- chuck-users - lists.cs.princeton.edu

CurveTable strange behaviour
by eduard 23 Aug '07

23 Aug '07

The following patch shows either a misbehaviour or a misuse of CurveTable, dunno. Two different functions (rand_coefs1 and rand_coefs2) are used to assign coefficients to CurveTable.coefs(). While the first seems to do the job alright, the latter doesn't and throws the following error message: [chuck](via CurveTable): usage: size, time1, value1, curvature1, [ timeN-1, valueN-1, curvatureN-1, ] timeN, valueN) Maybe I'm doing something wrong that | don't see, but can't understand why this happens. Could anyone explain? Also, don't understand why the error message tells me the size needs to be supplied. The examples say that coefs are arrays of the type [time1, value1, curvature1, ..., timeN-1, valueN-1, curvatureN-1, timeN, valueN] and Gen's size, if I'm not mistaken, is hard coded as 4096. Thanks in advance, eduard The patch: fun void rand_coefs1( CurveTable table ) { [-1, 0, 1 ] @=> int curvature[]; [0., 0., curvature[ Std.rand2(0, 2)], 1, Std.rand2f(.5,1.), curvature[ Std.rand2(0, 2)], 2, Std.rand2f(.5,1.), curvature[ Std.rand2(0, 2)], 3, Std.rand2f(0.,1.), curvature[ Std.rand2(0, 2)], 4, Std.rand2f(0.,1.), curvature[ Std.rand2(0, 2)], 5, Std.rand2f(0.,1.), curvature[ Std.rand2(0, 2)], 6, Std.rand2f(0.,1.), curvature[ Std.rand2(0, 2)], 7, Std.rand2f(0.,1.), curvature[ Std.rand2(0, 2)], 8, Std.rand2f(0.,1.), curvature[ Std.rand2(0, 2)], 9, 0] @=> float coefs[]; coefs => table.coefs; <<<"\n", "CURVE 1:\n" >>>; <<<"\t index", " value", " curvature">>>; <<<"\t----------------------------------","\t">>>; for( 0 => int i; i < 8; i++ ) <<<"\t",coefs[i++], coefs[i++], coefs[i]>>>; <<<"\t", coefs[8], coefs[9] >>>; <<<"\t----------------------------------","\t">>>; } fun void rand_coefs2( CurveTable table, int n_breakpoints ) { // this function does not assign values stored in coefs correctly to // CurveTable.coefs(). Why? It also throws an error message. [-1, 0, 1 ] @=> int curvature[]; n_breakpoints*3 - 1 => n_breakpoints; float coefs[ n_breakpoints]; 0. => coefs[0]; 0. => coefs[1]; for( 2 => int i; i < n_breakpoints; i++ ) { curvature[ Std.rand2(0, 2)] => coefs[i++]; i/3 => coefs[i++]; Std.rand2f(0.,1.) => coefs[i]; } 0 => coefs[n_breakpoints-1]; coefs => table.coefs; <<<"\n", "CURVE 2:\n" >>>; <<<"\t index", " value", " curvature">>>; <<<"\t----------------------------------","\t">>>; for( 0 => int i; i < n_breakpoints-2; i++ ) <<<"\t", coefs[i++], coefs[i++], coefs[i]>>>; <<< "\t", coefs[n_breakpoints-2], coefs[n_breakpoints-1] >>>; <<<"\t----------------------------------","\t">>>; } CurveTable curve1; CurveTable curve2; 1024 => int size; <<< "\n*Assigning coefs via rand_coefs1()","\t">>>; rand_coefs1( curve1 ); 0 => float sum; for( 0 => int i; i < size; i++ ) { curve1.lookup(i) +=> sum; } if( sum > 0 ) <<<"Ok", "sum[i=0,N=1023] =", sum>>>; else <<< "Error:", "sum[i=0,N=1023] =", sum, ". Coefs were not assigned. All values are 0." >>>; <<< "\n*Assigning coefs via rand_coefs2()","\t">>>; rand_coefs2( curve2, 10 ); 0 => sum; for( 0 => int i; i < size; i++ ) { curve2.lookup(i) +=> sum; } if( sum > 0 ) <<<"Ok", "sum[i=0,N=1023] =", sum>>>; else <<< "Error:", "sum[i=0,N=1023] =", sum, ". Coefs were not assigned. All values are 0.\n">>>;

