[Csnd] Signal Flow Graph Opcodes, Initialisation and Sean Costellos Reverb

Date	2011-03-31 11:21
From	Jan Jacob Hofmann
Subject	[Csnd] Signal Flow Graph Opcodes, Initialisation and Sean Costellos Reverb
	Dear list, in a setup where audio signals and i-rate signals are passed with Signal Flow Graph Opcodes I do experience clicks in a Sean Costellos reverb at the start of every note. The clicks will not be in the source-instrument, if the reverb is used there for test reasons, but at initialisation time of every new note in the sink instrument, where the reverb actually has to reside. So it looks like the clicks appear in connection to the use of alwayson within the instrument and the reinitialisation. There is an easy way to solve these clicks: Just set the iskip- argument at the reverbsc opcode to "1" and everything is fine. To make it a bit more complicated I had to modify that Sean Costello reverb by using the version which is a full blown instrument code. It works basically similar, but can be changed and modified in regard of the personal needs. See: http://www.adp-gmbh.ch/csound/effects/fdn_reverberator.html The only drawback: there is no "iskip" argument. So I do get clicks at init time here, unless I do not find a way to solve that problem also for that kind of reverb. I tried to put to skip the whole content of the instrument by using rigoto, but this had no effect. one possible solution might be to look at the source code of that opcode to see, which operations are affected by that "iskip" argument and to apply the same to the code of the reverb instrument, if possible. Unfortunately I cannot read C-Language. Or maybe someone has got a different idea? Putting alwayson in the header-section of the .csd would solve the problem too, but then I could not pass along other i-rate statements to the sink instruments. I would like to be able to do so, as I would like to produce granular clouds with overlapping notes using these instruments. Any suggestions welcome, test .orc can be found below. Best regards, Jan Jacob sound \| movement \| object \| space sonic architecture \| site: http://www.sonicarchitecture.de spatial electronic composition \| 2nd order ambisonic music -n connect "1", "audio_out", "2", "audio_in" ;connecting a-rate output from instr 1 to instr 2 ;alwayson 2 giprint init 0 instr 1 ; source instrument iskip init 0 ; set to 0: no clicks in test-setting of instr 1 itest init 0 ; set to "1" to play only instr 1 for test output: no clicks giprint = giprint + 1 asig oscil 0dbfs * 0.2, p4, 1 ; creating an audio signal kdeclik linseg 0, p3 * 0.01, 1 , p3 - (p3 * 0.02) , 1, p3 * 0.01, 0 ; generate declick envelope asig = asig * kdeclik ; apply declick if itest == 0 goto sinkinstrumentinvocation asigL, asigR reverbsc asig, asig, 0.7, 18000, sr, 1, iskip ; reverb effect asig = asigL + asigR ; sum both channels outc asig print giprint ; output the sound goto endtest sinkinstrumentinvocation: outleta "audio_out", asig ; sending the audio signal to channel "audio_out" alwayson 2, giprint ; starting instr 2 to play simultaneously - comment this out instead of the line above endtest: endin instr 2 ; sink instrument iskip init 0 ; set to 1 to skip initialisation: the clicks should vanish then giprint = p4 asig inleta "audio_in" ; reading channel "audio_in" asigL, asigR reverbsc asig, asig, 0.7, 18000, sr, 1, iskip ; reverb effect asig = asigL + asigR ; sum both channels outc asig ; output the sound print giprint endin f1 0 8192 10 1 /* sine */ i1 0 5 220 i1 1 4 440 i1 2 3 880 i1 3 2 1760 i1 4 1 3520 Send bugs reports to the Sourceforge bug tracker https://sourceforge.net/tracker/?group_id=81968&atid=564599 Discussions of bugs and features can be posted here To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"

Date	2011-03-31 17:42
From	Michael Gogins
Subject	Re: [Csnd] Signal Flow Graph Opcodes, Initialisation and Sean
	Why are you using the alwayson opcode within the scope of another instrument? Of course this will cause clicks. The idea is to turn the reverb instrument on only the instrument header. It will remain on during the entire performance. Every new instr 1 will automatically copy the connection to the reverb instrument and all will go well. You will need to extend performance with the s and e statements at the end of the score to prevent a click with a reverb tail at the end of the entire performance. I repeat, both the connections and the always on instruments should be defined only in the orchestra header. When the score creates new instrument instances, they will automatically inherit the declared connections. If this is not clear, let me know and I will rewrite your example so it works. Regards, Mike On Thu, Mar 31, 2011 at 6:21 AM, Jan Jacob Hofmann wrote: > Dear list, > > in a setup where audio signals and i-rate signals are passed with Signal > Flow Graph Opcodes I do experience clicks in a Sean Costellos reverb at