[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [sc-users] env ugen(s) don't function correctly??

I believe, your / our problem appears because of the moment the end level of the envelope is reached, together with the doneAction.(your end level of 0.001 is pretty high still) Compare: 

SynthDef(\bell1, { arg note = 300, dur = 1.4, vol = 0.5, doneAction = 2;
	var sound, input, env, together;
	input = BrownNoise.ar(0.5);
sound = Klank.ar(`[[note, note*3/2, note*8/5, note*18/8], [0.01, 0.02, 0.03, 
	0.04], [dur, 1.2, 1.3, 1.4]], input).dup;
	env =

	EnvGen.kr(/*perc*/Env.new([0.001,vol,0.001],[0.001,dur],'exponential'),
	doneAction: doneAction);

	together = sound*env;
	Out.ar(0, together);
}).send(s);


Synth(\bell1, [\dur, 3, \doneAction, 2]);
Synth(\bell1, [\dur, 3, \doneAction, 0])


// exponential with audio rate

SynthDef(\bell1, { arg note = 300, dur = 1.4, vol = 0.5, doneAction = 2;
	var sound, input, env, together;
	input = BrownNoise.ar(0.5);
sound = Klank.ar(`[[note, note*3/2, note*8/5, note*18/8], [0.01, 0.02, 0.03, 
	0.04], [dur, 1.2, 1.3, 1.4]], input).dup;
	env =

	EnvGen.ar(/*perc*/Env.new([0.001,vol,0.001],[0.001,dur],'exponential'),
	doneAction: doneAction);

	together = sound*env;
	Out.ar(0, together);
}).send(s);


Synth(\bell1, [\dur, 3, \doneAction, 2]);
Synth(\bell1, [\dur, 3, \doneAction, 0])




Sorry to hound, but I've been thinking about this a bit more.

I think there may be a misunderstanding because of your comment after the
test you suggested.

I find the beginnings of the envelopes to be quite good, the attack portion
always sounds proportioned nicely no matter which curve, preset or integer,
that I plug in. In almost all cases, the decay portion does not decay all of
the way to silence as .plot shows or as is programed in the code.

This does still occur with sine waves. I actually even hear it when using
.test so I'm convinced that it's not my synthesis method as you suggested,
esp. since the attack portions are fine and immediate (or only take 0.001
seconds, which is immediate enough for me).

Also, FWIW, I went and got the 3.1.1 from November on the sourceforge site
today, so I believe I'm all the way up to date on builds - the problem
remains.

Thanks again,
stwbass


Josh Parmenter wrote:


 Here is a better test... do you still hear the problem?


 SynthDef(\convolPulse, {arg note = 350, dur = 1.4, vol = 0.5, out = 0;
 var in, kernel, sound, env, together, pan;
	env = EnvGen.kr(Env.sine(dur, 1), doneAction: 2);
	together = SinOsc.ar(note, 0, vol);
	pan = Pan2.ar(together * env, out);
 Out.ar(0, pan);
 }).send(s);


 //the chord:

 (
 Synth(\convolPulse,[\note, 80.midicps, \dur,3]);
 Synth(\convolPulse,[\note, 66.midicps, \dur,3]);
 Synth(\convolPulse,[\note, 53.midicps, \dur,3]);
 Synth(\convolPulse,[\note, 45.midicps, \dur,3]);
 )
If so, what system and version of SC are you using? I imagine the difference in attack and decay that you are hearing has to do with your synthesis technique taking some time to get started, and the cutoff sounds faster then the attack. If the above gives you a problem, then perhaps there is a bug, but we need to know your version, etc. 

 Best,

 Josh

 On Dec 22, 2007, at 1:05 AM, stwbass wrote:



 hello list-
I've been having some interesting problems with the Env ugen and some of 
 it's presets...
The presets I've had the most trouble with are .sine and .triangle. Here's 
 an example of those acting funny.

