hi all,

my csd used to work for a while on OS X 10.11.1. installed yesterday.I needed to install liblo for el capitan, first for my own sender to get it even send something to oscdump. csound worked well wit the new library right away. Then *probably* I made a git pull, and now csound is stuck, ctrl-c gives only interrupts, only way to get back to command line is kill -9 PID.

Tried the same on ubuntu 14.04 LTS, and the same thing happens with a new git pull. Here’s he csd

; this is a quick hack csd for getting something out of segm2cs
; (c) -m
; copyright copyleft


-d -odac -+rtaudio=portaudio


sr=44100
ksmps=64
nchnls = 2
0dbfs=1

gihandle OSCinit 7778 ;
ga1 init 0
gifnum init 0
gSfile init "osc-initated.wav"
gkvol init 0
gkbasefreq init 0
gkfreqshift init 0
gkpanstart init 0
gkpan init 0
gkmodul init 0
gkquit init 0
turnon 1
;turnon 98 
;for writing to file

instr 1
      imaxinstr = 20
      print gihandle
nxtmsg:
   ;sprintf(msg,"ffffffi %f %f %f %f %f %f %d",vol,basefreq,freqshift,panstart,pan,modul,0);
      kk OSClisten gihandle,"/img2snd/run","ffffffi",gkvol,gkbasefreq,gkfreqshift,gkpanstart,gkpan,gkmodul,gkquit
      gkvol = gkvol / imaxinstr
      schedkwhen kk,0,imaxinstr,2,0,0.5. ; 
      schedkwhen gkquit,0,3,3,0,1 ; quit
      kgoto nxtmsg
ex:
      prints "waitimg for osc"
endin

instr 3 
      ;turnoff 98
      exitnow
endin

instr 2
      ;asig1,asig2 inch 1,2
      istr = i(gkpanstart)
      iend = i(gkpan)
      kt line istr,0.5,iend
      kdeclick linseg 0,0.1,1,0.3,1,0.1,0
      kdeclick *= 10 ; some scaling here and there
      amod oscil kdeclick*gkvol,gkmodul,1
      a1 oscil kdeclick*gkvol,gkbasefreq,1
      a2 oscil kdeclick*gkvol,gkbasefreq+gkfreqshift,1
      a1 = a1*amod; simple ring modulator
      a2= a2*amod 
      out gkpanstart*a1,gkpan*a2
endin
instr 98
      prints    "write\n"
      a1,a2 monitor
      fout gSfile,14,a1,a2
endin

instr 99
      
      exitnow
endin


f1 0 8192 10 1 0.33 0.12 0.8
f0 3 ; even after 3 secs nothing
e




When this worked, the first osc-message started the loop, now all I get is the print gihandle=0.00 and waiting for osc.

The first time I tried on ubuntu, a complete freeze of the machine.

A bug in my csd or csound? Seems strange that this cud used to work for a while, or I compiled csound again.

tnx

-m

ps. The code that sends the osc-messages is an OpenCV foreground segmentation, based on the position, pose etc. a different set of parameters are sent. I’ll probably put this to github, when osc works again.

Also, when 0dbfs is defined as 1, are negative values (log10(< 1)) allowed?

same

Csound mailing list
Csound@listserv.heanet.ie
https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND
Send bugs reports to
        https://github.com/csound/csound/issues
Discussions of bugs and features can be posted here

There have been some new things added to OSC in the devel repo, so maybe that’s what
is causing your problem.

========================
Dr Victor Lazzarini
Dean of Arts, Celtic Studies and Philosophy,
Maynooth University,
Maynooth, Co Kildare, Ireland
Tel: 00 353 7086936
Fax: 00 353 1 7086952 

