My Ideas for generative Music...

Cwejman, Livewire, TipTop Audio, Doepfer etc... Get your euro on!

Moderators: Kent, Joe., analogdigital, infradead, lisa, parasitk, plord

Post Reply
User avatar
ChibaCityBlues
Learning to Wiggle
Posts: 10
Joined: Fri Dec 30, 2016 1:42 pm
Location: Hamburg

My Ideas for generative Music...

Post by ChibaCityBlues » Thu Feb 09, 2017 5:13 am

I'm absolutely new to synthesizers. But over the last few years I acquired a taste for Music for Airports. I got into Brian Eno, Alva Noto, Bohren, Monolake and Murcof.

During my Christmas vacation I found out about modular synthesizers and generative music and fell in love with it. So I bought a very basic Setup and a Beatstep Pro. (modulargrid.net)

And obviously I want more.

So I started to watch videos and read threads on generative music to collect ideas on what modules might be useful on Modular Grid. (modulargrid.net)

But since I'm just learning this stuff, I want to go slow. Nevertheless I want to go in the right direction. So I made up a long term plan. (modulargrid.net)

What are your thoughts? Is the balance of the system okay? Am I missing something? Are there other problems with the rig? A important or interesting module for generative Music I did not include in the collection?
You do not have the required permissions to view the files attached to this post.

User avatar
Daisuk
Super Deluxe Wiggler
Posts: 3781
Joined: Tue Jul 23, 2013 3:16 pm
Location: Oslo
Contact:

Post by Daisuk » Thu Feb 09, 2017 6:49 am

Looks like a very good setup to me, looks like you've done your research. I don't see any switches though, which can be very handy for generative music. Look into the WMD Sequential Switch Matrix, Macro Machines Dynamic Destiny or the Doepfer A-152, for instance. Also, comparators and logic modules could be of interested to you - Ladik has a module called "derivator" which is mighty useful.

Although it's a DIY module, I've found that Ornaments + Crime can do a lot of good things for generative music (it has many interesting modes, and especially the Lorentz Attractor in it can do a lot of good things).

I did a semi-generative experiment with it a couple of months back - don't mean to be self-promoting in your thread, but maybe it could be of interest to you. Everything in this patch is basically based on the Lorentz Attractor app in Ornaments + Crime which is going through a comparator which then spreads out throughout the entire system. This one is still quite contained, of course, but you can really go wild with it and direct the CV it outputs anywhere for maximum effect.

[video][/video]

Good luck on the journey! Making generative stuff is quite challenging, but very engaging and a lot of fun. :)

User avatar
mlaszews
Common Wiggler
Posts: 126
Joined: Mon Dec 19, 2016 2:16 pm
Location: Boulder, CO

Post by mlaszews » Fri Feb 10, 2017 2:59 pm

I would ditch that mixer and headphone amp, and just use a cheap behringer mixer or something. Or replace with a voltage controlled mixer, which would be useful for generative stuff.

User avatar
Whelm
Wiggling with Experience
Posts: 301
Joined: Fri Sep 11, 2015 7:27 pm
Location: West Coast Ish

Post by Whelm » Fri Feb 10, 2017 5:55 pm

Hi! I'm also building what seems to want to become a generative rack. Your plan looks pretty well thought out to me. I would second the switch idea, and logic would be cool for generative stuff if you can find the space for it.

I had a headphone out, then I got rid of it for a Behringer mixer. Now I have one again. :despair: Whether it's worth the space will depend on your workflow/mindset. I have to take my stuff out and put it away every time I use it, and that extra step of setting up the little mixer turned out to be a drag for me. Then again, I have a 1U row for such things.

User avatar
Umcorps
Will patch for food
Posts: 2321
Joined: Fri Apr 17, 2009 6:33 am
Location: Currently unrecognisable

Post by Umcorps » Sat Feb 11, 2017 3:52 am

Can I suggest a different approach? Rather than leading with the gear maybe think about what the principle components of generative music might be and let that lead you to a rack plan.

What follows next is a clip from a mail discussion I had a few months back about generative music in Euro.

