rotating without

    Jan 23 2011 | 6:35 pm
    i am involved in a project where we want to spatialize sound in 3D.
    we have a speaker setup in the shape of a cube. this way we can pan the sound left and right, but also up and down.
    my job is to develop a user interface in which you can create 3D-paths for the sound.
    this is all working out nicely. i have created a gui where you can draw shapes, implement curves in between points, transpose and stretch the paths.
    what we want to add is a rotate function of the paths.
    basically, we want to be able to grab the path and transform it just like you can do with the object.
    but the thing is, the paths are made up of long lists of vectors that are put into a coll.
    so, using is not an option... or is it??
    below is a simple demo of how a path is put into a coll, and then how an object is following the path.
    the main goal would be to have the green torus still follow the path after it has been rotated by the rotatexyz message (top right of the patch)
    and help in a good direction would be verrry much appreciated :)

    • Jan 23 2011 | 9:03 pm
      that is a very nice patch! love the lissajou curves, lots to experiment with there... very clean and understandable layout.
      For rotation, you could hold all the points in a matrix instead of a [coll], which would probably be faster anyway. Then you could put that matrix as vertices into with matrixoutput enabled. When you rotate the mesh it would change the outputmatrix points accordingly, then use as needed. (Not to mention other fun matrix FX tweaks.) This is just a guess though, would need to try it out.
      I thought one could enable matrixoutput from its outlet claims that this is possible---but the docs say otherwise!
    • Jan 23 2011 | 10:02 pm
      thanks for the reply
      i thought i tried to run it into a mesh and then have it put out as a matrix again and it didn't work. the matrixoutput didn't work
      i'll have a look at it again though..
      and matrixoutput from doesn't work im sure
    • Jan 29 2011 | 8:40 pm
      i would like to now also how to output a matrix from
    • Feb 02 2011 | 10:20 am
      matrix output from a mesh just doesn't work
      i tried it again and im quite sure the way i did it should be correct
      (you just send the [matrixoutput 1] message right)
      i guess i'll have to calculate the rotations of my path data myself then??
      buhh... thats going to be cpu heavy
    • Feb 02 2011 | 7:40 pm
      hi, sorry i missed this post before.
      yes, matrixoutput is not enabled for or
      you can probably perhaps utilize the quat objects found in the cosm distribution to perform the rotations on each of your points. specifically the quatpov and euler2quat objects.