A Reaktor challenge of sorts

DFaught · 2022-02-05T15:10:12+00:00

There was an error rendering this rich post.

https://www.youtube.com/watch?v=uNz1XfVfJak

After watching Andrew Huang's video about generative ambient music above, I tried to build something similar with Reaktor Blocks using only free blocks from the User Library and the original blocks.

Only half knowing what I am doing with Blocks, I kept running into CPU problems before I got very far. I know there are a number of generative ensembles available in the User Library, but I would like to see what can reasonably be done to incorporate Andrew's specific ideas in a Reaktor ensemble, either using Blocks or a lower level. Anyone care to try?

Find more posts tagged with

Reaktor 6

Comments

Murat Kayi

Only half knowing what I am doing with Blocks

Why did you only half know what you are doing? I am asking, because maybe you should start from the ground up, create a subtractive synth first, try more basic sequencing and slowly branch out into generative patches?

running into CPU problems before I got very far

This really surprises me. I use toybox nano blocks (which granted are optmized for performance) on an ancient surface pro 3 with an i5 and 4 GB Ram and I have yet to crash it.

I would like to see what can reasonably be done

I know it's fun to watch other people explore stuff, but in this case I would say you should definitely get patching and try everything yourself. Especially if you felt inspired by Andrew's video. Start simple, get c&c, incorporate suggestions, fail, then fail better. Post here, the thread title fits. It's a challenge. Take it on.

DFaught

Okay, maybe it is more like 60% knowing. Here is a synth I built a few years ago - semi-generative.

https://www.youtube.com/watch?v=jxUdGL0rXLk

I run into problems when trying to copy some of the things that Andrew talks about.

Murat Kayi

Ok, based on that video I have a hint on how to get to a patch with more movement. The shift sequencer's main strength is that the lanes are flexible. Gates, Octaves, Pitch a.s.o. can have play ranges shorter than the full sequence. What you can do then, is to modulate the position and possibly length of the play range independently of any of the other lanes in the sequencer. Result: Octaves change through pitched melodies, loud passages become quiet, silent passages feature long notes, loud passages are staccato and all of that changes constantly.

You can use LFOs, mod sequencers...well basically any CV source to do that.

This alone generates so much movement that you can listen for several minutes without being able to recognize a pattern.

colB

https://community.native-instruments.com/discussion/791/a-reaktor-challenge-of-sorts

You can't really replicate Andrews specific ideas without having access to the specific modules he's using. You can reproduce a much more general approximation, but it won't sound the same. Even Andrews patch would sound pretty different with just a few parameter tweaks here and there.

Using only free stuff is a major limitation because even the Blocks Primes collection has a whole lot of really cool stuff that would help a lot... A huge part of the character of the sound he has there is down to the reverb delay fx. So even right there it's getting tricky with only free Blocks.

Why can't you use the Blocks Primes?...

EDIT: I misread your OP. So its all the factory blocks - primes included - plus free UL stuff?

colB

Here's a wee generative ambient rack with some kinda similar things using factory Blocks

gen01.zip

gen05.zip

DFaught

@colB I tried to use your gen05 rack, but it tells me that it can't find my Blocks Base library. That is properly installed as far as I can tell and I can pull blocks from it, so I don't know what is happening.

colB

https://community.native-instruments.com/discussion/comment/4579#Comment_4579

That patch I posted is a Rack that also uses Blocks primes. Maybe that's the problem - do you have Blocks Primes?

DFaught

https://community.native-instruments.com/discussion/comment/4675#Comment_4675

Yes, have that too. I have seen other posts abut the difficulty of dealing with paths for the NI products. I notice that in my installation Blocks Base and Blocks Primes are installed below C:\Users\Public\... but the other user paths are all below C:\Users\David\... so I'm wondering if that has something to do with it. Maybe you could post a screenshot of the rack?

colB

At this point, I wouldn't be worried about my dodgy patch (doesn't fit on one screen anyway) - more about getting Racks to load!

The user paths are different. Mine are also in my personal folder, while the factory Blocks are in the public documents folder.

Maybe Raise a support ticket somehow - @ one of the mods here anyway to try and get connected for some help with that. There were threads about this on the old forum, but maybe start a new thread here too...

