mxj~ and dynamic patching of DSP routines


    May 15 2006 | 2:01 pm
    David Ziccarelli wrote the following in his very interesting article "MSP Learns to Ride a Bike":
    "mxj~ is interesting because it shows a higher-level model for developing signal processing objects where the programmer can fill in a few Java classes and have a working object. It also supports a model for dynamic patching of DSP routines that we hope will be the basis of a future MSP."
    Please could you explain what does mean "It also supports a model for dynamic patching of DSP routines" ?
    Is this a way to allocate on the fly dynamic DSP objects similar to what Supercollider does ?
    All the best
    Alessandro Fogar

    • May 15 2006 | 6:48 pm
      If all your classes support the MSPPerformable or a similar interface you can easily build signal graphs on the fly. Topher
    • May 15 2006 | 7:55 pm
      I don't see any reason why you couldn't do this in conventional MSP externals - it's just code - I guess the DSP cycling is an artefact of the way Max/MSP deals with recompilation after rearrangements of patcher cords.
      -- N.
      nick rothwell -- composition, systems, performance -- http:// www.cassiel.com
    • May 15 2006 | 7:59 pm
      What I think you're missing, Nick, is that mxj~ provides an interface by which you can create unit generators in Java, and then mix and match those unit generators on the fly in another mxj~ object. There currently is no simple way to load another C msp external within a C external.
      Ben
    • May 15 2006 | 8:48 pm
      Fair enough.
      If I felt like it, I'm sure I could go away and invent some kind of dynamic dispatching framework to allow me to load up native DLL's for C signal vector processing and get the same kind of mixing and matching, but it would be hard work and awfully fiddly and messy, and I most certainly do not feel like it...
      -- N.
      nick rothwell -- composition, systems, performance -- http:// www.cassiel.com
    • May 16 2006 | 7:25 pm
      > If I felt like it, I'm sure I could go away and invent some kind of > dynamic dispatching framework to allow me to load up native DLL's for > C signal vector processing and get the same kind of mixing and > matching...
      absolutely, and it's our hope that we can provide this kind of framework sometime in the future.
      Ben
    • May 16 2006 | 8:25 pm
      Hi,
      If I remember well somebody on the list said he developed something similar but then stopped.
      Many thanks for all the replies.
      Alessandro Fogar