2 1

A example that needs correction
by Martin Ahnelöv 19 Aug '07

19 Aug '07

Hi! I was investigating what I could use the spork-keyword for when I stumbled on this problem in the examples: The example fm3.ck is pretty wrong. It says that fm-synthesis is at sync 0 on SinOsc's, but that isn't true. sync 0 is sync frequency. fm2.ck does it right however. Experience the difference yourself by changing both fm2.ck and fm3.ck's c.freq-variable. Anyways, this is the result of my tries: http://hauntedhouse.wordpress.com/2007/08/18/ambientic-monotonic-pieceus/ I'll probably go on and work a little more with it. Gasten

3 3

Re: [chuck-users] Lisa behaving strangely?
by Kassen 17 Aug '07

17 Aug '07

First of all, I seem to have inadvertedly send the message below to the dev-list which wasn't my intention. Oops. Secondly, sorry to reply to my own message and breach nettiquette but I found a second issue when trying to work around the one posted before. For some reason LiSa.duration() is write only. This is inconvenient when trying to work around this bug as it makes it impossible to go like this; l.playPos( l.duration() - samp); I can't realy see why duration would be write only while I can think of reasons to want to check the max duration of a buffer (aside from the posted bug). Could a readable version of .duration perhaps be considdered for a future update? Yours, Kas. On 8/8/07, Kassen <signal.automatique(a)gmail.com> wrote: > > Hi list, Hi Dan! > > > I have a slight issue with LiSa when she's supposed to play backwards. > Let's assume for a second we want to record a single piano note, then play > it back *once* in reverse; > ---------------------- > Rhodey r => LiSa l => dac; > > second => l.duration; > > 1 => l.record; > 1 => r.noteOn; > > second => now; > > 0 => l.record; > > 0 => l.loop; > second => l.playPos; > -1 => l.rate; > 1 => l.play; > > second => now; > --------------- > > The above looks entirely correct to me, however, it won't generate a > sound. for > > .999::second => l.playPos; > > it will work just fine and it will also work fine when .loop is turned on > but turning loop on for one-shot samples isn't always convenient. > > I believe this to be a bug, probably related to the way in which LiSa > determines when she is supposed to stop playing. When you, Dan, have a > moment, could you please look into this? > > Yours, > Kas. >

4 9

one of the two sX.freq() in a same while doesn't seem to get changed
by 2g 10 Aug '07

10 Aug '07

today i learned about .5::second => dur T; T - (now % T) => now; and wanted to modify my isthisokay.ck to something more overtone sequencer oriented but nothing seemed to be working properly so comment outed a lot and went into the debug mode with the followings being chucked i can see things like s1 changed 34.747145 s2 changed 95.554650 s1 changed 56.208618 s2 changed 140.521544 s1 changed 32.703196 s2 changed 89.933788 s1 changed 45.988869 s2 changed 160.961041 --- get displayed but the sound itself doesn't seem to be doing this one of the two sinewave ... maybe s1.freq() seems to be playing the same long tone (freq never changing) and i don't seem to hear the rhythm pattern it should have created if am not wrong i think there is three sinewaves coming out isthisokay.ck: .5::second => dur T; T - (now % T) => now; SinOsc s1 => Gain g1 => dac; SinOsc s2 => g1; SinOsc s3 => PRCRev g3 => dac.left; SinOsc s4 => PRCRev g4 => dac.right; .25 => g1.gain; .25 => g3.gain; .25 => g4.gain; 27.5 * Math.pow(2.0,(3.0/12.0)) => float oilroot; 2 * oilroot => s2.freq; fun void ri() { while(true) { 300::ms => now; s2.freq() * Std.rand2(16, 32) / 16.0 => s3.freq; } } fun void ga() { while(true) { 330::ms => now; s2.freq() * Std.rand2(16, 32) / 16.0 => s4.freq; } } //spork ~ ri(); //spork ~ ga(); while( true ) { Std.rand2(32, 64) * oilroot / 32.0 => s1.freq; <<<"s1 changed", s1.freq()>>>; //1000::ms => now; 7::T => now; s1.freq() * Std.rand2(8, 16) / 4.0 => s2.freq; <<<"s2 changed", s2.freq()>>>; //250::ms => now; 1::T => now; } -- 2g http://micro.ispretty.com

2 3

midi channels
by mike clemow 10 Aug '07

10 Aug '07

Hello, Is it possible to specify a MIDI channel to send information on? I can't find any information on channels in the manual. Cheers, Mike -- http://shadowofaculture.blogspot.com

