Re: [Sndobj-devel] (no subject)

[sonify <so...@av...>]

Before extending sndobj with math functions, it would be as well to  
test how much more efficient (if at all) this would be to using  
Numpy, which itself is a linked library written in C and pretty fast,  
especially for matrix algebra.

David
On 27/01/2007, at 5:29 AM, Victor Lazzarini wrote:

>
>
> At 18:03 26/01/2007, you wrote:
>> victor,
>> would there be any particular utility in having a series of SndObj
>> that performed math like functions. I am thinking in particular for
>> use with python or lisp scripting since they might not be as much use
>> in c++.
>
> Probably useful, if it would speed up calculation. These can be
> done in Python OK, but probably less efficiently
>
>> for instance
>>
>> class Asin : public SndObj{
>> public:
>>
>>     Asin(SndObj* input, int vecsize=DEF_VECSIZE, float sr=DEF_SR);
>>     ~Asin();
>>
>>     short DoProcess();
>>
>> };
>>
>> Asin::Asin(SndObj* input, int vecsize, float sr):
>>   SndObj(input, vecsize, sr) {
>> }
>>
>> Asin::~Asin() {
>> }
>>
>> short Asin::DoProcess(){
>>   if(!m_error){
>>     if(m_input){
>>       for(m_vecpos = 0; m_vecpos < m_vecsize; m_vecpos++){
>>         if(m_enable)
>>           m_output[m_vecpos] = asinf(m_input->Output(m_vecpos));
>>         else m_output[m_vecpos] = 0.f;
>>       }
>>       return 1;
>>     } else {
>>       m_error=3;
>>       return 0;
>>     }
>>   }
>>   else return 0;
>> }
>>
>> might not be the best way to do this in some situations but in others
>> it might be necessary.
>>
>>
>> in particular, and excuse my ignorance about python (i am starting to
>> learn it, btw), but do SndObj operators get imported in some way? for
>> instance would the '+' below be valid in python?
>>
>> line = Interp(0, 10000, 0.05)
>> osc = Oscili(tab, pitch, 0, None, line + .2)
>
> I don't think that would work in Python. It would not work in
> C++ either (you have to keep adding values every control
> period).
>
> However, I never tried the overloaded ops in Python.
>
>
>> for the cffi bindings it is not and really needs some type of Add
>> sndobj:
>>
>> class Add : public SndObj{
>> public:
>>
>>     Add(SndObj* input1, SndObj* input2, int vecsize=DEF_VECSIZE,
>> float sr=DEF_SR);
>>     Add(float input1, SndObj* input2, int vecsize=DEF_VECSIZE, float
>> sr=DEF_SR);
>>     Add(SndObj* input1, float input1, int vecsize=DEF_VECSIZE, float
>> sr=DEF_SR);
>>     ~Add();
>>
>>     short DoProcess();
>>
>> };
>>
>> then could be used
>>
>> (setf line (new_interp 0 10000 .05))
>> (setf osc (new_oscili tab pitch 0 nil (new_add line osc)))
>>
>>
>> i have a bunch of these defined in a basic way - would they be useful
>> or merely included just for the benefit of lisp. if the latter than
>> maybe i could but them as separate code that SWIG would wrap?
>
> We can think about it. I'll check what can and can't be done in  
> Python.
>
> At the moment, it's the Intel Mac bug that is making me worry. I  
> really
> can't see what the issue is. But then again I only hear reports.
>
> All the best
>
> Victor Lazzarini
> Music Technology Laboratory
> Music Department
> National University of Ireland, Maynooth
>
>
> ---------------------------------------------------------------------- 
> ---
> Take Surveys. Earn Cash. Influence the Future of IT
> Join SourceForge.net's Techsay panel and you'll get the chance to  
> share your
> opinions on IT & business topics through brief surveys - and earn cash
> http://www.techsay.com/default.php? 
> page=join.php&p=sourceforge&CID=DEVDEV
> _______________________________________________
> Sndobj-devel mailing list
> Snd...@li...
> https://lists.sourceforge.net/lists/listinfo/sndobj-devel
>

_________________________________________________
experimental polymedia:	www.avatar.com.au
Sonic Communications Research Group,
University of Canberra:	 www.canberra.edu.au/vc-forum/scrg
vip=Verbal Interactivity Project