[Csnd] Project SR for Opcode Development

The Csound Mailinglist
1

Hello everybody,

i can’t find it, but maybe you can help me:

i’m searching a way to get the current project sr into my opcode.
is there a data member in the OPDS struct or somewhere?

- Philipp
Csound mailing list
Csound@listserv.heanet.ie
https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND
Send bugs reports to
        Issues · csound/csound · GitHub
Discussions of bugs and features can be posted here

2

in csoundCore.h:

MYFLT GetLocalSr(OPDS *p)

3

In order to have one source build for both csound 6 and csound 7 I use the following:

static inline MYFLT _GetLocalSr(CSOUND *csound, OPDS *ctx) {
#if CS_API_VERSION_MAJOR == 7
IGN(csound);
return GetLocalSr(ctx);
#else
IGN(ctx);
return csound->GetSr(csound);
#endif
}

which can be called as MYFLT sr = _GetLocalSr(csound, &(p->h)), where p is the pointer to your opcode’s struct

This depends on CS_API_VERSION_MAJOR being set to the correct value, for example within cmake

4

Thanks all! This helped a lot!

I’m still working through some examples i try to understand this incrementing mechanism:

inc     = (int32)(*cpsp++ * CS_SICVT);

CS_SICVT is leading to a struct where sicvtis commented with ‚local sr‘. is this right?
if so, i don’t understand this incrementing mechanism:

static int32\_t randomh\(CSOUND \*csound, RANDOMH \*p\)
\{
    int32       phs = p\->phs, inc;
    uint32\_t offset = p\->h\.insdshead\->ksmps\_offset;
    uint32\_t early  = p\->h\.insdshead\->ksmps\_no\_end;
    uint32\_t n, nsmps = CS\_KSMPS;
    MYFLT       \*ar, \*cpsp;
    MYFLT       amp, min;

    cpsp = p\->xcps;
    min  = \*p\->min;
    amp  = \(\*p\->max \- min\);
    ar   = p\->ar;
    inc  = \(int32\)\(\*cpsp\+\+ \* CS\_SICVT\);
    if \(UNLIKELY\(offset\)\) memset\(ar, '\\0', offset\*sizeof\(MYFLT\)\);
    if \(UNLIKELY\(early\)\) \{
      nsmps \-= early;
      memset\(&ar\[nsmps\], '\\0', early\*sizeof\(MYFLT\)\);
    \}
    for \(n=offset; n<nsmps; n\+\+\) \{
      ar\[n\]     = p\->num1 \* amp \+ min;
      phs      \+= inc;
      if \(p\->cpscod\)
        inc     = \(int32\)\(\*cpsp\+\+ \* CS\_SICVT\);
      if \(phs >= MAXLEN\) \{
        phs    &= PHMASK;
        p\->num1 = randGab\(csound\);
      \}
    \}
    p\->phs = phs;
    return OK;
\}
\`\`\`
5

This is used to read function tables using fixed-point phase, converting frequency to a suitable
sampling increment to update the phase. Unless you understand fixed-point maths, best is to
use floating-point phase and ignore this. Also for csound 7, if fixed-point maths is used, an
alternative floating-point version needs to be implemented too, as we have scrapped the
limitation to power-of-two table sizes.

best