damped sine wave.

[ramp...@gmail.com]

 Hi.

   Thank you bery much for the library. it looks impresive!

   I have managed to include the midivox shield DAC (12 bit serial microchip MCP4921 ) to the library.

   i have forked on github if you are interested. (https://github.com/rampa069/Mozzi)

 Now i want to start building an analogue (DDS)  drumbox.

 I know i have samples. but i want to make it "virtual analogic" any idea on how to "damp" the sinewave?


 in my old code (generating the wave in ram in realtime) i have used this formula.

  sin(128*x)*exp(-2*x)

 but not sure on how to use it on mozzi.

[Mr Sensorium]

Hi,

great, I'm glad to hear about any forks and developments.

Would the Ead (Exponential Attack Decay) envelope generator do what you want? 
You can see it used in the envelopes>Ead_Envelope example.
It would probably sound better for drum sounds if it is used at audio rate. 

Let me know if you're looking for something else, or need a better explanation/example.

tim

[ramp...@gmail.com]

thanks.  i am getting som good results.

 in the example, i changed the noise to a sinewave and commented the setPhase line. now i get wath i want one of each 2 Beats :-O

 tried some values in the setPhase, but when i use this, the gain is not changed. any clue?

[ramp...@gmail.com]

 ok. changed the setPhase to 0 and it works. but not sure if the frequency is OK. it sounds very high to me.

 2 oscs 20+49Hz for a bassdrum sample sketch.


/*  Example playing an enveloped noise source
 *  using Mozzi sonification library.
 *
 *  Demonstrates Ead (exponential attack decay).
 *
 *  Circuit: Audio output on digital pin 9 (on a Uno or similar), or
 *  check the README or http://sensorium.github.com/Mozzi/
 *
 *  Mozzi help/discussion/announcements:
 *  https://groups.google.com/forum/#!forum/mozzi-users
 *
 *  Tim Barrass 2012.
 *  This example code is in the public domain.
 */

#include <MozziGuts.h>
#include <Oscil.h> // oscillator template
#include <tables/sin1024_int8.h> // sine table for oscillator
#include <Ead.h> // exponential attack decay
#include <EventDelay.h>


#define CONTROL_RATE 256 // powers of 2 please

Oscil <SIN1024_NUM_CELLS, AUDIO_RATE> aSin(SIN1024_DATA);
Oscil <SIN1024_NUM_CELLS, AUDIO_RATE> bSin(SIN1024_DATA);
EventDelay <CONTROL_RATE>  kDelay; // for triggering envelope start
Ead kEnvelope(CONTROL_RATE); // resolution will be CONTROL_RATE

int gain;



void setup(){
  startMozzi(CONTROL_RATE);
  aSin.setFreq(49);
  bSin.setFreq(20);

  kDelay.start(1000);
}


void updateControl(){
  // jump around in audio noise table to disrupt obvious looping
  //aNoise.setPhase(rand((uint)WHITENOISE8192_NUM_CELLS));
  //aSin.setPhase(0);
  //bSin.setPhase(0);
 
  if(kDelay.ready()){
    aSin.setPhase(0);
    bSin.setPhase(0);
    unsigned int attack = 10;
    unsigned int decay = 400;
    unsigned int duration = attack +decay ;
    kEnvelope.start(attack,decay);
    kDelay.start(duration+100);
  }
  gain = (int) kEnvelope.next();
}


int updateAudio(){
  int out=aSin.next();
          +bSin.next();
         
  return (gain*out>>4);
}


void loop(){
  audioHook(); // required here
}

[Mr Sensorium]

Hi rampamac,
how about this.. the cast to long in updateAudio() is the main change.
...apart from the >>4 to >>8, but maybe you're using your own 12 bit dac code ... anyway it sounds and looks as expected in a spectrum plot here.. is it any different for you?
I'm guessing you're using setPhase so the sine waves both start at 0 on each beat, where they would have been constantly shifting before, as the oscillators are playing continuously.

