Quick and Easy Generative Patch Tutorial

Tutorial | Oct 03, 2023

In this video, I demonstrate how to create a quick generative patch using Bitwig. Starting with a simple setup, I gradually make it more complex over time, showing easy steps for everyone to follow. By utilizing modules like the polygrid and counter, you can create unique melodies and rhythms, and further modulate them to add variation and interest to your patches.

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

Questions & Answers #

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

How do I create a generative patch in Bitwig? #

To create a generative patch in Bitwig, start with an empty project and use a polygrid. Choose a collection of nodes that you want to use, such as different pitches in a specific scale. Use a counter module to advance the playhead through the step sequencer, and experiment with modulation to create different melodies and rhythms.

How can I influence the melody and rhythm of the generative patch? #

You can influence the melody and rhythm of the generative patch by changing the counter size, using different modules to trigger the playhead, and modulating various parameters such as step size and mirror mode. By experimenting with modulation and different settings, you can create a wide range of melodic and rhythmic variations.

How can I add more complexity to the generative patch? #

To add more complexity to the generative patch, you can incorporate additional modules such as wavetable oscillators, filters, and effects like reverb and delay. Experiment with modulation to modulate different parameters and create evolving and dynamic sounds. By layering and stacking voices, using modulators, and adding effects, you can create a more intricate and unique generative patch.

How can I record the output of the generative patch in Bitwig? #

To record the output of the generative patch, create a new track in Bitwig and set the input to the track with the generative patch. Arm the new track for recording and create a new clip. Play the generative patch and it will be recorded into the clip, capturing all the notes, velocities, and other MIDI expressions.

Transcription #

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.