....

First off I think there is a split between tonal/melodic generative and the more Krell like stuff. Krell is fairly easy to create in the sense that there aren't that many rules involved. But generative music that's conventionally tonal is a bit of a challenge.

I've spent a lot of time working with generative music in software and in my version of that world there are four foundation rules to making tonal generative - scale rule (which notes are allowed to be played) next note rule (what intervals between adjacent notes are available), harmonic rule (which intervals between notes are considered harmonically ok) and rhythm rule (which note duration values can we play with). For ambient the rhythm rule mutates into more fuzzy definitions of how long notes and rests can last and is based on real time rather than metric division.

In Euro, scale rule = quantisation. rhythm rule = dynamic clock division, next note rule = dynamic pitch CV attenuation and harmony rule = you're buggered. Because I've not yet found a module smart enough to work out harmonisations on the fly. [Since writing that I've picked up an Ornament & Crime which doesn't do realtime harmonisation but can be set up to sound as if it is doing that]


For quantisers my view is its essential that they have a trigger out option. There are plenty of choices. The Doepfer dual is really nice. Even Disting is sweet. I thought the ADDAC would be the one for me but it wouldn't play nice in my old case so I stick with Doepfer, Disting and the TB Quantimator for nice chordal stuff.


Next note decisions makes the difference between something sounding totally random and something that might sound like it was composed. Melodic phrases tend not to jump around big ranges all the time. You usually get a bunch of notes that are more clustered together then maybe punctuated by a jump up or down. So rather than just sticking a fixed range random source into a quantiser it can be a lot more interesting to modulate that range in some way.

ADDAC complex random is a good solution to this because you can directly modulate both the lower and upper limit of the random voltages. With less flexible random sources one option could be to use an offset voltage generator to give you your lower limit and mix that with a stepped random CV via a VCA. Modulate the VCA to define the range above the fixed voltage. Doepfer copy of the Buchla Source of Uncertainty (149-1) is very good if you have the room because it seems to do this by default on the 2N output.

Loads of options for the rhythmic stuff. RCD, SCM, Pamela's etc etc. Most useful of all is a logic module, particularly AND logic because you can use that to bring the more unruly random modules (hi wogglebug) into line by ANDing it with the note division of choice.

Ambient generative is good fun because you can be a lot more creative about how you generate note durations. A pulse wave LFO becomes a really interesting gate source if you gently modulate both its frequency and width with random values. Multi it to the clock input of a S/H unit being fed noise and use that output to seed the LFOs freq and width modulation with a new value on each gate.

If you use something like that to clock a shift register type thing like the Turing Machine or the Disting v3 quantised ASR you can get some really good things going for very little effort.

For more regularised, recognisable but surprising kind of thing, sequential switches are very useful especially for rhythmic stuff. Send a standard, preprogrammed sequence to one input and a randomly generated one into the second. Then use a beat synced random gate to step back and forth between then. Even better to leave the 3rd input empty but include it as a step. That way you introduce rests which are always useful in generative as it gets a bit relentless if you don't add space.

....

User avatar
mopoco
Common Wiggler
Posts: 123
Joined: Sat Apr 18, 2015 4:47 am

Post by mopoco » Sat Feb 11, 2017 7:44 am

especially for "generative" music i'd like to recommend the ultra random analog from ssf


for long term plans i'd like to encourage you to think about a sequentix cirklon with cv i/o...

User avatar
oberdada
Common Wiggler
Posts: 202
Joined: Tue Nov 01, 2016 12:06 pm

Post by oberdada » Sat Feb 11, 2017 10:38 am

Umcorps wrote:Can I suggest a different approach? Rather than leading with the gear maybe think about what the principle components of generative music might be and let that lead you to a rack plan.
....

First off I think there is a split between tonal/melodic generative and the more Krell like stuff. Krell is fairly easy to create in the sense that there aren't that many rules involved. But generative music that's conventionally tonal is a bit of a challenge.
....
Lots of good ideas there, and I second that it's wise to begin thinking about what kind of music one wants to make, and then look for modules or other means to get there.

