that, delay taps included. And last, the distortion will process all of that once more, distorting the chorused and delayed guitar which should technically ruin everything you're trying to accomplish and should serve lesson to always put the distortion first in the chain. 

For each of the four FX units, Klystrack uses a fixed signal chain which goes Chorus > Delay > Crusher. In the software, these are called Stereo, Reverb and Crusher respectively, but I can't bring myself to call the first two that way. It's like calling a pixel "an image-dot" or calling a car "a moving people box". Let's call a dog a dog and use the proper terms for these effects shall we? We shall. Right then, moving on...

We'll start with the weirdest effect, because...reasons. What the multiplexer does is basically cycle through each of your channels one by one, playing the sound of the one it's on for a specific number of ticks while muting the others. Sounds confusing? Read on, it gets easier.

The multiplexer, also known as the what-the-hell-is-going-on effect.
Period is the number of ticks that each channel will play for. So for example let's say you have a song with four channels, running a speed 06 (6 ticks per row) with the multiplexer's Period set to 3. When you press play, the first row (00) will play only what's in channel 0 for the first three ticks, then only what's in channel 1 for the last three. On the second row, it will play channel 2 for three ticks, then channel 3 for three ticks and then this restarts from channel 0 on the next row and so on and so forth. Since this value is expressed in ticks, the lower the number, the faster the multiplexer will cycle through your channels.

Inaccuracy is a setting that will mess with the pitch of each note played in the song. Values range from 00 to 12. What it seems to do is grossly quantize the pitch of your notes. For example at it's highest setting, it really doesn't matter what note the multiplexer encounters, it will always play the same. At a slightly lower setting, it will group C,D, and E together at the same pitch, then F, G, A to another etc. All of these pitches are approximated and often discordant and this just as it was intended, the multiplexer is meant for glitchy experimental music.

The crusher is an effect that performs three functions. The first is bit crushing (hence the name) which reduces the bit depth of your sound. The second function is downsampling which reduces the sample rate of your sound. And the last function is a dither which smooths out the overall result of this effect. 

Ze bit crusher. Used to crush bits, should bits require to be crushed.
BITS is the number of bits by which to reduce. The bit depth of a sound is the "vertical" resolution of a digital sample. If you look at a waveform, you will see three lines. The one in the middle is 0, and the top and bottom lines represent the maximum bit resolution you have. The higher the bit depth, the more vertical "slices" you have between 0 and the top/bottom values. Remember that this is binary, not decimal, so for example 2 bits can express 4 numbers (00, 01, 10, 11) which would translate in a +2/-2 range for a waveform: two above zero and two below. This grows exponentially as you add bits. At 16 bits, you have a range of 65536 possible numbers, effectively giving you a range of +32,767/-32,768. The more you crush the bit depth, the lower the quality of the sound, and the closer to a pure digital square wave (1bit) you get. Since you are losing vertical resolution, you are also losing the ability to ramp up smoothly in volume so things like the volume envelope will start breaking up as you increase the BITS value. This effect can, in some ways, be used to create distortion effects or make some instruments sound more noisy.

DSMP is the downsampler. Values range from 00 to 64 and the higher you go the lower your sample rate gets. If the bit rate is the vertical resolution, then the sample rate is the horizontal resolution of a digital sound. Put (very) simply, the higher the sample rate, the more horizontal "slices" can be played per second. In rough not-too-technical terms, this means you can play higher frequencies more accurately. By downsampling a sound, you are effectively reducing the number of samples per second and thus, your ability to have proper high frequencies in your sound. On simple waveforms like you get from the synth's oscillators, this effect will give a lo-fi quality to your instruments. Be careful though, as you will quickly lose the ability to play high notes as you increase the DSMP value. Used creatively, when the sample rate syncs up with the frequency of your note, this can create pleasing "bell" sounds, or give nice tremolo effects.

Dither is the application of noise patterns to smooth out gross quantization. This permits a signal to be of relatively low quality, but still retain parts of its original character and sound. In Klystrack, the dither is very crude so do not expect it to make a crushed/downsampled instrument sound like the original signal, but it can be used to smooth out the effect if it gets too rough. It will introduce a sort of noise to your signal, which is very interesting when you want to achieve a certain lo-fi quality, which I assume is what you're going to be using the crusher for in most cases anyways.

The Volume is used to keep sounds in check. As explained before, reducing the bit depth of a sound has a tendency to increase the volume. This setting is basically just an adjustment to bring back a rogue sound at more manageable levels so you are able to mix it with the rest of the song without having to boost everything else in the stratosphere. 

A chorus is a time, and pitch-based effect that basically makes a copy of your sound and delays it a bit from the original while applying slight pitch variations. Klystrack's chorus is a little different from most chorus effects in that it uses the same settings for both the time delay, and the pitch modulation. While this is less flexible than a standard chorus, in most cases it will do just fines as very rarely will you want to drown your sound in thick chorus, and if you do you will probably want to go all out bat shit crazy anyways. 

It's called stereo, but it's really a chorus, dammit.
Now I don't know all the details on how this particular chorus was programmed, but what it seems to be doing is making a copy of the original sound, and applying the effect to both the copy and the original at opposite polarities. Meaning that when one of them is chorused at "+25ms" then the other one is chorused at "-25ms". I might be wrong, but since you lose the original signal entirely through this effect, I doubt that it would have been programmed to make two copies and then mute the original signal. But hey.. what do  I know?

