Hello everybody!
For a current project i want to setup a csound device that can analyse a broad spectrum of information from an incoming sound. I want to use these information to control sound transforming processes.
I know, that there are some possibilities in Csound for this, like RMS, central spectroid and other stuff.
I guess, a lot of you already done something like this and i wanted to ask for examples or someone has collect all build in opcodes that can provide some information of an incoming sound?
All ideas are welcomed for this.
Hope to hear from you 
Greetings,
Philipp
Csound mailing list
Csound@listserv.heanet.ie
https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND
Send bugs reports to
GitHub · Where software is built
Discussions of bugs and features can be posted here
Oeyvind has lots of experience on this. He should have a thing or two to say.
Hi Philipp,
attached is a CSD I wrote a while ago. It uses the lufs opcode to implement an EBU-meter with audible output. Lufs has three outputs for momentary, integrated and short term loudness.
I suppose a virtual analogue vocoder approach might also help in combination, i.e. isolating only certain frequency bands. If you work on monophonic signals, you could perhaps use a pitchtracking opcode again to control the frequency bands for analysis. I found that compression before pitchtracking may help, don't forget the extra latency introduced by compress or compress2 through their look-ahead time.
In the back of my mind there is something about converting allpass filter coefficients to classic centre frequencies and bandwidhts, but I can't remember the opcode or formula. I looked into that to achieve a kind of formant analysis in combination with lpc. You certainly have the new streaming lpc opcodes.
Hope any of that hels.
Best wishes,
Jeanette
ebu.csd (3.49 KB)