If one really wants full flexibility and control I'd recommend programming, maybe in MAX/MSP or such, and sending control data to the modular which takes care of sound generation and effects. Or maybe use some programmable module.

It's also worth pointing out that generative music, such as Eno's Music for Airports, don't always need to be generated in one go. You can do that with multitracking and make a piece from several recordings.

User avatar
propertyof
Wiggling with Experience
Posts: 312
Joined: Thu Dec 01, 2016 7:58 pm

Post by propertyof » Sat Feb 11, 2017 10:59 am

you might've seen this tips from mylarmelodies, but here i post again

[video][/video]

peripatitis

Post by peripatitis » Sat Feb 11, 2017 11:03 am

Of course you could also go with something like teletype and take it from there.

User avatar
Dcramer
Super Deluxe Wiggler
Posts: 5388
Joined: Fri Mar 29, 2013 12:11 pm
Location: Canada

Post by Dcramer » Sun Feb 12, 2017 11:48 am

oberdada wrote:
Umcorps wrote:Can I suggest a different approach? Rather than leading with the gear maybe think about what the principle components of generative music might be and let that lead you to a rack plan.
....

First off I think there is a split between tonal/melodic generative and the more Krell like stuff. Krell is fairly easy to create in the sense that there aren't that many rules involved. But generative music that's conventionally tonal is a bit of a challenge.
....
Lots of good ideas there, and I second that it's wise to begin thinking about what kind of music one wants to make, and then look for modules or other means to get there.

If one really wants full flexibility and control I'd recommend programming, maybe in MAX/MSP or such, and sending control data to the modular which takes care of sound generation and effects. Or maybe use some programmable module.

It's also worth pointing out that generative music, such as Eno's Music for Airports, don't always need to be generated in one go. You can do that with multitracking and make a piece from several recordings.
Yes, great ideas here about translating aspects of the music into patches. :tu:
In my own works I do this by creating sub-patches called Engines that work together to generate all the aspects of the music I desire. Lots of preplanning and testing goes into it.
Here's an example using the A-152 and many others to create an 8 voice fully generative, large scale piece in which I use a programable quantizer to switch through several different scales of limited transposition:

[video][/video]

User avatar
BenasPlentas
Learning to Wiggle
Posts: 36
Joined: Tue Nov 08, 2016 5:45 pm
Location: Vilnius, Lithuania

Post by BenasPlentas » Wed May 31, 2017 7:18 am

Here's an example using the A-152 and many others to create an 8 voice fully generative, large scale piece in which I use a programable quantizer to switch through several different scales of limited transposition:
Amazing! And the schematics of a patch are very useful to get the idea. So much to learn here :)[/quote]
--
Benas Plentas | Lithowiggler | SoundCloud

User avatar
damase
Super Deluxe Wiggler
Posts: 1613
Joined: Sun Apr 19, 2015 11:33 am
Location: Texas

Post by damase » Wed May 31, 2017 8:48 am

Umcorps wrote:Can I suggest a different approach? Rather than leading with the gear maybe think about what the principle components of generative music might be and let that lead you to a rack plan.

What follows next is a clip from a mail discussion I had a few months back about generative music in Euro.

....

First off I think there is a split between tonal/melodic generative and the more Krell like stuff. Krell is fairly easy to create in the sense that there aren't that many rules involved. But generative music that's conventionally tonal is a bit of a challenge.

I've spent a lot of time working with generative music in software and in my version of that world there are four foundation rules to making tonal generative - scale rule (which notes are allowed to be played) next note rule (what intervals between adjacent notes are available), harmonic rule (which intervals between notes are considered harmonically ok) and rhythm rule (which note duration values can we play with). For ambient the rhythm rule mutates into more fuzzy definitions of how long notes and rests can last and is based on real time rather than metric division.