Min/Max are two settings used to determine the minimum and maximum amount of delay that will be applied to the copy of your original sound. The chorus will oscillate between this two values according to the speed you set in Mod. Values here are in milliseconds, from 0.25ms to 25.5ms.  Note that these values are also used to determine by how much the pitch of the sound will be modulated.

Phase is the separation between channels (audio channels.. meaning left and right). Values go from 00 to 40, which essentially means from having the two signals dead center (mono) to having them completely separated left and right. Any value above 00 will give your sound increasingly bigger spatial placement. This can be very useful to separate two instruments that tend to muddy up each other in a mix. If two of your instruments are competing in frequencies and bury each other, try leaving one in mono, and then chorusing the other one into the hard left/right.

Mod is the speed, in hertz, at which your chorus will oscillate between the values set with Min/Max. Again, just to make sure the message came across, this affects both the delay and the pitch modulation of the effect. Also, this affects both the delay and pitch modulation. And always remember that it will affect both delay and pitch modulation.

Yes, it's called reverb, I know. And technically speaking it is a reverb since its taps are only played once. But to achieve a realistic reverberation effect, you'd need a metric shit ton more taps than the 8 that are offered here. And yes, I know, it's not a delay either since the taps do not repeat according to signal feedback and bla bla bla. Point is, at the end of the day you will use this for delay effects about 99% of the time, so there. And while it is neither usable as a reverb, or definable as a delay, this effect offers you a lot of possibilities with the way it is set up. With this effect, you'll be able to create (very) short reverbs, delays that can be synced to the tempo of the song, chorus effects without pitch modulation and flangers. Think of time as a time-based effect factory.

The big box of taps. Used for tap-dancing, mostly.
Roomsize, Volume, Decay and Set are settings that you use to generate the time and volume of each tap automatically. Those settings do not go very high, and thus will tend to produce exactly what was intended with this effect: SNES-like reverbs. That is, short and quickly fading echoes. Roomsize will determine the time of each tap. The higher the value, the longer the time between taps. Volume will decide at which volume the first tap is while Decay will determine how fast they fade out. Once your setting are where you want them, press Set and they will be applied to the taps.

Spread is kind of like the chorus effect and will only affect the taps, not the original signal. Values go from 00 (mono) to FF (complete separation left and right). Just like the chorus, this can be used to push a sound outwards from the center of the mix, but unlike the chorus there is no pitch modulation applied here.

Editing the taps yourself however, is where this effect shines. There are two settings to each of the 8 taps available. The first is the time, which can be expressed either in milliseconds, in increments of 1ms, which is what people used to "real" effects might prefer to use. Or you can have it in ticks, in increments of 0.05 ticks which might be more up the alley of people who learned music in trackers. In both cases, the increments are the same, 1ms = 0.05 tick. The second setting is the volume which is expressed in how many decibels away this tap will be from the original signal's volume.

Take a look at the settings shown in the above image. Each tap is 100ms apart starting at 50, and each tap is -3db lower than the previous one. What that means is that 50 milliseconds after a sound is sent to the delay, the same sound will play again 3 decibels softer, and again at 150ms, this time 6 decibels softer than the original volume and so on until all 8 taps have player.

You will notice there's a button between each set of arrows for the time. In the image above, these are marked with 0, 1, 2, 2, 3, 4, 5, and 6. The number represented here is a rough approximation of how many pattern rows away from the original note this particular tap will play. You can press this button to set the timing exactly to that number of rows. For example if you want a tap to play precisely two rows after your note is triggered, then move the ms/tick setting until you see the number 2 on the button, then click the button. This will adjust the timing to precisely what it needs to be to hit at exactly 2 rows. Note that while the actual the timing for each tap is not affected by the song's speed, the number shown on this button will actually change to show you the number of rows it represents at the new speed. If you change your song's tempo mid-project, it would be a good idea to re-set your delays to fit the new tempo.

Similarly, there are two buttons next to the arrows for the decibel setting. -INF means minus infinity, which essentially means this tap is silent and will not be heard. The 0 button is the opposite and sets the tap's volume to a 0db reduction, meaning it will play at the same volume as the original.

For classic delays, pick a number of rows, and increase each tap by that same amount (1,2,3,4,5,6,7,8 or 2,4,6,8,10,12,14,16 or 3,5,9,12,15,18,21,24 for example). Make sure each new tap is played at a lower decibel value than the last and there you have it, a bread n butter type of delay. You can of course get much more creative with it, but since you have four separate FX units, it's always a good idea to have one generic delay on hand.

By using very low values for the times, you can also create flanger and chorus effects. Setting a single tap  to 1ms with a 0db reduction will essentially give you a second identical copy of your sound delay by a single millisecond, thickening its sound and changing its timbre. Play around, explore the values and you'll discover a ton of ways to change the timbre of your instruments this way. Be careful not to use values that are too high for the timing, otherwise you'll fall in reverb/delay territory. 

And so that's about it for the FX units do and what each setting is used for. Remember that nothing beats noodling around, so explore and have fun and you'll find that these effects extremely useful for the sound they open up, and the time they save. Hell, who wants to waste channels and time to edit echoes by hand in this day and age?

Stay tuned, next time we'll go over how to personalize the Klystrack settings to suit your needs. And after that, with all the technical details out of the way we'll finally get to dig in the fun stuff and make some actual music.