Tags: posts polarity-music Bitwig Generative Sampling Modulators Presets Tutorial

Creating natural melody lines with complex LFO waveforms in modular synthesis

Tutorial | Feb 27, 2020

In this video, I share a patch that uses multiple LFO modules to create a more natural melody line. I blend different LFOs together to create a complex waveform for the pitch information, which is then narrowed down to a pitch scale and goes into a sample and hold module. The rhythm is also generated randomly using an LFO and the gates module. The whole patch is controlled by an overall volume LFO, but the only downside is that it's completely quiet after 32 bars. However, this can be easily fixed by offsetting the LFO in the second patch. Overall, this patch provides a more procedural approach to generating notes and rhythm, and experimentation is limitless.

You can watch the Video on Youtube - support me on Patreon

Questions & Answers

Maybe you dont watch the video, here are some important takeaways:

What was the inspiration behind this patch?

The inspiration behind this patch was to create more natural-sounding melody lines. The creator previously showed in one of their videos how to create a step sequencer with dice modules and a merge module, but they found that the resulting notes were too random. They wanted to make the melody lines more predictable, but still have some randomness to it.

How did the creator use LFO modules in this patch?

The creator used a lot of LFO modules in this patch. They started with a standard sine wave LFO over 8 bars, but then blended in other LFOs to create a more complex waveform. They then used this waveform for the pitch information, with the pitch scalar narrowing the values down to a pitch scale between D sharp 1 and D sharp 6. The pitch quantizer then narrowed it even further down to a node scale in D sharp minor. The LFOs were also used to control the rhythm, with a mixed-in LFO sine wave form generating a more interesting and randomized rhythm.

What was the problem with this patch and how did the creator fix it?

The only problem with this patch was that it was completely quiet after 32 bars. However, the creator fixed this by duplicating the patch and offsetting the controlling LFO by 180 degrees in the second patch. This created a back-and-forth effect between the two patches, with one fading in while the other faded out every 32 bars.

How can someone modify this patch for their own experimentation?

Someone can modify this patch by experimenting with different module types and parameters. Instead of just using blend modules, they can try using math modules for different effects on the resulting waveform. They can also try exchanging the blend module for ring modulation or amplitude modulation. There are no boundaries for experimentation with this patch, and the resulting waveform can be influenced by a variety of module types and parameters.


This is what im talking about in this video. The text is transcribed by AI, so it might not be perfect. If you find any mistakes, please let me know.
You can also click on the timestamps to jump to the right part of the video, which should be helpful.