 //preset SynthDef:

 SynthDef(\convolPulse, {arg note = 350, dur = 1.4, vol = 0.5, out = 0;
 var in, kernel, sound, env, together, pan;
	in = Pulse.ar(note, 0.2, vol);
	kernel = WhiteNoise.ar(0.05);
	sound = Convolution.ar(in, kernel, 2048, 0.5);
	env = EnvGen.kr(Env.sine(dur, 1), doneAction: 2);
	together = RLPF.ar(sound*env, note*2.2);
	pan = Pan2.ar(together, out);
 Out.ar(0, pan);
 }).send(s);

 //or this one:

 SynthDef(\convolPulse, {arg note = 350, dur = 1.4, vol = 0.5, out = 0;

 >> var in, kernel, sound, env, together, pan;

	in = Pulse.ar(note, 0.2, vol);
	kernel = WhiteNoise.ar(0.05);
	sound = Convolution.ar(in, kernel, 2048, 0.5);
	env = EnvGen.kr(Env.triangle(dur, 1), doneAction: 2);
	together = RLPF.ar(sound*env, note*2.2);
	pan = Pan2.ar(together, out);
 Out.ar(0, pan);
 }).send(s);


 //the chord:

 (
 Synth(\convolPulse,[\note, 80.midicps, \dur,3]);
 Synth(\convolPulse,[\note, 66.midicps, \dur,3]);
 Synth(\convolPulse,[\note, 53.midicps, \dur,3]);
 Synth(\convolPulse,[\note, 45.midicps, \dur,3]);
 )
The tail end of the envelope doesn't taper correctly, it seems to get about 
 3/4 of the way towards the bottom of the "curve" and the drops off
completely (at least on my end). The .triangle env seems much worse than 
 .sine, but they both sound like they have the same issue.

 I've also tried faking a new triangle env SynthDef below:

 //new SynthDef attempt:

 SynthDef(\convolPulse, {arg note = 350, dur = 1.4, vol = 0.5, out = 0;
 var in, kernel, sound, env, together, pan;
	in = Pulse.ar(note, 0.2, 0.5);
	kernel = WhiteNoise.ar(0.05);
	sound = Convolution.ar(in, kernel, 2048, 0.5);
	env =
 EnvGen.kr(/*tri*/Env.new([0.001,0.15*vol,vol,0.15*vol,0.001],
 [0.15*dur,0.35*dur,0.35*dur,0.15*dur],'exponential'),
 doneAction: 2);
	together = RLPF.ar(sound*env, note*2.2);
	pan = Pan2.ar(together, out);
 Out.ar(0, pan);
 }).send(s);
That one seems closer (the shape is a bit off), but the taper doesn't really 
 work either.
I've had similar problems in the past with Env.perc. SynthDefs and passage 
 below:

 // .perc SynthDefs

 SynthDef(\bell1, { arg note = 300, dur = 1.4, vol = 0.5;
 var sound, input, env, together;
 input = BrownNoise.ar(0.5);
sound = Klank.ar(`[[note, note*3/2, note*8/5, note*18/8], [0.01, 0.02, 0.03, 
 0.04], [dur, 1.2, 1.3, 1.4]], input).dup;
 env = EnvGen.kr(Env.perc(0.001, dur, vol), doneAction: 2);
 together = sound*env;
 Out.ar(0, together);
 }).send(s);

 SynthDef(\bell2, { arg note = 300, dur = 1.4, vol = 0.6;
 var sound, input, env, together;
 input = Impulse.ar(0.1, 0, 10);
sound = Klank.ar(`[[note, note*3/2, note*8/5, note*18/8], [0.02, 0.03, 0.04, 
 0.05], [dur, 1.2, 1.3, 1.4]], input).dup;
 env = EnvGen.kr(Env.perc(0.001,dur,vol), doneAction: 2);
 together = sound*env;
 Out.ar(0, together);
 }).send(s);

 //here's the passage:

 a = TempoClock(92/60);
 //m. 31-32
a.sched(1, {Synth(\bell1, [\note, 48.midicps, \dur, 1.3044, \vol, 0.2])}); 
 a.sched(1, {Synth(\bell2, [\note, 47.midicps])});
 a.sched(2, {Synth(\bell2, [\note, 56.midicps])});
 a.sched(2, {Synth(\bell2, [\note, 57.midicps])});
 a.sched(2.25, {Synth(\bell2, [\note, 81.midicps, \vol, 0.5])});
 a.sched(2.4375, {Synth(\bell2, [\note, 65.midicps, \vol, 0.5])});
 a.sched(2.625, {Synth(\bell2, [\note, 78.midicps, \vol, 0.5])});
 a.sched(2.8125, {Synth(\bell2, [\note, 80.midicps, \vol, 0.5])});
 a.sched(3, {Synth(\bell2, [\note, 49.midicps, \vol, 0.5])});
 a.sched(5, {
		Synth(\bell2, [\note, 50.midicps, \vol, 0.65]);
		Synth(\bell2, [\note, 51.midicps, \vol, 0.65]);
		Synth(\bell2, [\note, 54.midicps, \vol, 0.65]);
		Synth(\bell2, [\note, 61.midicps, \vol, 0.65]);
		Synth(\bell2, [\note, 64.midicps, \vol, 0.65]);
		Synth(\bell2, [\note, 65.midicps, \vol, 0.65]);
		});
 a.sched(5, {
		Synth(\bell1, [\note, 50.midicps, \dur, 6, \vol, 0.15]);
		Synth(\bell1, [\note, 51.midicps, \dur, 6, \vol, 0.15]);
		Synth(\bell1, [\note, 54.midicps, \dur, 6, \vol, 0.15]);
		Synth(\bell1, [\note, 61.midicps, \dur, 6, \vol, 0.15]);
		Synth(\bell1, [\note, 64.midicps, \dur, 6, \vol, 0.15]);
		Synth(\bell1, [\note, 65.midicps, \dur, 6, \vol, 0.15]);
		});
That one sounds fine, the very end still doesn't seem to taper correctly to me. Some other sounds in the piece the taper a bit more oddly; sorry about 
 not fining the best example for that one.
Again, I tried to make some perc-like envs... but I think the 
 >> preset .perc's
are better, though they still have issues. Even still, I don't understand 
 why these one's do not sound correct to me. Maybe there're some code
 suggestions for these?

 // perc-like envs - SynthDefs

 SynthDef(\bell1, { arg note = 300, dur = 1.4, vol = 0.5;
 var sound, input, env, together;
 input = BrownNoise.ar(0.5);
sound = Klank.ar(`[[note, note*3/2, note*8/5, note*18/8], [0.01, 0.02, 0.03, 
 0.04], [dur, 1.2, 1.3, 1.4]], input).dup;
 env =
 EnvGen.kr(/*perc*/Env.new([0.001,vol,0.001],
 [0.001,dur],'exponential'),
 doneAction: 2);
 together = sound*env;
 Out.ar(0, together);
 }).send(s);

 SynthDef(\bell2, { arg note = 300, dur = 1.4, vol = 0.6;
 var sound, input, env, together;
 input = Impulse.ar(0.1, 0, 10);
sound = Klank.ar(`[[note, note*3/2, note*8/5, note*18/8], [0.02, 0.03, 0.04, 
 0.05], [dur, 1.2, 1.3, 1.4]], input).dup;
 env =
 EnvGen.kr(/*perc*/Env.new([0.001,vol,0.001],
 [0.001,dur],'exponential'),
 doneAction: 2);
 together = sound*env;
 Out.ar(0, together);
 }).send(s);
The \bell2 sound doesn't have as much resonance as the \bell1 sound and I think that's why the "misshapen" envelope can pass for that sound; but I don't understand why the envelope has such a steep drop off at the end of 
 the \bell1 sound.

 In working on this piece I've become very confused by these envelope
tapering issues. I find that the attacks all sound nice, but the decays always have something funny. If it becomes neccessary (or would just help) I can post the whole piece and which events have the biggest problems to my 
 ears - maybe that will be the best way to diagnose the problem.

 Thanks in advance,
 stwbass


 --
 View this message in context:

http://www.nabble.com/env-ugen%28s%29-don%27t-function-correctly---tp14467510p14467510.html
 Sent from the Supercollider - User mailing list archive at Nabble.com.

 _______________________________________________
 sc-users mailing list
 sc-users@xxxxxxxxxxxxxxx
 http://www.create.ucsb.edu/mailman/listinfo/sc-users


 ******************************************
 /* Joshua D. Parmenter
 http://www.realizedsound.net/josh/
"Every composer - at all times and in all cases - gives his own interpretation of how modern society is structured: whether actively or passively, consciously or unconsciously, he makes choices in this regard. He may be conservative or he may subject himself to continual renewal; or he may strive for a revolutionary, historical or social palingenesis." - Luigi Nono 
 */


 _______________________________________________
 sc-users mailing list
 sc-users@xxxxxxxxxxxxxxx
 http://www.create.ucsb.edu/mailman/listinfo/sc-users




--
View this message in context: http://www.nabble.com/env-ugen%28s%29-don%27t-function-correctly---tp14467510p14477012.html 
Sent from the Supercollider - User mailing list archive at Nabble.com.


_______________________________________________
sc-users mailing list
sc-users@xxxxxxxxxxxxxxx
http://www.create.ucsb.edu/mailman/listinfo/sc-users



--





.