2 3

thangs on linux ppc
by 2g 09 Aug '07

09 Aug '07

experiencing that UB version (i don't know why but) works on osx10.3.9 i now understand that chuck is really what i wanted(RT-dsp@command-line) so wanted to go more CUI(VT) oriented and moved to linuxes on other volumes slackintosh was the only one i had to make(linux-alsa) by myself others got there by each apt oriented ways these are maybe cuz i didn't run jackd things before running chuck(is it?) but the result of each chuck moe.ck became as followings chuck on slack is alsa version but i don't know how the bin of debian&ubuntu got compiled (oh unstable is saying something related to jack..) any suggestions? tia slackintosh12.0: [chuck]: (via rtaudio): no devices found for given stream parameters: ... RtApiAlsa: error setting sample rate (48000) on device (hw:Snapper,0): Invalid argument. [chuck]: cannot initialize audio device (try using --silent/-s) debian-ppc-stable: [chuck]: (via rtaudio): no devices found for compiled audio APIs! [chuck]: cannot initialize audio device (try using --silent/-s) debian-ppc-unstable: JACK tmpdir identified as [/dev/shm] [chuck]: (via rtaudio): no devices found for compiled audio APIs! [chuck]: cannot initialize audio device (try using --silent/-s) ubuntu-ppc-7.04: [chuck]: (via rtaudio): no devices found for compiled audio APIs! [chuck]: cannot initialize audio device (try using --silent/-s) -- 2g http://micro.ispretty.com

3 6

order aspect when fun sharing things in main
by 2g 09 Aug '07

09 Aug '07

thanks to things taught via this list am now with isthisokay.ck like this: SinOsc s1 => Gain g1 => dac; SinOsc s2 => g1; SinOsc s3 => PRCRev g3 => dac.left; SinOsc s4 => PRCRev g4 => dac.right; .25 => g1.gain; .25 => g3.gain; .25 => g4.gain; fun void ri() { while(true) { 300::ms => now; s2.freq() * Std.rand2(16, 32) / 16.0 => s3.freq; } } fun void ga() { while(true) { 330::ms => now; s2.freq() * Std.rand2(16, 32) / 16.0 => s4.freq; } } //let's try sporking it spork ~ ri(); spork ~ ga(); while( true ) { Std.rand2(32, 64) * 27.5 / 32.0 => s1.freq; 1000::ms => now; s1.freq() * Std.rand2(8, 16) / 4.0 => s2.freq; 250::ms => now; } am happy with the result it gives but began to wonder why s2.freq() in the fun doesn't result into an error when activated funs are sporked before the main loops starts ... so s2.freq() might have nothing when funs want to claculate things maybe SinOsc s3 => PRCRev g3 => dac.left; SinOsc s4 => PRCRev g4 => dac.right; gives some sort of default for .freq? then at the beginning am having s3,s4 making non-overtone but random note that has nothing to do with 27.5hz? tia -- 2g http://micro.ispretty.com

2 1

NullPointerException
by eduard aylon 07 Aug '07

07 Aug '07

Hi list, I am trying to make a list of objects where each object has references to other objects, but I am not succeeding and I get a NullPointerException. The piece of code below may represent what I am trying to do. class object { Std.rand2f(0., 1.) => float f_; } class Item extends object { 2 => int size_; object item_[size_]; fun void print() { for( 0=>int i; i<size_; i++ ) <<< "item [",i,"]", item_[i].f_>>>; } <<<"----------------------------------------------">>>; <<<"calling print from within class definition">>>; <<<"----------------------------------------------">>>; print(); } 4 => int n; Item l[n]; // no problems up until here <<<"----------------------------------------------">>>; <<<"calling print from outside class definition">>>; <<<"----------------------------------------------">>>; for( 0=>int i; i<n; i++ ) l[i].print(); //offending line After execution I get: [chuck](VM): NullPointerException: shred[id=1:test.ck], PC=[37] Excuse me if it's too obvious, but I can't see why it rises the null pointer exception. Any ideas? thanks in advance, eduard

