j.pack_equal/j.pack.cpp

Go to the documentation of this file.

/** @file
 *
 * @ingroup implementationMaxExternalsAudioGraph
 *
 * @brief j.pack= : Max external packing several MSP audio signals onto one AudioGraph multichannel audio signal.
 *
 * @details This object functions as a source (generator) for #TTAudioSignal usable by a Jamoma AudioGraph dsp chain.
 *
 * @authors Tim Place, Trond Lossius, Nils Peters
 *
 * @copyright Copyright © 2008 by Timothy Place @n
 * This code is licensed under the terms of the "New BSD License" @n
 * http://creativecommons.org/licenses/BSD/
 */


#include "j.pack.h"



/************************************************************************************/
// Main() Function

int C74_EXPORT main(void)
{
    t_class* c;
    
    TTAudioGraphInit(); 
    common_symbols_init();
    
    c = class_new("j.pack=", (method)PackNew, (method)PackFree, sizeof(Pack), (method)0L, A_GIMME, 0);
    
    class_addmethod(c, (method)PackReset,               "audio.reset",      A_CANT, 0);
    class_addmethod(c, (method)PackSetup,               "audio.setup",      A_CANT, 0);
    class_addmethod(c, (method)MaxAudioGraphDrop,       "audio.drop",       A_CANT, 0);
    class_addmethod(c, (method)MaxAudioGraphObject,     "audio.object",     A_CANT, 0);
    class_addmethod(c, (method)PackDsp64,               "dsp64",            A_CANT, 0);     
    class_addmethod(c, (method)PackAssist,              "assist",           A_CANT, 0); 
    class_addmethod(c, (method)object_obex_dumpout,     "dumpout",          A_CANT, 0);  
        
    class_dspinit(c);
    class_register(_sym_box, c);
    sInClass = c;
    return 0;
}


/************************************************************************************/
// Object Creation Method

PackPtr PackNew(t_symbol* msg, long argc, t_atom* argv)
{
   PackPtr  self;
    TTValue sr(sys_getsr());
    long    attrstart = attr_args_offset(argc, argv);
    TTValue v;
    TTErr   err;
    
    self =PackPtr(object_alloc(sInClass));
    if (self) {
        self->maxNumChannels = 2;
        if (attrstart && argv)
            self->maxNumChannels = atom_getlong(argv);
        
        // Set the sample rate of the Jamoma AudioGraph environment
        ttEnvironment->setAttributeValue(kTTSym_sampleRate, sr);
        
        // Create an embedded Jamoma AudioGraph object functioning as a signal generator.
        v.resize(3);
        v[0] = "audio.generator";
        v[1] = 0; // no audio graph inlets (only msp inlets)
        v[2] = 1; // one audio graph outlet
        err = TTObjectBaseInstantiate(TT("audio.object"), (TTObjectBasePtr*)&self->audioGraphObject, v);
        self->audioGraphObject->addAudioFlag(kTTAudioGraphGenerator);
        // Self check and return error if this did not work out.
        if (!self->audioGraphObject->getUnitGenerator().valid()) {
            object_error(SELF, "cannot load audio.generator");
            return NULL;
        }

        attr_args_process(self, argc, argv);
        
        // Create rightmost outlet for dumping attributes etc.
        object_obex_store((void*)self, _sym_dumpout, (object*)outlet_new(self,NULL));
        
        // Create AudioGraph outlet
        self->audioGraphObjectOutlet = outlet_new((t_pxobject*)self, "audio.connect");
        
        // Register this object as a DSP object with MSP
        dsp_setup((t_pxobject*)self, self->maxNumChannels);
        
        // Set flags necsessary to ensure correct interaction with the MSP processing chain.
        self->obj.z_misc = Z_NO_INPLACE | Z_PUT_FIRST;
    }
    return self;
}

// Memory Deallocation
void PackFree(PackPtr self)
{
    dsp_free((t_pxobject*)self);
    // Release the AudioGraph object.
    TTObjectBaseRelease((TTObjectBasePtr*)&self->audioGraphObject);
}


/************************************************************************************/
// Methods bound to input/inlets

// Method for Assistance Messages
void PackAssist(PackPtr self, void* b, long msg, long arg, char* dst)
{
    if (msg==1)         //Packlets
        if (arg > 0) { snprintf(dst, 256, "(signal) single-channel input Nr. %ld", arg + 1); }
        else { strcpy(dst, "control messages and (signal) single-channel input Nr. 1");}
    else if (msg==2) {  // Outlets
        if (arg == 0)
            strcpy(dst, "multichannel output");
        else
            strcpy(dst, "dumpout");
    }
}


TTErr PackReset(PackPtr self, long vectorSize)
{
    return self->audioGraphObject->resetAudio();
}


TTErr PackSetup(PackPtr self)
{
    t_atom a[2];
    
    atom_setobj(a+0, (t_object*)(self->audioGraphObject));
    atom_setlong(a+1, 0);
    outlet_anything(self->audioGraphObjectOutlet, gensym("audio.connect"), 2, a);
    return kTTErrNone;
}


void PackPerform64(PackPtr self, t_object* dsp64, double **ins, long numins, double **outs, long numouts, long sampleframes, long flags, void *userparam)
{   
    for (TTUInt32 i=0; i < self->numChannels; i++)
        TTAudioGraphGeneratorPtr(self->audioGraphObject->getUnitGenerator().instance())->mBuffer->setVector64Copy(i, self->vectorSize, ins[i]);
}


void PackDsp64(PackPtr self, t_object* dsp64, short *count, double samplerate, long maxvectorsize, long flags)
{
    self->vectorSize = maxvectorsize;
    
    // COMMENT: In the following commented out code we attempted to reduce the number of channels of the generated AudioGraph signal in the case that the rightmost inlets of j.pack= did not have any audio signals connected. This would save us some CPU. However it also introduced a bug logged as issue #16 at github.com/jamoma/JamomaCore. It might be that instead of considering this a bug in j.pack=, we rather should consider it a bug when combining several AudioGraph signals, to be resolved elsewhere. But this fixes the issue for now. [TL-2013-08-04]
    /*
    // Find the rightmost inlet that an audio signal is connected to.
    TTUInt16    highestIndexForConnectedSignal = 0;
    
    self->numChannels = 0;
    for (TTUInt16 i=0; i < self->maxNumChannels; i++) {
        self->numChannels++;
        if (count[i])
            highestIndexForConnectedSignal = i;
    }
    
    self->audioGraphObject->setOutputNumChannels(0, highestIndexForConnectedSignal+1);
    */
    
    // We always set j.pack= to process all channels:
    self->numChannels = self->maxNumChannels;
    self->audioGraphObject->setOutputNumChannels(0, self->numChannels);
    
    self->audioGraphObject->getUnitGenerator().set(kTTSym_vectorSize, self->vectorSize);
    self->audioGraphObject->getUnitGenerator().set(kTTSym_maxNumChannels, self->maxNumChannels);
    self->audioGraphObject->getUnitGenerator().set(kTTSym_sampleRate, samplerate);
    
    object_method(dsp64, gensym("dsp_add64"), self,PackPerform64, 0, NULL); 
    //dsp_add64(dsp64, (t_object*)self, (t_perfroutine64)PackPerform64, 0, NULL);
}