Forums > MaxMSP

Designing a Granular Synthesiser with grains generated via Cellular Automata


Ali
August 25, 2010 | 8:55 pm

Greetings all, this is my first post so please bear with me on the technical terms.

I am a doing my masters degree thesis on granular synthesis and cellular automata. I am trying to design a grnular synth and to use the data created by forms of cellular automata to control various parameters. Ideally the creation of the grains themselves but also grain time, duration, offset, stereo location of grains etc etc… whatever seems interesting and will produce interseting sounds.

I have a pretty good backgorund in the sound synthesis side of things but less so in the computing, coding, programming side but I have managed to work out Csound which seems a lot more taxing than Max so far (I’d be happy not to see Csound for a long time lol – but it is powerful). The idea for this project came after using Chaosynth

http://www.nyrsound.com/Chaosynth/CsynInformation.htm

I have started working through the tutorials and am making some decent progress. I have used the following patches as a reference on my computer (windows 7 running Max 5):
- Jit.conway
- mdegranular
(while certain objects/externals won’t run on my version of max I can download trail version on my or uni computers if/when needed)

Both have provided me with the means of trying to reverse engineer (if thats a real term) through the patches learning the objects as I go. I have only been using Max for about a month now so am far from fluent but I get the main concepts of which objects I need to understand but am struggling at understanding all the maths and expressions as easily.

At this point it would be extremely helpfull if anyone had any advice on which areas I should get into first such as:
- learning all relevant math related topics (which ones are most fundamantal?)
- focussing on a specific area (such as Matricies, Controlling data flows, MSP specific stuff etc)
- perhaps learning existing patches inside out first ( CA, Granular, etc)

The ‘Granular Cloud Generator’ is a good example of the granular synth i would try to emulate (partially because I have been told it overcomes some time rleated problems that a simple MAX patch will encounter); if I could generate Cellular Automata output data and feed this into a similar granular instrument that would be perfect.

Also I trying to work out how to make the output of a matrix audible in a way that I can map each cell to a specific frequency and amplitude ( so when the Cellular automata evolve over time so will the corresponding amp of freq)

Sorry for the long post guys, any advice would be much much appreciated.
(thanks for taking the time to read it all!)


August 25, 2010 | 10:24 pm

brain candy attached.


August 26, 2010 | 10:16 am

I did my thesis on composition with cellular automata. I suggest you read Shiff’s book on cellular automata, it is the most recent and quite accessible http://www.amazon.com/Cellular-Automata-Discrete-Mathematics-Optimization/dp/047016879X

Robin Foster designed a 1D-CA patch. I added an extraction of the data so I could see which cell was alive and if a cell was alive it would trigger a note. I am now on vacation so i do not have my external HD with me but I will mail you when I get home (monday). Meanwhile, download my thesis (in Portuguese) on my website http://www.samuelvanransbeeck.be Although it is not in English, you should check out the biography for sure.



Ali
August 26, 2010 | 11:38 am

Thanks Noob4Life that sounds great but how do I use .mxb files? Are they just an old file extension?

To Thinksamuel I shall get into the book you reccomend today. I’ve been reading Wolfram’s A new kind of Science so it would be great to have another text to get into. I look forward to trying out your addition to the 1D-CA patch and will look through your bibliography right now.

Thanks for the help guys, its nice to find a forum where people really do help, thanks again


August 26, 2010 | 2:24 pm

really great patch, noob. very inspiring. i already have some ideas for making things more complicated. hah.


August 26, 2010 | 4:55 pm

hey, glad you like it, guys. but it’s not my patch, it used to be attached here created originally by Michele Verità, someone should probably reattach it there… or maybe this thread can be its new home… whatevs:

http://cycling74.com/forums/topic.php?id=14178

and it’s a max version of an even older patch made in Reaktor as part of the Reaktor Nu-Skool collection created long time ago.

(@Ali, you can just drag .mxb files to max5, they’ll open up fine, but if you alter them, you’ll be prompted to save a new Max5 version… no biggie, it’ll work and then when you save as max5, you won’t have to ‘save as’ again after that… but ya, basically, it’s an old Max4 file extension… still works with Max5)

happy hunting!


August 27, 2010 | 12:30 pm

Had a lot of fun with the cellular matrix sequencer last night. Thanks for reposting it, Noob!



Ali
August 27, 2010 | 1:16 pm

got the .mxb files to work, it wouldn’t let me drag and drop them but now one works they all do,

good stuff! – its a great example of waht im trying to do, some of the sub-patch coding is pretty intense but its good to see it all as a working example.

Now I have to find a way to map the evolution of the CA to virtual oscillators allocated to grid coordinates, thanks for the help again guys


Viewing 8 posts - 1 through 8 (of 8 total)