In Euro, scale rule = quantisation. rhythm rule = dynamic clock division, next note rule = dynamic pitch CV attenuation and harmony rule = you're buggered. Because I've not yet found a module smart enough to work out harmonisations on the fly. [Since writing that I've picked up an Ornament & Crime which doesn't do realtime harmonisation but can be set up to sound as if it is doing that]


For quantisers my view is its essential that they have a trigger out option. There are plenty of choices. The Doepfer dual is really nice. Even Disting is sweet. I thought the ADDAC would be the one for me but it wouldn't play nice in my old case so I stick with Doepfer, Disting and the TB Quantimator for nice chordal stuff.


Next note decisions makes the difference between something sounding totally random and something that might sound like it was composed. Melodic phrases tend not to jump around big ranges all the time. You usually get a bunch of notes that are more clustered together then maybe punctuated by a jump up or down. So rather than just sticking a fixed range random source into a quantiser it can be a lot more interesting to modulate that range in some way.

ADDAC complex random is a good solution to this because you can directly modulate both the lower and upper limit of the random voltages. With less flexible random sources one option could be to use an offset voltage generator to give you your lower limit and mix that with a stepped random CV via a VCA. Modulate the VCA to define the range above the fixed voltage. Doepfer copy of the Buchla Source of Uncertainty (149-1) is very good if you have the room because it seems to do this by default on the 2N output.

Loads of options for the rhythmic stuff. RCD, SCM, Pamela's etc etc. Most useful of all is a logic module, particularly AND logic because you can use that to bring the more unruly random modules (hi wogglebug) into line by ANDing it with the note division of choice.

Ambient generative is good fun because you can be a lot more creative about how you generate note durations. A pulse wave LFO becomes a really interesting gate source if you gently modulate both its frequency and width with random values. Multi it to the clock input of a S/H unit being fed noise and use that output to seed the LFOs freq and width modulation with a new value on each gate.

If you use something like that to clock a shift register type thing like the Turing Machine or the Disting v3 quantised ASR you can get some really good things going for very little effort.

For more regularised, recognisable but surprising kind of thing, sequential switches are very useful especially for rhythmic stuff. Send a standard, preprogrammed sequence to one input and a randomly generated one into the second. Then use a beat synced random gate to step back and forth between then. Even better to leave the 3rd input empty but include it as a step. That way you introduce rests which are always useful in generative as it gets a bit relentless if you don't add space.

....
I found this post so incredibly informing and inspiring. Thank you

User avatar
dumbledog
Super Deluxe Wiggler
Posts: 1165
Joined: Sun Aug 16, 2015 10:10 pm

Post by dumbledog » Thu Jun 01, 2017 10:19 am

Umcorps wrote:What follows next is a clip from a mail discussion I had a few months back about generative music in Euro.
Just wanna give props to what may be the best post I've read on here. This was super insightful, thank you :)

I love, love, love making tonal generative music things in Eurorack. Last weekend's jam was a Dark Time sequencer's CV being added to a Turing Machine that was going at a much slower rate, and running the sum through a quantizer, so that the sequencer's melody changed up and down the scale. The turing machine was also quantized alone providing a bass. Good stuff. I have an Arpitecht coming today which should do something very similar.

User avatar
gis_sweden
Veteran Wiggler
Posts: 595
Joined: Mon Dec 07, 2015 12:57 am
Location: Sweden
Contact:

Post by gis_sweden » Sat Nov 11, 2017 12:56 pm

Time for a new post! Quiet here :help: :despair:
Okay, you can use square wave LFOs to create a sequence. But you can also use gates. I use Doepfer A-160/A-161. Easier. Just connect, for an example, three gates to a mixer. From mixer to quantizer (if you want). Remaining gates and some logic get you some strange music.
Here is a simple example. :tu:
https://soundcloud.com/gis_sweden/pinhole-camera

User avatar
Orange
Wiggling with Experience
Posts: 366
Joined: Thu Jan 28, 2016 3:57 pm
Location: The Hague - Holland

Post by Orange » Sun Nov 12, 2017 4:45 am

damase wrote:
Umcorps wrote:Can I suggest a different approach?.......
....
I found this post so incredibly informing and inspiring. Thank you
Thank you indeed! :cloud:

Post Reply

Return to “Eurorack Modules”