#include <MozziGuts.h>
#include <Oscil.h> // oscillator template
#include <tables/sin1024_int8.h> // sine table for oscillator
#include <Ead.h> // exponential attack decay
#include <EventDelay.h>


#define CONTROL_RATE 256 // powers of 2 please

Oscil <SIN1024_NUM_CELLS, AUDIO_RATE> aSin(SIN1024_DATA);
Oscil <SIN1024_NUM_CELLS, AUDIO_RATE> bSin(SIN1024_DATA);
EventDelay <CONTROL_RATE>  kDelay; // for triggering envelope start
Ead kEnvelope(CONTROL_RATE); // resolution will be CONTROL_RATE

byte gain;

void setup(){
  startMozzi(CONTROL_RATE);
  aSin.setFreq(49);
  bSin.setFreq(20);

  kDelay.start(1000);
}


void updateControl(){
  // jump around in audio noise table to disrupt obvious looping
  //aNoise.setPhase(rand((uint)WHITENOISE8192_NUM_CELLS));
  //aSin.setPhase(0);
  //bSin.setPhase(0);
 
  if(kDelay.ready()){
    aSin.setPhase(0);
    bSin.setPhase(0);
    unsigned int attack = 10;
    unsigned int decay = 400;
    unsigned int duration = attack +decay ;
    kEnvelope.start(attack,decay);
    kDelay.start(duration+100);
  }

  gain = kEnvelope.next();

}


int updateAudio(){
  int out=aSin.next();
          +bSin.next();

  return ((long)out*gain)>>8;

[ramp...@gmail.com]

On Friday, 14 June 2013 15:00:59 UTC+2, Mr Sensorium wrote:

 better now. now. now another problem summing the wave.
 
 in my prevoous code (and in yours) in the updateAudio function, there was a semicolon in the addition to eliminate it. can you check wath happens if you sum the waves? in my side i only can get very low volume waves or distorted waves depending on the bit shifting. (in my case with 12 bit >>4 sounds distorted, >>5 sounds very low.
 

[Mr Sensorium]

Aha... sin1024_int8.h has reduced dynamic range - a left-over from something I should have removed -woops.  That explains the low levels.  But maybe it's also worth having a master level to adjust for particularities of the synthesis.

Here's another version ... it looks like an improvement could be to make a "drum" class, except classes introduce overhead in the audio calculations, reducing capability by about 10-20% in my experience.  That might be improved by having conditional audio updating for the class, depending on whether the envelope is below a threshold...just thoughts..

#include <MozziGuts.h>
#include <Oscil.h> // oscillator template

#include <tables/sin2048_int8.h> // sine table for oscillator

#include <Ead.h> // exponential attack decay

#include <EventDelay.h> // for metronome

#define CONTROL_RATE 256 // powers of 2 please

Oscil <SIN2048_NUM_CELLS, AUDIO_RATE> aSin(SIN2048_DATA);
Oscil <SIN2048_NUM_CELLS, AUDIO_RATE> bSin(SIN2048_DATA);
EventDelay <CONTROL_RATE>  kBeat; // for triggering envelope start
Ead kBassEnvelope(CONTROL_RATE); // resolution will be 1000ms/CONTROL_RATE

byte bass_env; // carries bass envelope signal from updateControl() to updateAudio()

void setup(){
  aSin.setFreq(49);
  bSin.setFreq(20);

  unsigned int attack = 10;
  unsigned int decay = 400;

  kBassEnvelope.set(attack,decay);

  kBeat.start(500); // beat rate approx 500ms ~ 120bpm
  startMozzi(CONTROL_RATE);
}


boolean metro(){
  if(kBeat.ready()){
    kBeat.start();
    return true;
  }
  else{
    return false;
  }
}


void bassStart(){
  // align sine waves
  aSin.setPhase(0);
  bSin.setPhase(0);
  // start envelope
  kBassEnvelope.start();
}


void bassNext(){
  bass_env = kBassEnvelope.next();
  if(bass_env<20) bass_env = 0; // gate off grainy tail on rendered audio
}


void updateControl(){
  if(metro()){
    bassStart();
  }
  bassNext();
}


int updateAudio(){
  // bass
  int bass=aSin.next()+bSin.next(); // 8+8=9 bits
  bass = ((long)bass*bass_env)>>8; // 9 bits
  // master
  byte master_volume = 215; // adjust for max level without overflow
  // 9+8=17 bits, -2 bits = 15bits (but less because master_vol is less than full 8 bits)
  int out = ((long)bass*master_volume)>>2; // for 14 bit output, note this would be >>4 for 12 bit output
  return out;

[Jean-Luc Deladrière]

Hi 

I cannot compile this line

EventDelay <CONTROL_RATE>  kBeat; // for triggering envelope start

I have corrected  it to  :

EventDelay  kBeat; // for triggering envelope start

to get it work 

Is this ok or am I missing something ?

[Tim Barrass]

Hi,
your correction is correct!
An example for EventDelay is accessable in Arduino examples menu, File>Examples>Mozzi>02.Control>EventDelay.

--
You received this message because you are subscribed to the Google Groups "Mozzi-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to mozzi-users...@googlegroups.com.
To post to this group, send email to mozzi...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/mozzi-users/0c2f188f-1e48-4de0-99c4-17fcef2e2784%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.