the > start of every note. The clicks will not be in the source-instrument, if the > reverb is used there for test reasons, but at initialisation time of every > new note in the sink instrument, where the reverb actually has to reside. So > it looks like the clicks appear in connection to the use of alwayson within > the instrument and the reinitialisation. > > There is an easy way to solve these clicks: Just set the iskip-argument at > the reverbsc opcode to "1" and everything is fine. To make it a bit more > complicated I had to modify that Sean Costello reverb by using the version > which is a full blown instrument code. It works basically similar, but can > be changed and modified in regard of the personal needs. See: > > http://www.adp-gmbh.ch/csound/effects/fdn_reverberator.html > > The only drawback: there is no "iskip" argument. > > So I do get clicks at init time here, unless I do not find a way to solve > that problem also for that kind of reverb. I tried to put to skip the whole > content of the instrument by using rigoto, but this had no effect. > > one possible solution might be to look at the source code of that opcode to > see, which operations are affected by that "iskip" argument and to apply the > same to the code of the reverb instrument, if possible. Unfortunately I > cannot read C-Language. > > Or maybe someone has got a different idea? > > Putting alwayson in the header-section of the .csd would solve the problem > too, but then I could not pass along other i-rate statements to the sink > instruments. I would like to be able to do so, as I would like to produce > granular clouds with overlapping notes using these instruments. > > Any suggestions welcome, test .orc can be found below. > > Best regards, > > Jan Jacob > > > > sound \| movement \| object \| > space > sonic architecture \| site: http://www.sonicarchitecture.de > spatial electronic composition \| 2nd order ambisonic music > > > > > > -n > > > > connect "1", "audio_out", "2", > "audio_in" ;connecting a-rate output from instr 1 to instr 2 > ;alwayson 2 > > giprint init 0 > instr 1 ; source instrument > iskip init 0 ; set to 0: > no clicks in test-setting of instr 1 > itest init 0 ; set to "1" > to play only instr 1 for test output: no clicks > giprint = giprint + 1 > asig oscil 0dbfs * 0.2, p4, 1 ; creating > an audio signal > kdeclik linseg 0, p3 * 0.01, 1 , p3 - (p3 * 0.02) , 1, p3 * > 0.01, 0 ; generate declick envelope > asig = asig * kdeclik ; apply > declick > > if itest == 0 goto > sinkinstrumentinvocation > asigL, asigR reverbsc asig, asig, 0.7, 18000, sr, 1, iskip > ; reverb effect > asig = asigL + asigR ; sum both > channels > outc asig > print giprint > ; output the sound > goto endtest > sinkinstrumentinvocation: > > > outleta "audio_out", asig ; sending the > audio signal to channel "audio_out" > alwayson 2, giprint ; starting > instr 2 to play simultaneously - comment this out instead of the line above > endtest: > endin > > > > > > instr 2 ; sink instrument > > iskip init 0 ; set to 1 > to skip initialisation: the clicks should vanish then > giprint = p4 > asig inleta "audio_in" ; reading > channel "audio_in" > asigL, asigR reverbsc asig, asig, 0.7, 18000, sr, 1, iskip > ; reverb effect > asig = asigL + asigR ; sum both > channels > outc asig ; output the > sound > print giprint > endin > > > > > > > f1 0 8192 10 1 /* sine > */ > i1 0 5 220 > i1 1 4 440 > i1 2 3 880 > i1 3 2 1760 > i1 4 1 3520 > > > > > > > > > > Send bugs reports to the Sourceforge bug tracker > https://sourceforge.net/tracker/?group_id=81968&atid=564599 > Discussions of bugs and features can be posted here > To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe > csound" > > -- Michael Gogins Irreducible Productions http://www.michael-gogins.com Michael dot Gogins at gmail dot com Send bugs reports to the Sourceforge bug tracker https://sourceforge.net/tracker/?group_id=81968&atid=564599 Discussions of bugs and features can be posted here To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"

Date	2011-03-31 17:52
From	Michael Gogins
Subject	Re: [Csnd] Signal Flow Graph Opcodes, Initialisation and Sean
	I said instrument header, I meant orchestra header... On Thu, Mar 31, 2011 at 12:42 PM, Michael Gogins wrote: > Why are you using the alwayson opcode within the scope of another > instrument? Of course this will cause clicks. The idea is to turn the > reverb instrument on only the instrument header. It will remain on > during the entire performance. Every new instr 1 will automatically > copy the connection to the reverb instrument and all will go well. You > will need to extend performance with the s and e statements at the end > of the score to prevent a click with a reverb tail at the end of the > entire performance. > > I repeat, both the connections and the always on instruments should be > defined only in the orchestra header. When the score creates new > instrument