pland

I'm trying to port a Puredata object to max/msp. We used the Flext framework until now, but I'm just trying a clean Max5 implementation in C/C++.

I have this perform method in the main file:

    t_soundspotter* x = (t_soundspotter *)(w[1]);

    t_sample* inL = (t_sample*)(w[2]);

    t_sample* outL = (t_sample*)(w[3]);

x->spotter.run(n, &inL[0], &inL[1], &outL[0], &outL[1]);

float result = (float)x->spotter.reportResult();

which I'm hoping will (eventually) send the incoming signal data from max to the soundspotter run method, which is contained in an implementation file:

void SoundSpotter::run(int n, ss_sample* ins1, ss_sample* ins2, ss_sample* outs1, ss_sample* outs2)

My biggest problem right now is that reading the Max5 SDK, I can't get a clear idea of how the main max object gets and passes signal data around:

the above is just a guess, but can anyone see where I'm going wrong? right now, I get nothing from an adc~ if I hook it up to the audio inlet I've given the object.

	t_soundspotter* x = (t_soundspotter *)(w[1]);
	t_sample* inL      = (t_sample*)(w[2]);
	t_sample* outL     = (t_sample*)(w[3]);
	int n             = (int)w[4];

    x->spotter.run(n, &inL[0], &inL[1], &outL[0], &outL[1]);

    float result = (float)x->spotter.reportResult();

dsp-perform-function-confusion

Quote: pland wrote on Mon, 16 February 2009 14:50

----------------------------------------------------

> void SoundSpotter::run(int n, ss_sample* ins1, ss_sample* ins2, ss_sample* outs1, ss_sample* outs2)

OK - first a t_sample * is a pointer to a 32 bit float - I'd assume that your ss_sample* is the same, but I can't be sure with the code you've provided.

Secondly, your guess is almost certainly not right - but you've not really made it clear *exactly* what you want to do. You are passing the pointers to the first and second samples of the inL and outL variables - which I would guess is not what you intend.

I assume that you might be trying to send 2 channels of audio in and out? - inL is not a pointer array of channels, simply an array of samples. Without your dsp routine it's not possible to check what input or output inL really refers to, or if it even holds a valid pointer to an array of samples, as the array of values to the perform routine is defined by the code in the dsp routine.

If you do want 2 channels in and out then you'll probably need to change your dsp routine to pass an array with more than 1 channel in and out - your perform routine implies you've copied an example for only one channel in or out. I'd suggest looking at the SDK examples, as well as the docs - it should make things a lot clearer. Check the object structures, main, new, dsp and perform routines - they'll all need to be right to get audio in and out.

Beyond that it's quite hard to help further without more info. Ideally, at least the new, main and dsp routines would be needed to see what would actually happen with the code you've provided. Outlining exactly what you're trying to achieve would also be helpful.

Quote: pland wrote on Mon, 16 February 2009 14:50
----------------------------------------------------
>
> void SoundSpotter::run(int n, ss_sample* ins1, ss_sample* ins2, ss_sample* outs1, ss_sample* outs2)

Alex: thanks for your patient response. I'll gather my thoughts and come back with more illustrative code and better questions later today.

dsp perform function confusion