OpenAL for 3D sound in max/msp


    Sep 03 2008 | 1:10 pm
    HI,
    Is there a way to do 3D sound (14 speakers in a sphere) in Max/Msp.
    Using OpenAL ?
    Thanks
    jav

    • Sep 03 2008 | 3:39 pm
      In regard to questions like this, it's often best
      to have a look at maxobjects.com, typing in key
      words you might be searching for - in addition to
      doing the same search of the archives of the Max
      list.
      It would appear that the answer is "no", or that
      anyone who's done it has decided to remain mum
      about the matter. The relative paucity of umptyspeaker
      3d arrays in the real world might also contribute to
      a certain lack of critical mass. :-)
    • Sep 03 2008 | 9:33 pm
      Thank you for the answer,
      It will be fun to found my way :)
    • Sep 03 2008 | 9:43 pm
    • Sep 04 2008 | 5:52 am
      Wavefield synthesis, very interesting hint, thanks a lot.
      I want to suggest different accoustic space/field and transform/merged them. I thought of using rtconv~.
      Spat or ambiosonic object don't seems to have an Elevation factor.
      Inside of this changing field, I want to control sound movements in a 3D array with a 3D representation/trigger system.
    • Sep 04 2008 | 6:04 am
      Ambisonics allow you to control a sound source on the surface of a
      sphere, so there are typically 2 angles (like in 3D spherical
      coordinates), azimuth and elevation. They don't really lend
      themselves to placing sources inside our outside the speaker array
      though. Wavefield synthesis in 3D is not feasible in realtime at this
      point. It can be done in 2D but requires a good understanding of the
      technique to get everything right.
      wes
      On Wed, Sep 3, 2008 at 10:52 PM, j-Ambroise Vesac wrote:
      >
      > Wavefield synthesis, very interesting hint, thanks a lot.
      >
      > I want to suggest different accoustic space/field and transform/merged them. I thought of using rtconv~.
      >
      > Spat or ambiosonic object don't seems to have an Elevation factor.
      >
      > Inside of this changing field, I want to control sound movements in a 3D array with a 3D representation/trigger system.
      >
      >
    • Sep 04 2008 | 6:13 am
      That's help me to have a better understanding on both techniques. I thought it was possible to move a source along a path across a periphonic space.
    • Sep 04 2008 | 7:35 am
      I do believe, however, that you can do 3d with vbap, if that would
      work for you.
      M
      On Sep 4, 2008, at 2:13, j-Ambroise Vesac wrote:
      >
      > That's help me to have a better understanding on both techniques. I
      > thought it was possible to move a source along a path across a
      > periphonic space.
    • Sep 04 2008 | 10:15 am
      > Ambisonics allow you to control a sound source on the surface of a
      > sphere, so there are typically 2 angles (like in 3D spherical
      > coordinates), azimuth and elevation. They don't really lend
      > themselves to placing sources inside our outside the speaker array
      > though.
      Not to blow my own horn, but I address this in a paper I presented at SMC this year:
      It's only a proposal and a work in progress, but if anyone cares to give some feedback I'd be glad to discuss it.
      All the best,
      Rui
    • Sep 20 2008 | 8:56 pm
      Quick summary: vbap and ambisonic share the property of not encoding
      distance (by default), but in both cases distance cues can be
      simulated using appropriate filter/delay combos. Ambisonics has had
      some proposed extensions to more accurately reconstruct distance
      effects, but I don't believe these have been implemented in any of
      the various Max/MSP externals yet. Reverberation is likewise absent
      from vbap and ambisonics by default, but can be added in a more or
      less ad-hoc manner. In general, vbap is going to give you a slightly
      less 'spatial' and more 'pointillistic' ambience, while ambisonics
      (if properly configured) will grant an expanded 'sweet' listening
      area (especially with the higher order ambisonics). vbap is
      algorithmically simpler, but when using max externals that won't make
      much difference. both vbap and ambisonics depend upon having a
      regularly distributed array of speakers (as equidistant from the
      center point and to each neighbor as possible). both vbap &
      ambisonics can be 2d or 3d.
      As Wesley stated, WFS will be superior in terms of spatial imaging
      and distance effects, but unless you have hundreds of well-matched
      speakers almost side by side, it's not going to happen; real-time
      moving objects may be extremely CPU intensive and the 3D situation is
      very complex indeed.
      On Sep 4, 2008, at 12:35 AM, mattyo wrote:
      > I do believe, however, that you can do 3d with vbap, if that would
      > work for you.
      >
      > M
      >
      >
      > On Sep 4, 2008, at 2:13, j-Ambroise Vesac wrote:
      >
      >>
      >> That's help me to have a better understanding on both techniques.
      >> I thought it was possible to move a source along a path across a
      >> periphonic space.
      >
      Be seeing you
      grrr waaa
      www.grahamwakefield.net
    • Sep 21 2008 | 10:22 pm
      Quote: j-Ambroise Vesac wrote on Wed, 03 September 2008 14:10
      ----------------------------------------------------
      > Is there a way to do 3D sound (14 speakers in a sphere) in Max/Msp.
      >
      > Using OpenAL ?
      ----------------------------------------------------
      BTW. There is some source code for an Open AL external for pd on Windows that I found some time ago.
      It would need developing ... (not least since it supports only 5 sources and is pd not Max/MSP)