instances, they will automatically inherit the declared > connections. > > If this is not clear, let me know and I will rewrite your example so it works. > > Regards, > Mike > > On Thu, Mar 31, 2011 at 6:21 AM, Jan Jacob Hofmann > wrote: >> Dear list, >> >> in a setup where audio signals and i-rate signals are passed with Signal >> Flow Graph Opcodes I do experience clicks in a Sean Costellos reverb at the >> start of every note. The clicks will not be in the source-instrument, if the >> reverb is used there for test reasons, but at initialisation time of every >> new note in the sink instrument, where the reverb actually has to reside. So >> it looks like the clicks appear in connection to the use of alwayson within >> the instrument and the reinitialisation. >> >> There is an easy way to solve these clicks: Just set the iskip-argument at >> the reverbsc opcode to "1" and everything is fine. To make it a bit more >> complicated I had to modify that Sean Costello reverb by using the version >> which is a full blown instrument code. It works basically similar, but can >> be changed and modified in regard of the personal needs. See: >> >> http://www.adp-gmbh.ch/csound/effects/fdn_reverberator.html >> >> The only drawback: there is no "iskip" argument. >> >> So I do get clicks at init time here, unless I do not find a way to solve >> that problem also for that kind of reverb. I tried to put to skip the whole >> content of the instrument by using rigoto, but this had no effect. >> >> one possible solution might be to look at the source code of that opcode to >> see, which operations are affected by that "iskip" argument and to apply the >> same to the code of the reverb instrument, if possible. Unfortunately I >> cannot read C-Language. >> >> Or maybe someone has got a different idea? >> >> Putting alwayson in the header-section of the .csd would solve the problem >> too, but then I could not pass along other i-rate statements to the sink >> instruments. I would like to be able to do so, as I would like to produce >> granular clouds with overlapping notes using these instruments. >> >> Any suggestions welcome, test .orc can be found below. >> >> Best regards, >> >> Jan Jacob >> >> >> >> sound \| movement \| object \| >> space >> sonic architecture \| site: http://www.sonicarchitecture.de >> spatial electronic composition \| 2nd order ambisonic music >> >> >> >> >> >> -n >> >> >> >> connect "1", "audio_out", "2", >> "audio_in" ;connecting a-rate output from instr 1 to instr 2 >> ;alwayson 2 >> >> giprint init 0 >> instr 1 ; source instrument >> iskip init 0 ; set to 0: >> no clicks in test-setting of instr 1 >> itest init 0 ; set to "1" >> to play only instr 1 for test output: no clicks >> giprint = giprint + 1 >> asig oscil 0dbfs * 0.2, p4, 1 ; creating >> an audio signal >> kdeclik linseg 0, p3 * 0.01, 1 , p3 - (p3 * 0.02) , 1, p3 * >> 0.01, 0 ; generate declick envelope >> asig = asig * kdeclik ; apply >> declick >> >> if itest == 0 goto >> sinkinstrumentinvocation >> asigL, asigR reverbsc asig, asig, 0.7, 18000, sr, 1, iskip >> ; reverb effect >> asig = asigL + asigR ; sum both >> channels >> outc asig >> print giprint >> ; output the sound >> goto endtest >> sinkinstrumentinvocation: >> >> >> outleta "audio_out", asig ; sending the >> audio signal to channel "audio_out" >> alwayson 2, giprint ; starting >> instr 2 to play simultaneously - comment this out instead of the line above >> endtest: >> endin >> >> >> >> >> >> instr 2 ; sink instrument >> >> iskip init 0 ; set to 1 >> to skip initialisation: the clicks should vanish then >> giprint = p4 >> asig inleta "audio_in" ; reading >> channel "audio_in" >> asigL, asigR reverbsc asig, asig, 0.7, 18000, sr, 1, iskip >> ; reverb effect >> asig = asigL + asigR ; sum both >> channels >> outc asig ; output the >> sound >> print giprint >> endin >> >> >> >> >> >> >> f1 0 8192 10 1 /* sine >> */ >> i1 0 5 220 >> i1 1 4 440 >> i1 2 3 880 >> i1 3 2 1760 >> i1 4 1 3520 >> >> >> >> >> >> >> >> >> >> Send bugs reports to the Sourceforge bug tracker >> https://sourceforge.net/tracker/?group_id=81968&atid=564599 >> Discussions of bugs and features can be posted here >> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe >> csound" >> >> > > > > -- > Michael Gogins > Irreducible Productions > http://www.michael-gogins.com > Michael dot Gogins at gmail dot com > > > Send bugs reports to the Sourceforge bug tracker > https://sourceforge.net/tracker/?group_id=81968&atid=564599 > Discussions of bugs and features can be posted here > To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound" > > -- Michael Gogins Irreducible Productions http://www.michael-gogins.com Michael dot Gogins at gmail dot com Send bugs reports to the Sourceforge bug tracker https://sourceforge.net/tracker/?group_id=81968&atid=564599 Discussions of bugs and features can be posted here To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"