[00:00:00] Hey, folks, welcome back to another video.
[00:00:02] I want to show you in this video how
[00:00:04] to make a quick generative patch.
[00:00:06] I start simple, and then I make it more complex over time.
[00:00:10] And I try to make it quick and easy for everyone to follow.
[00:00:15] So we have here an empty project.
[00:00:18] Maybe I choose a slower VVM here, 85.
[00:00:22] And in here, we use a polygrid.
[00:00:23] And the first thing we need is some nodes.
[00:00:29] And you can create nodes in a lot of different ways.
[00:00:32] You saw me use the step mod in my recent videos a lot.
[00:00:38] But you also can take here the pitches module for that.
[00:00:42] And I want to use here the pitches module more like a collection
[00:00:48] of nodes you want to use.
[00:00:50] So you want to stay in A minor.
[00:00:52] So I'm using A probably a lot.
[00:00:54] And I want to use C maybe.
[00:00:56] And I want to use a D. I want to have F in there.
[00:00:59] Maybe let's say a G also.
[00:01:03] Maybe another A, maybe an E, something like this.
[00:01:07] Or maybe this E here.
[00:01:10] We can also use maybe more nodes, let's say 16 steps.
[00:01:15] And choose more different nodes here from the scale.
[00:01:19] So all the white nodes basically.
[00:01:23] D down here, maybe another A. Or maybe an A also here.
[00:01:27] So I choose here a lot of A's.
[00:01:28] So the probability that we get an A is pretty high.
[00:01:33] So that's a thing behind it.
[00:01:35] With disable here, the phase pre-code.
[00:01:38] So the playhead stops playing.
[00:01:40] And you want to introduce yourself modules to advance
[00:01:44] through the step sequencer.
[00:01:46] And some question I saw today on Reddit
[00:01:48] was how do I can use triggers or gate signals to advance
[00:01:53] basically this playhead.
[00:01:55] We can do this by using a counter.
[00:02:03] Go with triggers into the counter.
[00:02:05] The counter counts up.
[00:02:06] And then it spits out here a phase signal that steps
[00:02:11] through this pictures module here.
[00:02:15] And you can see it skips certain nodes, right?
[00:02:18] Every second node, every other node, it skips.
[00:02:21] And that's because the step count of this module here is 16.
[00:02:25] And the counter only has eight.
[00:02:28] So we need to increase this to 16.
[00:02:31] And it takes up here every single step.
[00:02:36] So this already gives you an idea that you
[00:02:38] can influence here the counter and influence how
[00:02:42] the playhead here advances through this module.
[00:02:47] So if you change this on modulators, later on,
[00:02:51] you can influence how the melody sounds.
[00:02:54] So we can use here a wave table real quick.
[00:02:58] Waste table oscillator uses a pitch input.
[00:03:01] We don't need a bridge pre-chord.
[00:03:02] We need to increase here the amount.
[00:03:06] Use a simple envelope, AD envelope,
[00:03:10] and just an output here.
[00:03:12] And then we use here maybe convolution reverb.
[00:03:23] Use a delay plus, a reverb on there.
[00:03:28] We maybe use a chorus plus, maybe in between here.
[00:03:34] Delay two, lots of nice things.
[00:03:42] And then, yeah, use also trigger output here
[00:03:46] to trigger the AD.
[00:03:51] So this is a melody, right?
[00:03:57] It chooses always the same notes out of this box,
[00:04:02] out of this module here.
[00:04:04] But if you change here the counter size,
[00:04:11] it plays a different melody because it
[00:04:13] chooses different notes.
[00:04:18] So just already with this small little patch here,
[00:04:34] you can do a lot of things because you can modulate this
[00:04:37] inside of Bitwig.
[00:04:37] You can also modulate it to trigger size
[00:04:41] so you get a different--
[00:04:45] --a different rhythm, different rhythm, different melody.
[00:04:52] OK, so that's an idea.
[00:04:58] You can also take here the output of this counter
[00:05:01] with just a face signal.
[00:05:03] And you can use this to influence basically
[00:05:05] the index of the wave table.
[00:05:12] So now, the problem is every note that you play in here
[00:05:26] gets the same index of the wave table.
[00:05:30] So if you always play this note, you
[00:05:32] get the same index for this note inside of this wave table,
[00:05:34] all right?
[00:05:36] So that's maybe not that interesting.
[00:05:41] So what you can do is you can just make it simple.
[00:05:45] Use a long delay here.
[00:05:48] Maybe this size is a bit--
[00:05:51] and just delay the signal by maybe, let's say, four, eight
[00:05:56] notes, and then go into this wave table here.
[00:06:05] So now it's basically the same as before,
[00:06:06] just to use a different note, or one note that's exactly
[00:06:11] four, eight notes earlier.
[00:06:15] And you use this note then for the wave table position,
[00:06:18] or the face signal at this point in time
[00:06:21] for the position of the index, right?
[00:06:25] That's also always the same thing.
[00:06:28] But when you modulate your the counter
[00:06:30] and you play a different melody each time,
[00:06:32] you basically delay--
[00:06:34] yeah, you delay the notes when you play the notes
[00:06:37] and when you play a different position in the wave table.
[00:06:40] So it gets you always a different note
[00:06:43] for a different wave table position.
[00:06:46] I hope this makes sense.
[00:06:48] You have to think in time.
[00:06:50] It's a bit of time travel.
[00:06:52] You have to make in your head.
[00:07:20] So when you modulate basically these two values
[00:07:23] continuously, differently, maybe with a randomized button
[00:07:27] or something like this, you always
[00:07:30] get a different position for a different note,
[00:07:33] which makes it more interesting, in my opinion.
[00:07:37] But before we do this here, we can also influence, of course,
[00:07:42] maybe with some face modules here,
[00:07:46] we can influence how we play this thing here, this step--
[00:07:52] yeah, step sequencer.
[00:07:54] So let's go here to 16 notes.
[00:07:57] Let's increase here the speed.
[00:07:59] So you can see we step here through all the steps,
[00:08:03] continuously from start to finish.
[00:08:05] But we can also use here reverse.
[00:08:08] So now we are playing basically this whole sequence
[00:08:12] in reverse.
[00:08:13] So you can see from the end to the start.
[00:08:17] I mean, when we remove this, it's the other way around.
[00:08:24] You can also say, let's use a mirror.
[00:08:31] Right, so we get the step signal here.
[00:08:39] And the signal is then folded, in a way.
[00:08:45] So we play back and forth all the time.
[00:08:53] OK, so then we take here an octaver.
[00:09:07] So you can influence the octave.
[00:09:10] Maybe we start at minus 2.
[00:09:14] There's some kind of bass for that.
[00:09:17] But then I want to introduce you some voice stacking.