[00:00.000] So today I want to talk about this patch here, I post it recently on my YouTube channel
[00:07.400] and what my thinking was about it.
[00:15.720] Use my code polarity in the shop to save 10% on the price and support my channel.
[00:21.440] So as you may be already noticed in this patch are a lot of LFO modules happening and this
[00:28.120] is for a reason.
[00:30.320] In one of my videos from last year I think I showed you how you can build a step sequencer
[00:35.520] with dice modules and a merge module where I have generate random values for each step,
[00:42.880] for each pitch step and sometimes this is pretty fine but all these values are all over
[00:49.000] the place so I have notes occurring pretty randomly and I wanted to behave my melody lines
[00:56.000] more naturally so I tried to accomplish this with the LFOs and I started here with this
[01:03.120] bottom LFO and this output is a pretty normal standard sine wave over 8 bars and I tried
[01:12.040] to make this sine wave more complex by mixing in other LFOs and you can see I mixed these
[01:17.760] two together or blended these two together, then I blended these two together and then
[01:22.400] I used the blend for these two pairs and then I have in the end here a much more complex
[01:27.800] waveform and this waveform is used for the pitch information and also you can change
[01:37.080] how this waveform looks by changing the speed of each of these LFOs right so this gives
[01:46.800] you some kind of predictable random output that behaves in a slowly evolving manner and
[01:55.480] these pitch information or these sine wave form in the end here, this complex wave form
[02:03.440] goes straight into the pitch scalar which narrow the values down to a pitch scale between
[02:09.440] D sharp 1 and D sharp 6 so it's audible and audible range and then I go into the pitch quantizer
[02:15.560] which narrows it further down to a scale to a node scale and this case here it's D sharp
[02:22.760] minor only missing this one here and then it goes in the sample and hold and the sample
[02:29.360] and hold samples the pitch information only when a gate trigger is occurring right and
[02:36.640] then it goes into the sine oscillator and then to the output and in the output here you
[02:42.320] can see I have a multiply installed and I use this to change the volume of the whole
[02:48.840] patch and this LFO here outputting and sine wave form over 32 bars is used with this multiply
[03:00.000] and you can see now it's pretty loud so it's on the loudest part and then it slowly
[03:05.760] goes back to zero and then also the patch is quiet you don't hear anything.
[03:12.040] And this line hits zero also this year is outputting a triggering signal which triggering
[03:20.520] these dice modules and these dice modules generating values for the LFO speed information
[03:28.480] so every time the patch is quiet a new pair of values here is generated the LFO is generating
[03:34.720] a new wave form and this new wave form is a new melody and then the LFO here starts to
[03:41.800] fade in the patch again and we can hear a new melody occurring okay.
[03:48.680] So we have a patch that's slowly fading in and fading out and every time it's faded
[03:54.400] out we generate a new melody and then we start to fade in again.
[03:59.120] I use the same method as with the pitch information here I used for the rhythm so I used basically
[04:08.840] the phase information from the transport and mixed in an LFO sine wave form I blended these
[04:18.960] two together to have a much more complexer wave form and I used this wave form to drive
[04:25.400] basically this gates module here.
[04:31.440] So when I go here straight to only input 1 we have the original phase input from the transport
[04:39.280] going straight into the gates module here and have a pretty synchronized playback but
[04:45.080] this is also pretty static and pretty on grid and pretty boring.
[04:50.120] So I used an LFO here with some random speed and blended this in so you can see sometime
[04:57.120] it's playing faster sometimes it's playing slower and sometimes it's going even backwards.
[05:06.960] So the rhythm is also basically generated randomly on the fly to make it more interesting
[05:15.960] on the rhythm part so we have the same methodology for rhythm and pitch information here and we
[05:25.840] have this LFO controlling the whole patch with an overall volume.
[05:34.240] So this patch has one problem it's basically completely quiet after 32 bars that's the
[05:44.360] only problem with this patch but you can fix that pretty easily.
[05:49.640] You just go in here and hit control and D and now we have two of these patches and you
[05:57.280] can go to this patch or the second patch and basically offset this LFO which controls
[06:03.200] the sequence by 180 degrees and then you just hit start and stop to trigger these LFOs
[06:14.520] together or again and now when one patch is quiet the other patch is loud and vice versa
[06:22.680] and you have basically every 32 bars coming in in new melody and the other melody is fading
[06:29.080] out so you have some back and forth between these two patches which can be interesting.
[06:39.080] No it is interesting so this basically gives you a more kind of procedural approach to generating
[06:48.840] notes and rhythm instead of the random approach from my dice pitch step sequencer and it's
[06:58.120] also more interesting because you can more like go in here and tweak some of the parameters
[07:04.920] and you have more kind of predictable outcome in the end right you can see what's happening
[07:12.040] here with the Sine-Waves and how they are influencing each other and what you get here
[07:18.760] at the end right so this is maybe more interesting if you want to have more like natural melody
[07:27.000] lines in your crit patches so just blend in these LFOs and get more complex waveforms for
[07:35.320] the pitch input you can not only use the blend here you can exchange this maybe also for ring
[07:42.680] modulation or amplitude modulation so there's no boundary for your experimentation this is
[07:52.680] just and this was just my approach here to use blend modules you can also use some math modules
[07:59.560] to change the behavior of your resulting waveform so that's it for this video maybe I gave you
[08:10.360] some new inspiration for new projects and new patches maybe I gave you a solution for a problem
[08:16.440] you have let me know in the comments also leave a thumbs up if you liked the video subscribe to the
[08:22.360] channel and if you want to support my content then maybe think about the subscription on patreon
[08:29.400] or use my voucher code in the bitwix store it helps me a lot so thanks for watching and I see you
[08:35.560] in the next video bye