[less concepts]: max for live

[nnn]

one more! I let olafur play the part with less notes thanks @dan_derks

[chailight]

thanks @dan_derks - this is an amazing tool. currently building a new piece around this.

I always like including an visual element - I had a look at the max patch in case there was good starting point for visualising what is happening under the hood with the cellular automata, but wasn’t obvious what to really latch onto.

Maybe I could just feed the seed out to a separate CA visualiser, it might not be the same as what is generating specific notes, but if the same rule were being used, there might be some relationship.

Has anyone tried anything similar???

[dan_derks]

oh cool, i’m so glad to hear that the device has made it into your work

@chailight

I had a look at the max patch in case there was good starting point for visualising what is happening under the hood with the cellular automata, but wasn’t obvious what to really latch onto.

I’m tuning up an abstraction of the CA engine – essentially, it’ll output raw values 0-255 and you can scale that to your needs. you can also have a ton of them running, modulating various parameters of a visual patch.

recently, a friend mentioned that they were interested in grabbing four of the bits, summing, and averaging their output to use for control – that’d be super easy to roll on your own once I get this patcher clean

excited where you end up!

(also, paging @zebra for any thoughts on CA visualization, as his brain is a damn treasure trove)

[chailight]

thanks @dan_derks - will look forward to playing with that - happy to beta test stuff as appropriate.

[zebra]

visualizing elementary binary CA seems pretty straightforward, the state is just a 1-d binary array… dunno if there’s much to say about that. a scrolling list of cell states is all there is to it.

let’s see… i banged out this JS version a while back for some reason… most expedient reference i’ve got. with the UI you can randomize or set patterns, and see which neighbor-sets correspond to which rule-bits. (you can also change boundary point and conditions, but i didn’t add UI glue for those.)
https://jsfiddle.net/ebuchla/nj6bfkha/

(i also have an audio-rate elca~ object i should package up, if that is of interest. it’s nice to listen to several simultaneous cell states with interpolation by cosine segment.)

@dan_derks - i have max but not ableton, so i’m not sure how to look at this. i remember thinking that the norns version of less concepts could maybe have benefited from using an 8-cell window into a larger state array. (may more possible behaviors that way.) boundary mode also makes a big difference.

IOW:

@chailight

it might not be the same as what is generating specific notes

yeah you’d want to pay close attention to the state vector size and whether bounds are clamped high, clamped low, or wrapped.

[chailight]

thanks @zebra - great visualisation example - i’ll definitely have a play with the ideas there. it got me thinking about somehow visually revealing both the “background” pattern of the CA as well as the specific musical events that are selected from that background - perhaps just by using colour on a cell when it happens to be the one generating a note.

I’d also be really interested in the audio rate object. Am happy to play with experimental max objects, so no need to put any major effort into packaging. I’m as equally interested in driving sound from "visual"processes as I am in going the other way.

The background to my current focus on is working on a piece for the SeenSound Audio/Visual series I help curate - I’ve naturally been thinking about all that is going on, and playing with processes and events that have cascading effects seems very on point in the current context.

[zebra]

getting deeper into (non-elementary) CA can be really rewarding…

at root, a CA model is just one kind of computational representation of a system of differential equations. they can have continuous and stochastic rules as well as bitwise rules.

let’s see… here’s an example of something i coded up ages ago for a real-world modelling problem (difussion of time-release tablets in bloodstream)
{ https://github.com/catfact/celldiff/blob/master/CellModel.cpp }
(includes some stochasticity)

and here’s something that i found rather aesthecially satisfying, but never developed much beyond a single performance installation: replacing binary cell value with continuous value, and bitwise update rule with a continuous weighting. this really is “just” a simple linear dynamic system. but it’s a nice connection
{ https://github.com/catfact/catfact_sc/blob/master/catfact_util/CfModels.sc#L110 }

(wow that is both messy/throwaway stuff and quite unrelated to the thread. woops!)

---

i’ve been (very) slowly working my way towards releasing a collection of chaotic oscillators for max/msp. will be nice to share here, and i’ll do so sooner rather than later.

first, gotta get the next norns update out! it’s a big one

[dan_derks]

@zebra

i remember thinking that the norns version of less concepts could maybe have benefited from using an 8-cell window into a larger state array. (may more possible behaviors that way.) boundary mode also makes a big difference.

for sure, i didn’t mean for it to come off that i’d ignored the suggestion — i just didn’t have a clear understanding of how to implement it. your examples are illustrative, thank you! maybe a CA thread would be good? i want to dig into this but don’t really want to shackle you to this topic.

[nervous_circuits]

Just wanted to say thanks for making and sharing this, proved to be a really fun and interesting M4L tool

[eblomquist]

Greetings Dan,

I’m finally getting a chance to dip my toe into Less Concepts, and wow!

Already some lovely stuff happening, as I start to explore the randomization function in the meta sequencers… I’m running it into Lekko and it’s gorgeous!

One question about the scale variable, could it be made randomizable, and if so, how? I tried to send it numbers, thinking that perhaps each scale option corresponded to its position in the list, but that didn’t do anything…

I’m still pretty clueless yet when it comes to Max, so I don’t know how it parses values in a text list…

Grateful for any clues you might share!

Ed

[dan_derks]

hey ed!

oh, fantastic to know you’re digging in

@eblomquist

One question about the scale variable, could it be made randomizable

you can definitely address the scales in the meta sequencer – scale: aeolian, for example.

you could also just MIDI map Live’s LFO device to the scale dropdown with a random waveshape. this got me pretty great random “church mode” results:

hope that helps!

[eblomquist]

Thanks Dan.

I had figured out shifting scales in the meta sequencer, but was hoping for a way to add random. The random wave shape is a great idea! I will try that next…

Very very fun, thanks so much!

[Blit]

I really enjoy using less concepts for generating ideas or just jamming. Thanks for making it into a M4L device
Im using it here to play a microfreak

[unicity]

Oof I missed this one until now. Thanks so much for sharing this! Loving the meta seq!

[merchant]

This is so good! Thanks for building this.

[rbxbx]

Decided to play around with using less concepts to drive some percussion this evening. Kind of mixed results, but going to keep refining the idea. Heavy use of the meta sequencers can make for some pretty interesting semi-evolving patterns!.

Anyone else exploring this idea?

[nkozi]

Can’t believe that was only 5 minutes - I’m not usually one for jams but you really put lc to great use!

[Blit]

Thanks! That’s the kindest words I ever had for one of my jams. : )

[bestia]

Absolutely love this as a generative sequencer for leads and melodies, but has anyone found a way to create long chords and more pad oriented sounds?

Would really love to use this as an ambient foundation then improvise over it. Any ideas?

[closeyoureyes]

When I first tried less concepts I mostly sequenced piano libraries and other sample-based VSTs, but lately I discovered how beautiful it may sound with hardware synth. Here I sequence Behringer Deepmind with less concepts and play couple of acoustic instruments on top of it. Digitakt does subtle drums using amazing probability function of the recent firmware which turns basic kick/tom pattern into never repeating itself figure.
Big thanks for creating less concepts and sharing it with the world!

Less | Ambient Electro-Acoustic Generative Set