> On 29 Nov 2015, at 17:15, Matti Koskinen  wrote:
> 
> hi all,
> 
> my csd used to work for a while on OS X 10.11.1. installed yesterday.I needed to install liblo for el capitan, first for my own sender to get it even send something to oscdump. csound worked well wit the new library right away. Then *probably* I made a git pull, and now csound is stuck, ctrl-c gives only interrupts, only way to get back to command line is kill -9 PID.
> 
> Tried the same on ubuntu 14.04 LTS, and the same thing happens with a new git pull. Here’s he csd
> 
> ; this is a quick hack csd for getting something out of segm2cs
> ; (c) -m
> ; copyright copyleft
> 
> 
> -d -odac -+rtaudio=portaudio
> 
> 
> sr=44100
> ksmps=64
> nchnls = 2
> 0dbfs=1
> 
> gihandle OSCinit 7778 ;
> ga1 init 0
> gifnum init 0
> gSfile init "osc-initated.wav"
> gkvol init 0
> gkbasefreq init 0
> gkfreqshift init 0
> gkpanstart init 0
> gkpan init 0
> gkmodul init 0
> gkquit init 0
> turnon 1
> ;turnon 98 
> ;for writing to file
> 
> instr 1
>      imaxinstr = 20
>      print gihandle
> nxtmsg:
>   ;sprintf(msg,"ffffffi %f %f %f %f %f %f %d",vol,basefreq,freqshift,panstart,pan,modul,0);
>      kk OSClisten gihandle,"/img2snd/run","ffffffi",gkvol,gkbasefreq,gkfreqshift,gkpanstart,gkpan,gkmodul,gkquit
>      gkvol = gkvol / imaxinstr
>      schedkwhen kk,0,imaxinstr,2,0,0.5. ; 
>      schedkwhen gkquit,0,3,3,0,1 ; quit
>      kgoto nxtmsg
> ex:
>      prints "waitimg for osc"
> endin
> 
> instr 3 
>      ;turnoff 98
>      exitnow
> endin
> 
> instr 2
>      ;asig1,asig2 inch 1,2
>      istr = i(gkpanstart)
>      iend = i(gkpan)
>      kt line istr,0.5,iend
>      kdeclick linseg 0,0.1,1,0.3,1,0.1,0
>      kdeclick *= 10 ; some scaling here and there
>      amod oscil kdeclick*gkvol,gkmodul,1
>      a1 oscil kdeclick*gkvol,gkbasefreq,1
>      a2 oscil kdeclick*gkvol,gkbasefreq+gkfreqshift,1
>      a1 = a1*amod; simple ring modulator
>      a2= a2*amod 
>      out gkpanstart*a1,gkpan*a2
> endin
> instr 98
>      prints    "write\n"
>      a1,a2 monitor
>      fout gSfile,14,a1,a2
> endin
> 
> instr 99
> 
>      exitnow
> endin
> 
> 
> f1 0 8192 10 1 0.33 0.12 0.8
> f0 3 ; even after 3 secs nothing
> e
> 
> 
> 
> 
> When this worked, the first osc-message started the loop, now all I get is the print gihandle=0.00 and waiting for osc.
> 
> The first time I tried on ubuntu, a complete freeze of the machine.
> 
> A bug in my csd or csound? Seems strange that this cud used to work for a while, or I compiled csound again.
> 
> tnx
> 
> -m
> 
> ps. The code that sends the osc-messages is an OpenCV foreground segmentation, based on the position, pose etc. a different set of parameters are sent. I’ll probably put this to github, when osc works again.
> 
> Also, when 0dbfs is defined as 1, are negative values (log10(< 1)) allowed?
> 
> same
> 
> Csound mailing list
> Csound@listserv.heanet.ie
> https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND
> Send bugs reports to
>        https://github.com/csound/csound/issues
> Discussions of bugs and features can be posted here

Csound mailing list
Csound@listserv.heanet.ie
https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND
Send bugs reports to
        https://github.com/csound/csound/issues
Discussions of bugs and features can be posted here

> On 29 Nov 2015, at 7:34 PM, Victor Lazzarini  wrote:
> 
> There have been some new things added to OSC in the devel repo, so maybe that’s what
> is causing your problem.
> 
> ========================
> Dr Victor Lazzarini
> Dean of Arts, Celtic Studies and Philosophy,
> Maynooth University,
> Maynooth, Co Kildare, Ireland
> Tel: 00 353 7086936
> Fax: 00 353 1 7086952 


So the proof of pudding is to eat it:-) I downloaded 6.06, but why cmake can’t find e.g .fluidsynth or liblo -headers, although they are in /usr/local/include really visible? Libs are found but headers not. There are no options to set the headers, only libs. Or some -Dyabbada=
I’m so used to using cmake-gui, so real-unix-men write as little as possible :-) 

tnx

-m
 
Csound mailing list
Csound@listserv.heanet.ie
https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND
Send bugs reports to
        https://github.com/csound/csound/issues
Discussions of bugs and features can be posted here

eaders msy be in a -devel package


On Sun, 29 Nov 2015, Matti Koskinen wrote:

>> On 29 Nov 2015, at 7:34 PM, Victor Lazzarini  wrote:
>>
>> There have been some new things added to OSC in the devel repo, so maybe that’s what
>> is causing your problem.
>>
>> ========================
>> Dr Victor Lazzarini
>> Dean of Arts, Celtic Studies and Philosophy,
>> Maynooth University,
>> Maynooth, Co Kildare, Ireland
>> Tel: 00 353 7086936
>> Fax: 00 353 1 7086952
>
>
> So the proof of pudding is to eat it:-) I downloaded 6.06, but why cmake can’t find e.g .fluidsynth or liblo -headers, although they are in /usr/local/include really visible? Libs are found but headers not. There are no options to set the headers, only libs. Or some -Dyabbada=
> I’m so used to using cmake-gui, so real-unix-men write as little as possible :-)
>
> tnx
>
> -m
>
> Csound mailing list
> Csound@listserv.heanet.ie
> https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND
> Send bugs reports to
>        https://github.com/csound/csound/issues
> Discussions of bugs and features can be posted here
>

Csound mailing list
Csound@listserv.heanet.ie
https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND
Send bugs reports to
        https://github.com/csound/csound/issues
Discussions of bugs and features can be posted here

Am I reading your code rigt?  Looks like a k-rate infinite 
loop in instr 1 
jumping back to nxtmsg, with no get-out so once started i1 will never end 
to give anyone else a chance.

Please explain I am wrong!

On Sun, 29 Nov 2015, Matti Koskinen wrote:

> hi all,
>
> my csd used to work for a while on OS X 10.11.1. installed yesterday.I needed to install liblo for el capitan, first for my own sender to get it even send something to oscdump. csound worked well wit the new library right away. Then *probably* I made a git pull, and now csound is stuck, ctrl-c gives only interrupts, only way to get back to command line is kill -9 PID.
>
> Tried the same on ubuntu 14.04 LTS, and the same thing happens with a new git pull. Here’s he csd
>
> ; this is a quick hack csd for getting something out of segm2cs
> ; (c) -m
> ; copyright copyleft
> 
> 
> -d -odac -+rtaudio=portaudio
> 
> 
> sr=44100
> ksmps=64
> nchnls = 2
> 0dbfs=1
>
> gihandle OSCinit 7778 ;
> ga1 init 0
> gifnum init 0
> gSfile init "osc-initated.wav"
> gkvol init 0
> gkbasefreq init 0
> gkfreqshift init 0
> gkpanstart init 0
> gkpan init 0
> gkmodul init 0
> gkquit init 0
> turnon 1
> ;turnon 98
> ;for writing to file
>
> instr 1
>      imaxinstr = 20
>      print gihandle
> nxtmsg:
>   ;sprintf(msg,"ffffffi %f %f %f %f %f %f %d",vol,basefreq,freqshift,panstart,pan,modul,0);
>      kk OSClisten gihandle,"/img2snd/run","ffffffi",gkvol,gkbasefreq,gkfreqshift,gkpanstart,gkpan,gkmodul,gkquit
>      gkvol = gkvol / imaxinstr
>      schedkwhen kk,0,imaxinstr,2,0,0.5. ;
>      schedkwhen gkquit,0,3,3,0,1 ; quit
>      kgoto nxtmsg
> ex:
>      prints "waitimg for osc"
> endin
>
> instr 3
>      ;turnoff 98
>      exitnow
> endin
>
> instr 2
>      ;asig1,asig2 inch 1,2
>      istr = i(gkpanstart)
>      iend = i(gkpan)
>      kt line istr,0.5,iend
>      kdeclick linseg 0,0.1,1,0.3,1,0.1,0
>      kdeclick *= 10 ; some scaling here and there
>      amod oscil kdeclick*gkvol,gkmodul,1
>      a1 oscil kdeclick*gkvol,gkbasefreq,1
>      a2 oscil kdeclick*gkvol,gkbasefreq+gkfreqshift,1
>      a1 = a1*amod; simple ring modulator
>      a2= a2*amod
>      out gkpanstart*a1,gkpan*a2
> endin
> instr 98
>      prints    "write\n"
>      a1,a2 monitor
>      fout gSfile,14,a1,a2
> endin
>
> instr 99
>
>      exitnow
> endin
> 
> 
> f1 0 8192 10 1 0.33 0.12 0.8
> f0 3 ; even after 3 secs nothing
> e
> 
> 
>
>
> When this worked, the first osc-message started the loop, now all I get is the print gihandle=0.00 and waiting for osc.
>
> The first time I tried on ubuntu, a complete freeze of the machine.
>
> A bug in my csd or csound? Seems strange that this cud used to work for a while, or I compiled csound again.
>
> tnx
>
> -m
>
> ps. The code that sends the osc-messages is an OpenCV foreground segmentation, based on the position, pose etc. a different set of parameters are sent. I’ll probably put this to github, when osc works again.
>
> Also, when 0dbfs is defined as 1, are negative values (log10(< 1)) allowed?
>
> same
>
> Csound mailing list
> Csound@listserv.heanet.ie
> https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND
> Send bugs reports to
>        https://github.com/csound/csound/issues
> Discussions of bugs and features can be posted here
>