Seqsual

Some points I got from Andrew's video:

bass is less frequent that mid range and plays the most limited selection of notes;
middle range notes are the main, most regulary clocked part;
high register notes/phrases are occasional;
all three "channels" (above) are in the same scale,
some "out-of-sync" sounds are used to spice it up (he has "Floaty arp", but IMO speech samples works very well for that kind of things);
all washed in spacey reverbs or similar efx;
and if you let it go for long time use some very slow modulation of the timbre.

I made a rather simple rack consisting of 4 channels with identical structure:

Rounds LFO used as clock source driving Clock Divider
some 3 divisions are summed together (Bento Box Mix) to form a complex clock/gate source
another Rounds LFO is feeding Quantizer (Level and Offset adjusted)
MIDI Out receives pitch from Quantizer and gate from the Mix.

Each of those structures is using separate MIDI channel to trigger different sound. I only used Massive X stock presets in the examples below. (I honestly don't feel like trying to rebuild Massive's synth structure 4 times in one Reaktor patch...)

https://youtu.be/z0oV2TDpe8s

Here is the rack if you are interested:

5x something.nksr.zip

Btw, I had the same problem when I tried to open colB's gen05 rack ("Reaktor can't find Blocks Base library..." message).

---

I used a similar concept to build a rack with my Blue Set, again - just MIDI on separate channels playing several Massive X instances, every version using same rack but different sound presets:

https://youtu.be/jm2LLbY6rwY

https://youtu.be/WmD3YbUXgnw

https://youtu.be/FhiZ95x7XPw

colB

What about this one - does it load?

gen08.zip

DFaught

@Seqsual very nice. If this were a competition, you're in the lead. I especially liked the last Blue one. But now I'm curious, are you using something like MIDI Yoke, Rewire to drive Massive X on the same system? And why do that instead of a simple synth inside Reaktor? I've seen something about DSPs requiring that only one CPU core be used, maybe related to that? Or you just like Massive X? And yes, your rack loads just fine.

@colB I get the same error when I try to load gen08, so no go.

So what does @Seqsual know about saving/loading racks that @colB and I don't?

Thank you both for your responses!

colB

Was gen01 broken as well, or did that one work?

It would be good to get to the bottom of this. Maybe there is something in my installation that is non standard in some way... a hang over from a previous version or something. Would be good to know.

DFaught

I just added 4 instances of the Monark Mini synth (from the user library) to replace the MIDI outs, added a 4-channel mixer and a Rounds Reverb to the rack that @Seqsual uploaded. It gets some pretty interesting sounds but could probably use more work. Reaktor tells me it's running around 80%, which is a lot higher than what this original rack runs. I am hesitant to upload it without knowing why that might not work.

colB

OK, lets try again. After hopefully solving the racks issue that was completely on my end, I've rebuilt the patch I made.

Basics are similar to Andrews patch (although totally different sound, and nowhere near as Pro ;-)).

There is a probabilistic melody/bass:

This uses the west coast sequencer, and drives it by sending probabilities to the offsets via modulation A and B. The probabilities are created using noise from a sample and hold (S&H), then multiplying it with itself with a VCA to generate a curve, VCA is set to exp, so there is an extra curve, so you get the white noise but to the 3rd power. That makes values close to zero way more common than values near to 1. So for the X axis, the notes to the left of the grid are more probably than the notes to the right.

I used trig/gate processor combined with another S&H to reduce the likelihood of a gate firing when the pitch is lower, so the low notes have longer pauses after them.

The arp is done using trig/gate/S&H to get a random firing similarly to the pauses/rests for the melody. It's notes are generated using a curve sequencer combining both channels, and quantising. The gates for this come from a square wave LFO. The rate of the LFO is modulated per burst with a sample and hold, and during the burst by the curve sequence to give a curved change in pace - with the settings tweaked this gives a nice organic effect. The timbre is modulated per burst as well using the same S&H as the pitch offset...

Finally, the 'granular' FX is hacked together using delay and verb, with some feedback, and the whole thing amplitude modulated by a slewed divided clock gate to get that nice hypnotic pulse effect.

Obviously, this could benefit from using some more specialised Blocks (e.g. ASR would be good to get some harmony in there), and also from better mixing/FX. But I wanted to keep it Factory only.

Also would be good to spend some time really focussing on note choices for the sequencer and the arp quantiser.

genPhoenix04.zip

I sure hope this one works :-p

Quick links

Unanswered questions