Max/msp theory

[anarchosyn]

[anarcho note: I'm really bloody sleep deprived, so bear with me if I start ranting ]

Lastnight I swung out to catch Hrvatski do a live PA at the BlackBird here in portland, and I found myself bitten by the Max/msp bug once again. Sadly, I langish in PC land and my only experience with computer music has been through cubase vst systems of note arrangment (generally all hand programed notes, I'm getting into loop manipulation now that I reside digitally though). Before I continue, let me give some audio examples of what I'd love to be doing:

http://www.planet-mu.com/discog/ziq057/ziq057_02_hi.mp3
http://www.planet-mu.com/discog/ziq057/ziq057_06_hi.mp3
http://www.planet-mu.com/discog/ziq044/ziq044_01_hi.mp3
http://www.planet-mu.com/discog/ziq032/ziq032_03_hi.mp3

Not sure if these examples are even done with Max/msp, but I've noticed that the majority of the acts I see doing music like this use the proggie, and I have to wonder whether this is really a coincidence (compounded by the coincidence that trax like these didn't really start appearing on the scene until msp because more of a staple in the track writing community). Now, I don't want to rip these cats off, don't get me wrong.. and I'm not looking for some easy to use program that will make me aphex twin overnight (i wouldn't want that anynight).. what I want to ask is how one would go about structuring a composition to produce music like this entirely in MAX/MSP ? I know i have many sleepless nights ahead of me making crap music until I get around the learning curve, I've used real modulars, reaktor, the nord modular and PD (max/msp's bastard cousin for the pc).. but, like I said, I'm from the "open drum editor- click notes here.. here.. and one here" mindset. I have a weird idea that music like this is structured more along the lines of manipulating loops of pregenerated rhythm lines through live msp effects, but never touching the programs interface I have no clue how easy this stuff would be to sequence ? I've also noticed many artist using algorhythmic sequences, stuff like the ball bouncing rhythm aphex made so famous on RDJ.. are these rhythms particulary easy to generate? if so, are they typically used live or sampled out for cut'n paste via max/msp later?

Thanx for any help you can give (programs like this are a LARGE investment of time and energy, so I hope you can understand why I stand before you asking such knob like questions)..

btw- for those that care, I do stuff like this now.. but I work with static samples mostly. I generate a rhythm in cubase, export it as audio, manipulate it in soundforge and spit full bars (or warped hits) back into my sampler for triggering as variations. Doing stuff like twerk, richard devine (yeah, I know he uses reaktor), hrvatski or any of these other cats (squarepusher, kid 606, etc) is generally more like musique concrete (tendious sample hacking and reimporting just to sound once, and not be used again) or completely prohibitive cause I can't simulate that organic (i.e. fluid, changing over time) quality of the deconstruction. I'm just looking for the theoretical notions to get my creativity going..

[anarchosyn]

Which tells me I'm either a very silly sod with no story telling abilities, y'allz just too drunk to wanna read all my shite, or y'allz just a bunch of stuck-up nancy-boys (like myself ;). I'll assume the former, as I know the latter probably isn't true.

Let me take another stab at the root of my question:

Ok, I know max/msp is very functional, what I was curious about was the limits of its power and how easy the sequencer functions tend to be.

MAX:
Is it plausable to construct a sequencer fluid enough to allow me to write music (ala example #1 above) WITHOUT needing sample loops? I know doing stuff like autchre's peel sessions 2, or the new album is very simple (well, structure wise--it's basically one complex pattern with slight changes done by hand).. what I'd need would be a very extensive sequencer that would allow me to que patterns.. is this fairly easy to construct in the max enviroment (fairly easy = something people have seen others do, regardless of complexity.. i.e. is it possible and plausible)? Could one chain together these patterns (i.e. to write linear music, if they chose)?

MSP:
On a g3 (which we'll just assume is comparable to a p4 1.4gHz system) would doing music like above (again, see mp3 example #1 ) where the effects and modulations of tones (which will probably be mostly samples) are 100% live in MSP really bog down the 'puter? Keep in mind I'd like to be using MSP as my sampler as well, not effecting external input from my rack sampler. Is this just too intensive for a modern system, or would it not even click over?

As analogy, I'll use the nord modular. In theory, before one understood the 100% DSP limitation (which, btw, they are changing for the v4 update--or so I've been told), you could imagine slewing together 300 OSC, 400 filter, tons of sequencers all set to flip flop and morph around to create new sequences for music. Once the reality of the limitations set in, one saw that doing full compositions (ala my mp3 examples) would be impossible, since creating just a few drum sounds would max out the DSP power, plus the sequencers were very simple, at best. BUT somebody who didn't have one wouldn't be able to really understand how big the confines of the room were or were not until they actually stood in that room (yes, I'm mixing metaphores).

btw- I did manage to create some rather complex noodles that shifted through many measures of a track, some that would go through about 64 bars before changing back, but they were static loops .. plus, if one wanted to change patterns on the fly, they were shite out of luck 9 times out of 10.. I made a joke noodle that did it for the clavia list, and a few others did it with switches and cross fade modules, but the result was hardly what I'm looking for above (i.e. inaccurate, notes would sputter when crossfaded cause values would mix if you turned the knob to slow, or glitch if you hit the switch button 1/10th of a second to late ;)

[Dialect]

this is a very interesting subject, and I wish some one would answer the guy, i find that too much in the music world, producers just want to keep their tricks of the trade to themselves, i think this is very selfish and un-productive to the expansion of a scene.

I want to ditch my old powermac in favour for a new G4 or G5 when they come out, and I'm very interested in playing with this Max/MSP thing I keep hearing so much about. So some theory would be good. whats the the other one Super_Collider is it, supposed to be an easier but similar piece of software. anyway...get on the chit chat.

[benmilstein]

hi
max/msp is an open ended programming environment. to answer your questions anarch yes you can create just about anything you can think of in there if you have the time/dedication/programming chops.
supercollider is certainly not any easier than max/msp.. usually its quite a headfuck for people. sure was for me. more info on that one @ audiosynth.com
max/msp programming takes places in a graphical environment where objects are patched together with virtual patch cords etc. SC is a text code based language based on smalltalk. also very similar structurally to java. but all about the real time audio...
both of these languages require an extremely high level of time and brain committment.
but the results may be worth it to you.......
i still believe that whatever comes out of the speakers is what counts. not how it got there........
peace
ben