beyond Olive's Hybrid math

topic posted Thu, October 11, 2007 - 1:21 AM by  Alien Jon
Ok so I've been working with a more in-depth model for "hybrids" "composite spinning" whatever you want to call it... It's all just relations in configuration space. At any rate its starting to gel to where I can talk about it semi-cohesively, thanks to playing with Cat Eyes as of late. So if you aren't already familiar with Olive's post about possible hybrid combinations here's your homewor:
...doing the math on hybrids:
techpoi.tribe.net/thread/78...d54c0dc644

Weeellll, his math was incomplete (or maybe I just like to over complicate things). So, I'm really starting to get comfortable with spinning a Cat Eye. Cat eye is Arash'ese for a 1beat antispin ellipse:
techpoi.tribe.net/thread/55...60ee9f6b34
Zan post as well:
techpoi.tribe.net/thread/6d...6bac1a188)


As a result, I've been thinking of all those Olive hybrids in terms of one beat and unit circle. The poi unit circle idea is that your poi length = 2units, in other words, poi length is the diameter of the circle an isolation makes. Same thing for the extension that can be substituted in for an iso hybrid style... and same thing for your hands path in a cat eye. I can't yet do a horizontal cat eye, but I think it's possible.
Anyhow, once I started analyzing all those Olive hybrids as unit circle one-beats instead of polyrhythms, The relational poi thing started to become way more clear to me. *deep breath*

So you have to look at the relation of:
a) Lhand vs Lpoi [0,0]
b) Rhand vs Rpoi [0,0]
c) Lpoi vs Rpoi [0,0]
d) Lhand vs Rhand [0,0]
e) Lhand vs Rpoi [0,0]
f) Rhand vs Lpoi [0,0]

Each relation has 2 binary variables:
Orientation (spin)= same[0] or opposite[1]
Phase (timing)= same[0] or opposite[1] (I'm leaving quarter-time and polyrhythmn out for now, more complex model later)

So as far as hand orientation and phase, it's determined by a vector pointing through the hand, away from the pivot for the arm; just like for the poi it is determined by a vector through the head, pointing away from the pivot as it spins. When the vector has cycled through all orientations in the spin plane, that is one beat... or cycle.

Now if you have the same values in a-f, you have a high level of symmetry in poi "configuration space". This gives you 4 different non-hybrid movements:
same-time extensions
same-time isolations
split-time extensions
split-time isolations

The 4 "same poi vs same hand" relationships determine the unit circle "driving styles":
[0,0]=extension
[0,1]=isolation
[1,0]=vertical cat eye
[1,1]=horizontal cat eye

We are all familiar with the 4 poi vs poi and hand vs hand spin but lets write it out anyway:
same direction, same time [0,0]
same direction, split time [0,1]
opposite direction, same time [1,0]
opposite direction, split time [1,1]

In Olives model he applied the 4 to hand/arm motion, but only direction and not timing to poi motion, assuming that they naturally fit in certain ways. In the case of some of the hybrids with more uniform configurations, if you change the poi timing they turn into one of the 4 non-hybrids from above. However, the more complex ones have different configurations based on timing: ie a vertical cat eye as opposed to a horizontal cat eye.


I'm still digesting and grokking the importance of poi vs opposite hand, as something more than an epiphenomena, but it feels like it is another angle that is pertinent.

Ultimately this the beginning of a model to composite movements together that encompasses and goes beyond hybrids. I'd like to try and make it extensible. It should include stalls, compound timings and hence compound circles, linear movements as well as the circular movement, polyrhythmns, etc. This should lay the foundation to describe other driving styles such as spin/antispin, pundula, whips, parallels, etc.

I'm working on animating all this stuff. I have the original Olive list transposed to unit circle 1-beats. I need to render it and compile a video... and of course start to expand it exhaustively to show all of the combinations, with an attribute overlay, so people can start to see the relations at a glance.

Ok 4:20am,
-Alien signing off
posted by:
Alien Jon
Boston
  • Re: beyond Olive's Hybrid math

    Thu, October 11, 2007 - 6:57 PM
    awesome man i definitely need to grok this over a bit but i'm glad you're really running iwth this, I like how it's starting to converge,
    • Re: beyond Olive's Hybrid math

      Thu, October 11, 2007 - 10:07 PM
      Hmm all of Olive's original post made sense but by the time I finished it I just want to get up and spin. So maybe I'll tackle yours at a later date or next time I see you in person. From reading the posts a bit, I think I'm working on a lot of this stuff but I'd rather practice them then figure out how to put them in words. So hopefully soon I'll get to it and make a sensible post about this. But for tonight I have too many antibrid/head orbit patterns in my head to sort out. Off to spinning!
      • Re: beyond Olive's Hybrid math

        Sat, October 13, 2007 - 9:19 PM
        *Quick update*
        The videos are all rendered for the unit circle versions of The hybrids listed by Olive. I just need to finish making text overlays to show code for each and compile it all.

