jpff@maths.bath.ac.uk wrote:
> 
> There is no noctdft opcode in the Csound that I know -- it is not in
> the manual. 

Yes it is. It's in the manuals - both in HTML and the 3.0 one which we all 
started from. It also WAS in Csound until fairly recently - I happen to have
some snips from -95, and it's right there in Entry.C (together with octdown)
just before the spec* opcodes. Also I believe I saw them - commented out - 
in Entry.C just one or two versions ago, but that may be a hallucination.

So, I guess we remove those too, then...


Regards, 

	re

>I was referring to the manual from MIT which is the one by Barry.  
>It is that manual which lives on my shelf.
>  Yes it was there, but as the spectral opcodes do the same task
>better, it was removed.
>==John ffitch

  I ask for a little indulgence if it is wrong, but from
  the manual - help format for windows - I understand:

  - octdown converts a, or k, signals to d signals
  - noctdft converts d signals to w signals
  - spectral opcodes use the w signals 

  - octdown + noctdft are the only opcodes which produce
    w-signals and the spectral opcodes are the only which
    can use them (are specaddm, specfilt, specptrk, ecc. the
    spectral opcodes?).

  Such the spectral opcodes seem inserviceables without
  octdown and noctdft.

  I have ignored something?

  (I am interested in particular to specptrk, for check
  the base frequency of the samples).

  Thanks
  Silvano Zambon