[00:09:20] So I select the device, go to maybe three voices.
[00:09:26] So now we play this whole patch here three times
[00:09:30] at the same time.
[00:09:30] You can hear it's already a bit phasing.
[00:09:35] And we use a voice stack modulator.
[00:09:39] And we have three voices, of course.
[00:09:41] And when we use this modulator here,
[00:09:43] we have a value from 0 to 1.
[00:09:46] So because we have three voices, we modulate this here
[00:09:49] by exactly three.
[00:09:55] Which means every octave, or we play basically three voices
[00:10:03] at one octave.
[00:10:06] So the first octave is minus 2.
[00:10:09] The second voice is minus 1.
[00:10:11] And the third voice plays at the octave 0.
[00:10:16] OK, so then we can say the first voice plays 3.
[00:10:24] The second plays at a different rhythm.
[00:10:28] Let's go to 5.
[00:10:31] And the third voice plays at maybe 7.
[00:10:45] It's a nice, small, little pulley rhythm.
[00:10:51] Maybe I need a filter in here.
[00:11:05] Let's go for a selling key filter.
[00:11:09] So we disable also the pitch pre-chord
[00:11:12] and pull this down to C3.
[00:11:14] Just double click here, right?
[00:11:15] And then we use a bandpass filter, increase the resonance.
[00:11:22] And then we want to do the same thing here with that long delay.
[00:11:25] But instead of using the phase signal,
[00:11:27] we use the pitch signal here and go into the selling key
[00:11:34] filter, maybe delay it even more, and change here
[00:11:39] the frequency, the center frequency of this bandpass filter.
[00:11:43] But now because we have this delay,
[00:11:45] we don't use the current note as a pitch input
[00:11:49] for the frequency here.
[00:11:50] We use always the note that is 8 notes earlier.
[00:11:58] And because we change the melody all the time here,
[00:12:02] we get always a different note that currently plays
[00:12:04] in a different pitch from the note earlier, if this makes
[00:12:08] sense.
[00:12:09] Of course, you have to change here this all the time
[00:12:11] to actually make it different each time.
[00:12:14] But you get the idea.
[00:12:22] Maybe we ordered this here a bit more because the polyrhythm
[00:12:45] is nice.
[00:12:46] But it's a bit too busy.
[00:12:50] So we can use a clock quantizer here and trigger--
[00:13:00] or use a trigger here as an clock source.
[00:13:34] And when you want to change the melody,
[00:13:36] you can just change the counter size here.
[00:13:38] And you get different selection of notes.
[00:13:43] Or maybe instead of C3, let's go to C5.
[00:13:54] So we amplify certain upper harmonics.
[00:14:03] So you lose a bit of tonality because we kind of remove
[00:14:09] the fundamental frequency of the sound.
[00:14:13] But we can maybe use a sub oscillator here.
[00:14:18] Maybe put this over here.
[00:14:20] And use a sub oscillator playing on the same octave
[00:14:25] as the sine.
[00:14:28] Use a mixer and bring this back.
[00:14:35] So this is here C5, maybe C6.
[00:14:46] So C3 always uses the center frequency, basically.
[00:14:54] But you can just fake that you want to shift at one octave
[00:14:58] higher, but instead of using C3, you use C4, C5, or C6.
[00:15:05] In a way, we can use an octave over here
[00:15:14] instead of this track here.
[00:15:22] Or maybe using two.
[00:15:30] Let's use here maybe a step mode for that.
[00:15:46] Nice uneven number, maybe nine steps.
[00:15:52] And modulate this here.
[00:15:57] Yeah, and because we amplify a certain overtones,
[00:16:13] what's better to bring in some different overtones
[00:16:16] than using maybe a transfer and distort the whole signal here?
[00:16:23] And if it becomes too dissonant, you just
[00:16:40] mix in the sub oscillator, which basically nails the fundamental
[00:16:47] and gives the ear something to orientate itself
[00:16:53] in which kind of pitch we are, right?
[00:17:08] We can also, let's say, is there something you can use here
[00:17:19] to--
[00:17:23] Yeah, we need to modulate this here in certain ways.
[00:17:34] Maybe use a different step mode here.
[00:17:38] Or I need to explain.
[00:17:42] Eight notes.
[00:17:48] So we use this here to maybe change up the count size or the mirror.
[00:18:01] Maybe a bit slower.
[00:18:15] And maybe a different one, different step size, maybe seven.
[00:18:29] And for triggers here for the rhythm.
[00:18:37] OK, kind of like it.
[00:19:30] Yeah, that's basically how you create these kind of patches.
[00:19:35] I started simple with some basic ideas.
[00:19:39] And then when you advance this patch even more,
[00:19:42] you get more and more ideas how you can influence this patch with modulators.
[00:19:46] And the more modules you have in here, of course,
[00:19:49] the more modulation targets you have.
[00:19:52] And the only thing you need to avoid or try to avoid
[00:19:56] is that it becomes too random.
[00:19:58] Because when it's too random, then all of this becomes obsolete
[00:20:03] and you can just take a random modulator here
[00:20:06] and spit out some random notes, right?
[00:20:09] And modulate everything at random,
[00:20:12] and you have basically more or less the same outcome.
[00:20:14] So there's a certain sweet spot between it's too random
[00:20:19] and it's too monotone or too, yeah, you know, too straight.
[00:20:25] So there's always a sweet spot in between
[00:20:28] and you need to find the sweet spot, in my opinion,
[00:20:31] to make it lessened.
[00:20:36] And something I get asked all the time is,
[00:20:48] yes, you can record the notes of this here.
[00:20:51] Just take a note out, for instance, right?
[00:20:56] You take here to trigger out, you take the notes,
[00:21:00] pitch output and all the other stuff you want to use,
[00:21:04] maybe here velocity or the MIDI channel or pressure
[00:21:07] or some other MPE stuff, right?
[00:21:10] You get the notes here.
[00:21:12] Then you create another instrument track.
[00:21:14] And as an input, you go down to tracks.
[00:21:17] You can see that the polygrid is already flashing nicely.
[00:21:20] It's record and then just create a new clip here
[00:21:25] and record everything that you get from the grid.
[00:21:29] And then you can use it in kinds of different ways.
[00:21:34] Oh, you can see we have already some MPEs
[00:21:36] and micro pitches in there,
[00:21:38] because we don't use actually here a sample and hold.
[00:21:45] So we can do this.
[00:21:48] Actually, we have to use here a sample note here,
[00:21:54] trigger here, trigger there, bam.
[00:21:59] Then we get the same thing,
[00:22:03] but without all the micro expressions.
[00:22:07] OK, that's it for this video.
[00:22:09] Thanks for watching.
[00:22:10] Please leave a like if you liked the video.
[00:22:11] Subscribe to the channel.
[00:22:12] If you have some questions, please leave it down in the comments.
[00:22:15] See you in the next video.
[00:22:17] Bye.