        It occurs to me that for now L(eft) and R(ight) should be taken as 1st and 2nd. That way the patterns can be thought of in both directions. If I'm to include actual left and right then I also need to define clockwise or anti-clock wise spins linked to it like Rev does. I think I'll probably revise this later to include "clock".

        -Alien Jon
        • Re: beyond Olive's Hybrid math

          Sun, October 14, 2007 - 11:15 PM
          To clarify the Poi Unit Circle: I refer to the circle an isolation makes because it is the smallest possible circle that BOTH poi head and hand can make at the same time. If you go further and approach a head (point) isolation then the circle your hand makes increases in size. It just so happens that this unit circle is the size of the hand motion in a cat eye as well.
          With that in mind, here is a simulation of all of the Hybrids Olive listed translated into unit circle 1-beats:
          youtube.com/watch

          This keeps more true to the original definition of hybrids. It also makes further interesting patterns appear as we play with the timing values for poi that Olive didn't specify. So far I've only shown patterns that fit Olive's math, but I've omitted horizontal Cat Eyes because I can't spin them yet.
          So far I'm good 1A, 2A, 3A, 1B, and 3B; almost 4B. I can kind of manage 4A and 2B after some flailing.

          -Alien Jon
          • Re: beyond Olive's Hybrid math

            Mon, October 15, 2007 - 9:08 AM
            Oh yeah so I switched 3b and 4b around in the videos on accident. It was late, I was nodding off, so what da ya want ;-P
            • Re: beyond Olive's Hybrid math

              Mon, October 15, 2007 - 6:45 PM
              very very cool. I like the elipsed trammelish one. 3A I believe it was.
              • Re: beyond Olive's Hybrid math

                Mon, October 15, 2007 - 6:59 PM
                3A, 4a, 1B, and 2B all have Cateyes in them. A Cateye could be thought of as an expanded trammel. For that matter, if you just linear isolate instead of taking your hand around the unit circle, that IS a poi trammel.
                • Re: beyond Olive's Hybrid math

                  Mon, October 15, 2007 - 11:53 PM
                  That was nice, I like that you can simulate the trail before learning the move.
                  Kind of lets you figure out if its worth the time.

                  You are truly a tech-poi. =D (compliment)
                  • This is the maximum depth. Additional responses will not be threaded.

                    Re: beyond Olive's Hybrid math

                    Wed, October 17, 2007 - 7:36 PM
                    And the trails aren't a video after effect. They are full 3d particle systems... which means if I start animating camera pans and fly-bys you can still see exactly whats going on the whole time!
  • Re: beyond Olive's Hybrid math

    Sun, October 21, 2007 - 7:07 PM
    So what we're saying is that while hybrids, in poi, can, without going all the way to the staff-style '2 driving styles definition', the "one giant one isolated" hybrid is, in fact, part of a sub-set of hybrids that could be called unit-circle hybrids...am I right? And these are the permutations therein...thoughts?
    • Re: beyond Olive's Hybrid math

      Sun, October 21, 2007 - 7:37 PM
      Pretty much. I feel like if your hands are moving in the unit circle that it is much closer to the original definition of hybrids (It's all just silly words in the end). These aren't all the permutations though. I didn't include horizontal Cateyes cuz I cant spin them yet and I didn't get into substituting isolations for long arm vs the cateyes. Ultimately, though the "Alien code" is the 1st simple model for relating the elements of relational poi spinning. That encompasses describing anything from simple spinning to stuff that goes beyond what I'd call hybrid. Pretty much the broad sense of "Composite spinning" I was talking about.
      • Re: beyond Olive's Hybrid math

        Mon, December 3, 2007 - 1:21 PM
        Update:
        The link to Zan's post in my original Post doesn't work. So here is the working link:

        techpoi.tribe.net/thread/6d...f6bac1a188
        • Re: beyond Olive's Hybrid math

          Fri, April 11, 2008 - 8:44 AM
          Update: Here is a new visualization I rendered showing the relationship between the different Poi Unit Circle driving styles and their linear boundaries:
          www.youtube.com/watch

          Disclaimer: the horizontal trammel and (to a slightly lesser degree) horizontal cateye are next to impossible to spin. So don't beat your self up over it if you have a hard time with it. Here are examples of the vertical versions (which also appear in this vid):

          Trammel:
          www.youtube.com/watch

          Cateye (antispin ellipse):
          www.youtube.com/watch

          Enjoy!
          -Alien Jon
          • Re: beyond Olive's Hybrid math

            Sun, April 13, 2008 - 10:30 AM
            Good stuff! Glad to see "the dark side of the circle" in action.
            Something tells me that futuristic poi-beings will spin horizontal trammels as a warm-up.
            • Re: beyond Olive's Hybrid math

              Sun, April 13, 2008 - 11:45 AM
              Yeah, I'm pretty sure we are just beginning to produce a new breed of bender (ala the Avatar cartoon series): The Gravity Bender. ;-P
              • Re: beyond Olive's Hybrid math

                Mon, April 21, 2008 - 1:59 PM
                how do you make them all? Like what program are you using? or how are you creating it?
                • Re: beyond Olive's Hybrid math

                  Mon, April 21, 2008 - 3:26 PM
                  I use 3D Studio MAX 9, from Autodesk/Discreet... Cuz I'm familiar with it. =) It's one of the more well rounded industry standard 3d packages out there. So far, I'm giving a predominantly 2D view of patterns as viewed from performance plane. But when the time comes, I can pan around any angle of a 3D pattern as well.