5 10

Re: [chuck-users] can there be many ::ms in while?
by Stephen Ball 07 Aug '07

07 Aug '07

just to chime in here - i'm VERY new to programming. how do you create this kind of concurrency? if, say, you wanted to create two while loops, but wanted them to run "side-by-side", but looping at different speeds, how would you do it?for example:while (true){ do something; 100::ms => now;}andwhile (true) {do something else; 150::ms => now;}cheers> Date: Fri, 3 Aug 2007 05:53:57 +0900> From: electriclightheads(a)gmail.com> To: chuck-users(a)lists.cs.princeton.edu> Subject: Re: [chuck-users] can there be many ::ms in while?> > oh> maybe in the second way> am just looping the 1st while only?> > maybe it was a bit too early for me to ask things on the list> (only i can do is modifying the tut)> i thought> 100::ms => now;> was something like "100msec later do ..." or "every 100msec do ..."> but i happened to find that in () expression of while> > mmm> i'll go through a bit more ahead in the manual> yes> then i'll find Blit-UGen waiting for me?> > On 8/2/07, Martin Ahnelöv <operagasten(a)gmail.com> wrote:> > tor 2007-08-02 klockan 07:34 +0900 skrev 2g:> > > have just started coding on chuck> > > was looking for way to let s1 have 3 overtones> > > flying around him> > > i know it should be something other than rand2f to handle integers> > > but my q is about ::ms in while> > >> > > i wonder which one of these are the right way to use them> > > i thought the latter might be safe but it seems to give one never> > > changing long note> > > and maybe only one another is floating around> > > but none of them give me an error> > >> > >> > > style one:> > >> > > SinOsc s1 => PRCRev g => dac;> > > SinOsc s2 => g;> > > SinOsc s3 => g;> > > SinOsc s4 => g;> > > .25 => g.gain => g.mix;> > > while( true ) {> > > 500::ms => now;> > > Std.rand2f(30.0, 220.0) => s1.freq;> > > 170::ms => now;> > > s1.freq() * Std.rand2f(8.0, 16.0) / 8.0 => s2.freq;> > > 75::ms => now;> > > s2.freq() * Std.rand2f(16.0, 32.0) / 16.0 => s3.freq;> > > 111::ms => now;> > > s2.freq() * Std.rand2f(16.0, 32.0) / 16.0 => s4.freq;> > > }> > >> > >> > > style two:> > >> > > SinOsc s1 => PRCRev g => dac;> > > SinOsc s2 => g;> > > SinOsc s3 => g;> > > SinOsc s4 => g;> > > .25 => g.gain => g.mix;> > > while( true ) {> > > 200::ms => now;> > > Std.rand2f(30.0, 220.0) => s1.freq;> > > }> > > while( true ) {> > > 170::ms => now;> > > s1.freq() * Std.rand2f(8, 16) / 4.0 => s2.freq;> > > }> > > while( true ) {> > > 75::ms => now;> > > s2.freq() * Std.rand2f(16, 32) / 8.0 => s3.freq;> > > }> > > while( true ) {> > > 111::ms => now;> > > s2.freq() * Std.rand2f(16, 32) / 8.0 => s4.freq;> > > }> > >> >> > Okey, I'm gonna teach you something about while-loops. A while loop will> > do whatever is in the brackets behind the statements until the> > expression inside the parenthesis is false. When it is false, it> > continues to the next thing after the close bracket (in your case the> > next while-loop).> >> > illustrated with chuck-code:> >> > 1 => int i;> >> > while (i < 10) {> > <<<i>>>; //print i> > i + 1 => i;> > }> > <<<"done">>>;> >> > Notice how it prints the numbers 1 to 9, and finally "done"?> >> > Now, look at your second example. Does the expression never return> > false? No, because you have hard coded "true" inside it.> >> > So, your first one should work as you want it, and your second won't.> > (Yes, you can advance time how many times as you want in whatever> > code-context you want)> >> > By the way, have you checked out the Blit-UGen? I think that one might> > satisfy your needs.> >> >> > Hope that helps,> > Gasten> >> > _______________________________________________> > chuck-users mailing list> > chuck-users(a)lists.cs.princeton.edu> > https://lists.cs.princeton.edu/mailman/listinfo/chuck-users> >> > > -- > 2g> http://micro.ispretty.com> _______________________________________________> chuck-users mailing list> chuck-users(a)lists.cs.princeton.edu> https://lists.cs.princeton.edu/mailman/listinfo/chuck-users _________________________________________________________________ Try Live.com - your fast, personalised homepage with all the things you care about in one place. http://www.live.com/?mkt=en-gb

7 15

Bug; Delay floating point denormalisation
by Kassen 07 Aug '07

07 Aug '07

Fellow Chuckists, "Delay" definately has a floatingpoint denormalisation issue (well, either Delay or Envelope). My feedbackloop went something like this; soundsource => Delay del => dac; del => Envelope feedback => del: .9 => feedback.gain; In this form it has to loop for quite a while before the issue shows up in cpu usage and it will probably be most evident in situations where other Ugens are also using floatingpoint opperations. A smaller value then .9 for the gain should speed up the phenomenon. In my case my "soundsource" had filled the Delay with sound, then went quiet, after some time the CPU meter rose fairly sharply, because of the controll using HID that I had over the feedback envelope I could make the CPU usage go down again with one switch, meaning I'm absolutely positive it was this loop that was to blame. This is on Windows, using the unnoficial ASIO build that Philippe posted, delay time was around one second or so, neither of which should matter that much. Kas.

2 2