Csound mailing list
Csound@listserv.heanet.ie
https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND
Send bugs reports to
        https://github.com/csound/csound/issues
Discussions of bugs and features can be posted here

> On 29 Nov 2015, at 9:23 PM, jpff  wrote:
> 
> Am I reading your code rigt?  Looks like a k-rate infinite loop in instr 1 jumping back to nxtmsg, with no get-out so once started i1 will never end to give anyone else a chance.
> 
> Please explain I am wrong!

It might be black magic, but this exact code worked earlier and it’s copied mainly from  the examples. But aren’t the schekwhens supposed to trigger other instruments  based on what is received in instr 1?  turnon  1 should start the i1. it used to be stuck on waiting for osc. Then after receiving the first osc-message, things kept running until the gkquit was 1 and instr 3 scheduled and it executed exitnow.

Now it’s stuck on waiting for osc, and ctrl-c doesn’t interrupt the execution, only kill -9 works.

But I’ll try 6.06, if I get osc included.

tnx

-m

Csound mailing list
Csound@listserv.heanet.ie
https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND
Send bugs reports to
        https://github.com/csound/csound/issues
Discussions of bugs and features can be posted here

> On 29 Nov 2015, at 10:12 PM, Justin Smith  wrote:
> 
> yes, that would normally be a conditional goto, only done if there was a message (to check if there is another after that in the same k cycle), it looks like here the intention is to infinitely loop one k cycle until there are no more messages in that cycle, then ask csound to exit?

my code sends quite a lot of osc-messages,, so I thought the messages would choke csound, but just sending  only one from oscsend had no influence either.

How to hack the cmakefiles to find osc includes and fluidsynth includes? I’m stuck on that compiling csound6.06 both on os x and linux.

tnx

-m

Csound mailing list
Csound@listserv.heanet.ie
https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND
Send bugs reports to
        https://github.com/csound/csound/issues
Discussions of bugs and features can be posted here

On Sun, 29 Nov 2015, Matti Koskinen wrote:

>> On 29 Nov 2015, at 10:12 PM, Justin Smith  wrote:
>>
>> yes, that would normally be a conditional goto, only done if there was 
>>a message (to check if there is another after that in the same k cycle), 
>>it looks like here the intention is to infinitely loop one k cycle until 
>>there are no more messages in that cycle, then ask csound to exit?

That is how I read it
>
> my code sends quite a lot of osc-messages,, so I thought the messages 
> would choke csound, but just sending  only one from oscsend had no 
> influence either.

because if there is no message you loop, as yoou dio if tere is one.  The 
effect is tat instr 1 never gets to the end of its k-cycle, and so csound 
is in a busy loop.
  >
> How to hack the cmakefiles to find osc includes and fluidsynth includes? 
> I’m stuck on that compiling csound6.06 both on os x and linux.
Which linux?  On Debian you need liblo-devel for OSC and 
libfluidsynth-devel for fluidsynth opcodes.

On OpenSuse tyou also need lliblo-devel and fluidsynth-devel

No idea for OSX

>
> tnx
>
> -m
>
> Csound mailing list
> Csound@listserv.heanet.ie
> https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND
> Send bugs reports to
>        https://github.com/csound/csound/issues
> Discussions of bugs and features can be posted here
>

Csound mailing list
Csound@listserv.heanet.ie
https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND
Send bugs reports to
        https://github.com/csound/csound/issues
Discussions of bugs and features can be posted here

> On 29 Nov 2015, at 10:52 PM, jpff  wrote:
> 
> 
> 
> On Sun, 29 Nov 2015, Matti Koskinen wrote:
> 
>>> On 29 Nov 2015, at 10:12 PM, Justin Smith  wrote:
>>> 
>>> yes, that would normally be a conditional goto, only done if there was a message (to check if there is another after that in the same k cycle), it looks like here the intention is to infinitely loop one k cycle until there are no more messages in that cycle, then ask csound to exit?
> 
> That is how I read it
>> 
>> my code sends quite a lot of osc-messages,, so I thought the messages would choke csound, but just sending  only one from oscsend had no influence either.
> 
> because if there is no message you loop, as yoou dio if tere is one.  The effect is tat instr 1 never gets to the end of its k-cycle, and so csound is in a busy loop.