                  Then I've just been using Windows Movie Maker to edit titles in.

                  -Alien Jon
                  • This is the maximum depth. Additional responses will not be threaded.

                    Re: beyond Olive's Hybrid math

                    Sat, April 26, 2008 - 12:24 PM
                    sweet man, its cool to find out the logistics behind it, thanx for sharing that.

                    I was thinking down the line of trying to develop a program in flash,
                    so that you could pick the actions of what each hand is doing.
                    And potentially having a list of moves you could choose, and being able
                    to mix and match it all... very similar to your tech vids, but with interactivity.

                    ase x
                    • Re: beyond Olive's Hybrid math

                      Sat, April 26, 2008 - 1:07 PM
                      That would be poi bot, over on spherculism.net

                      But it's out of date and clunky. MCP and I have talked a bit about using the J-monkey java-based 3d game engine for such an app. Probably pretty easy with out a human body in there. But getting the inverse-kinematics right so that the body looks like it is spinning the prop properly... instead of twisting around like Gumby... that is tricky!

                      Ase do you have any Java and/or 3d coding experience? Cuz that would be sweet!

                      -Alien Jon
                      • Re: beyond Olive's Hybrid math

                        Sat, May 10, 2008 - 1:45 AM
                        unfortunately I have very minimal experience with Java and no 3d coding exp.
                        although I have always wanted to get into 3d coding...3d max a good place to start?

                        yeah I could definitely imagine the trickyness of getting the body to form right, have you
                        ever messed with half life's game engine called... gary's mod? You might be able to
                        go from there, but I assume it would still take some tricky coding to get it all right...

                        oh btw I am a friend of the great Justice of Dragons, he told me to say hi :)
                        I was actually the one who showed him one of your lastest vids with Zan (poi trans.) and
                        inspired him to get a hold oh ya, after that, its like its a lot smaller world then I thought...

                        but that would be cool to see the Jmonkey in action... keep us updated :) until then...

                        keeps it spinnin'

                        ase x
  • Re: beyond Olive's Hybrid math

    Fri, May 30, 2008 - 11:35 AM
    ok...lots of cateye derived flowers coming up in the next tech lounge we film.

    problem with the vertical and horizontal cateyes is that the vertical has 1 or 2 tention points and the horizontal has to have 3 tention points.....its kinda polyrhythmic in a way.....
    • Re: beyond Olive's Hybrid math

      Fri, May 30, 2008 - 1:25 PM
      Why would they be different?
      • Re: beyond Olive's Hybrid math

        Fri, May 30, 2008 - 1:34 PM
        its not actually poly rhythmic, its just to make a vertical cateye, you have tention points at the top and bottom, and then extention through both sides.

        to make a good looking horizontal cateye, you extend through to top, and then you spike at the bottom, so you get the left and right limits of the extention on top, and the vertex of the angle that completes that triangle on the bottom.

        you can actually do horizontal cateyes pretty slowly....video to come.
        • Re: beyond Olive's Hybrid math

          Thu, June 5, 2008 - 9:16 AM
          So... I just saw this video on Home Of Poi. It's a fire club swinging video and I was wondering if some of the stuff he is doing is actually possible with poi. I've been trying to figure out some of the stuff, so far not much luck. Perhas some of you crazy minds can think of something, posting it here since by my definition they are definitely hybrids.:

          www.dailymotion.com/shadowbl..._creation

          At least the stuff I'm interested in =)
          • Re: beyond Olive's Hybrid math

            Fri, June 6, 2008 - 5:13 AM
            Okay, I'm giving up poi and taking up club swinging...just because of that amazing video!
            My favorite moves of his are when he has the wick high and the hand low, and then he brings the arm up, and does an antispin-to-linear isolation out to the side. (1:37)
            I tried doing this with poi, and while all the parts are doable, the point isolation to get the hand above the poi kills all the momentum, so that the linear-iso can't just flow out of it. I think a lot of the beauty of the move lies in the total consistency of speed therein.
            • Re: beyond Olive's Hybrid math

              Fri, June 6, 2008 - 6:02 AM
              yeah, that was sick....does kinda make me think about takin up clubs! i do think *most* of it is possible(highly difficult, mostly) with poi, but some of the point isos just linger against gravity for too long to be feasible.