Definitively is a bug, thanks for reporting, added on the TODO list.
Stéphane
> Le 13 nov. 2020 à 18:00, Oleg Nesterov <o...@redhat.com> a écrit :
>
> On 11/13, Karl Schultheisz via Faudiostream-users wrote:
>>
>> process = os.osc(55) : rwtable(1024, 0.1, 42, _, 42);
>>
>> The result, as demonstrated using faustide, is that the hypothesis is wrong.
>> The output signal is a constant 0.1, not a 55Hz sinusoid.
>
> I think this is a bug. faust wrongly assumes that if ridx is constant then
> the output is constant too.
>
> Perhaps this is because
>
> Tree sigWriteReadTable(Tree n, Tree init, Tree widx, Tree wsig, Tree
> ridx)
> {
> return sigRDTbl(sigWRTbl(gGlobal->nil, sigTable(gGlobal->nil, n,
> sigGen(init)), widx, wsig), ridx);
> }
>
> and the optimization above is valid for sigRDTbl.
>
>
> For example,
>
> process = rwtable(1024, 0.0, ffunction(int widx(),"",""), _, 0);
>
> compiles to
>
> virtual void instanceConstants(int sample_rate) {
> fSampleRate = sample_rate;
> mydspSIG0* sig0 = newmydspSIG0();
> sig0->instanceInitmydspSIG0(sample_rate);
> sig0->fillmydspSIG0(1024, ftbl0);
> fConst0 = ftbl0[0];
> deletemydspSIG0(sig0);
> }
> virtual void compute(int count, FAUSTFLOAT** inputs, FAUSTFLOAT**
> outputs) {
> FAUSTFLOAT* input0 = inputs[0];
> FAUSTFLOAT* output0 = outputs[0];
> for (int i = 0; (i < count); i = (i + 1)) {
> ftbl0[int(widx())] = float(input0[i]);
> output0[i] = FAUSTFLOAT(fConst0);
> }
> }
>
> Oleg.
>
_______________________________________________
Faudiostream-users mailing list
Faudiostream-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/faudiostream-users