But a single orc-message from oscdump has no effect either 
> >
>> How to hack the cmakefiles to find osc includes and fluidsynth includes? I’m stuck on that compiling csound6.06 both on os x and linux.
> Which linux?  On Debian you need liblo-devel for OSC and libfluidsynth-devel for fluidsynth opcodes.
> 
> On OpenSuse tyou also need lliblo-devel and fluidsynth-devel
> 
> No idea for OSX
> 

I’ve installed these on ubuntu, the fluidisynth and liblo-incs are there. On OS X homebrew has installed them. Python, java all are found. But 
 OSC_HEADER="", so disabling BUILD_OSC_OPCODES and same with fluidsynth, maybe others, but these should definitely be needed.

tnx

-m

Csound mailing list
Csound@listserv.heanet.ie
https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND
Send bugs reports to
        https://github.com/csound/csound/issues
Discussions of bugs and features can be posted here

On Sun, 29 Nov 2015, Matti Koskinen wrote:

> But a single orc-message from oscdump has no effect either

Naturally.  Your code loops whether thereare essages or not. 
it never gets to the end of the perf cycle so nothing else runs


>>>
>>> How to hack the cmakefiles to find osc includes and fluidsynth includes? I’m stuck on that compiling csound6.06 both on os x and linux.
>> Which linux?  On Debian you need liblo-devel for OSC and libfluidsynth-devel for fluidsynth opcodes.
>>
>> On OpenSuse tyou also need lliblo-devel and fluidsynth-devel
>>
>> No idea for OSX
>>
>
> I’ve installed these on ubuntu, the fluidisynth and liblo-incs are there. On OS X homebrew has installed them. Python, java all are found. But
> OSC_HEADER="", so disabling BUILD_OSC_OPCODES and same with fluidsynth, maybe others, but these should definitely be needed.



fluidsynth-devel
           ^^^^^^
and liblo-devel
          ^^^^^^
> tnx
>
> -m
>

Csound mailing list
Csound@listserv.heanet.ie
https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND
Send bugs reports to
        https://github.com/csound/csound/issues
Discussions of bugs and features can be posted here

> On 29 Nov 2015, at 11:29 PM, jpff  wrote:
> 
> 
> 
> On Sun, 29 Nov 2015, Matti Koskinen wrote:
> 
>> But a single orc-message from oscdump has no effect either
> 
> Naturally.  Your code loops whether thereare essages or not. it never gets to the end of the perf cycle so nothing else runs
> 
> 
the thousand pound question is: why did the cod worked fine before, even today?

I used the same template last year to freeze audio, the jitter as Øivind pointed out, made a bad drift, in the end the video frame and sound were totally out of sync. 
> 
> fluidsynth-devel
>          ^^^^^^
> and liblo-devel

I’ve installed them on ubuntu, that makes things bit even stranger. But on further investigations, CMake version may have an effect. Using some old version couldn’t find png, newer found it. I remember when Victor created the cuda-opcodes, the only cmake was 2.4-11 that compiled the code correctly.Now  ccmake 3.3.2 can’t either find -lpng,, 3.4 can. (On OS X)
>          
>> tnx
>> 
>> -m
>> 
> 
> Csound mailing list
> Csound@listserv.heanet.ie
> https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND
> Send bugs reports to
>       https://github.com/csound/csound/issues
> Discussions of bugs and features can be posted here

Csound mailing list
Csound@listserv.heanet.ie
https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND
Send bugs reports to
        https://github.com/csound/csound/issues
Discussions of bugs and features can be posted here

nxtmsg:
   ;sprintf(msg,"ffffffi %f %f %f %f %f %f %d",vol,basefreq,freqshift,panstart,pan,modul,0);
      kk OSClisten gihandle,"/img2snd/run","ffffffi",kvol,kbasefreq,kfreqshift,kpanstart,kpan,kmodul,kquit
      if(kk==0) goto ex <—————————— I must have accidentally destroyed this line
      kvol = kvol / imaxinstr
      schedkwhen kk,0,imaxinstr,2,0,0.5,kvol,kbasefreq,kfreqshift,kpanstart,kpan,kmodul ; 
      schedkwhen kquit,0,3,3,0,1 ; quit
      kgoto nxtmsg
ex:
      ;prints "waitimg for osc"
endin

Oh sh………………..t

sorry for the noise

but in these cases for us idiots, csound shouldn’t be needed to  kill -9 to get out these freezes.

I looked almost the same kind of code, and noticed this goto.

-m

Csound mailing list
Csound@listserv.heanet.ie
https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND
Send bugs reports to
        https://github.com/csound/csound/issues
Discussions of bugs and features can be posted here