[Csnd] Probable bug in ATSadd
Date | 2011-06-28 04:31 |
From | Ernesto Illescas |
Subject | [Csnd] Probable bug in ATSadd |
Attachments | original.ats |
Dear All, First of all thanks for the great source of information this list is. I hope soon I can contribute with solutions... Anyway to the issue: I believe I've found a bug with the ATSadd opcode. I'm working on Debian Squeeze (adm64) with "Csound version 5.12 (double samples) Aug 5 2010" (quoting from my terminal). I ignore if this happens on other architectures-builds. What I've been able to establish is that if I can reduce the number of partials for the re-synthesis (p4 in my score), but once I got to a value, I can't increment it without getting an error. This is evidenced by commeting-uncommenting the last score-statement in my .sco (the commented statement shows no error). From what I've read on the internet about the error, it seems that once I reduce the number of partials I use, memory is freed, and I can't access it again; but Iḿ not sure. Here is the error message, the .orc, the and .sco. I'm also attaching my .ats analysis file. The error: PortAudio real-time audio module for Csound PortMIDI real time MIDI plugin for Csound virtual_keyboard real time MIDI plugin for Csound 0dBFS level = 32768.0 Csound version 5.12 (double samples) Aug 5 2010 libsndfile-1.0.21 Reading options from $CSOUNDRC: /home/ernesto/csound/.csoundrc orchname: temp.orc scorename: temp.sco rtaudio: ALSA module enabled RAWWAVE_PATH: /usr/share/stk/rawwaves/ rtmidi: PortMIDI module enabled orch compiler: Macro definition for AtsNumPartials Macro definition for AtsDur instr 1 Elapsed time at end of orchestra compile: real: 0.001s, CPU: 0.000s sorting score ... ... done Elapsed time at end of score sort: real: 0.001s, CPU: 0.000s Csound version 5.12 (double samples) Aug 5 2010 0dBFS level = 1.0 orch now loaded audio buffered in 256 sample-frame blocks writing 512-byte blks of shorts to /home/ernesto/csound/soundfiles/test.wav (WAV) SECTION 1: ftable 1: new alloc for instr 1: file original.ats (791640 bytes) loaded into memory instr 1: iAtsNumPartials = 77.000 B 0.000 .. 2.482 T 2.482 TT 2.482 M: 0.85321 instr 1: iAtsNumPartials = 77.000 B 2.482 .. 4.964 T 4.964 TT 4.964 M: 0.85063 instr 1: iAtsNumPartials = 77.000 B 4.964 .. 7.446 T 7.446 TT 7.446 M:123786381.31152 *** glibc detected *** csound: malloc(): memory corruption: 0x00000000007a61c0 *** The orchestra: sr = 44100 kr = 4410 ksmps = 10 nchnls = 1 0dbfs = 1.0 ; Macro definitions #define AtsNumPartials #3# #define AtsDur #7# instr 1 SAtsfile = "original.ats" ; Extract data from header of ATS anlysis. iAtsNumPartials ATSinfo SAtsfile, $AtsNumPartials iAtsDur ATSinfo SAtsfile, $AtsDur print iAtsNumPartials iDur = p3 if p4 > 0 then iNumPartials = p4 else iNumPartials = iAtsNumPartials endif kTimePointer line 0, iDur, iAtsDur - .01 aSig ATSadd kTimePointer, 1, SAtsfile, 1, iNumPartials, 0 out aSig endin The score: f1 0 16384 10 1 ; a sine wave i1 0 2.482 0 i1 + 2.482 50 i1 + 2.482 60 I hope this helps better the code, apart from helping me ;~). Best, Ernesto |
Date | 2011-06-28 12:41 |
From | Victor Lazzarini |
Subject | Re: [Csnd] Probable bug in ATSadd |
Thanks, this is fixed now in GIT. Victor On 28 Jun 2011, at 04:31, Ernesto Illescas wrote:
Dr Victor Lazzarini Senior Lecturer Dept. of Music NUI Maynooth Ireland tel.: +353 1 708 3545 Victor dot Lazzarini AT nuim dot ie |
Date | 2011-06-29 02:06 |
From | Ernesto Illescas |
Subject | Re: [Csnd] Probable bug in ATSadd |
Dear Victor, Thanks for your fast response. I guess I'll have to download the sources from Guitorius, and use the development version. With my skills, this seems like a free-weekend project. One thing wasn't clear, though. Was the bug already fixed, or was it a simple correction that you just made? I ask this because I ignore if it is relevant that I report bugs from v5.02 (the stable Debian repos must be months behind, at the least). If sending bugs from this version is relevant, I also found a bug in ATSsinnoi: the ipartialincr parameter gives a segmentation fault if set to anything but 1. If it is useful, I'll send a simple .orc and .sco that reproduce the
bug. Best, Ernesto De: Victor Lazzarini <Victor.Lazzarini@nuim.ie> Para: csound@lists.bath.ac.uk Enviado: martes, 28 de junio, 2011 6:41:58 Asunto: Re: [Csnd] Probable bug in ATSadd Thanks, this is fixed now in GIT. Victor On 28 Jun 2011, at 04:31, Ernesto Illescas wrote:
Dr Victor Lazzarini Senior Lecturer Dept. of Music NUI Maynooth Ireland tel.: +353 1 708 3545 Victor dot
Lazzarini AT nuim dot ie |
Date | 2011-06-29 05:35 |
From | jpff@cs.bath.ac.uk |
Subject | Re: [Csnd] Probable bug in ATSadd |
> Dear Victor, > > Thanks for your fast response. I guess I'll have to download the sources > from Guitorius, and use the development version. With my skills, this > seems like a free-weekend project. One thing wasn't clear, though. Was the > bug already fixed, or was it a simple correction that you just made? I ask > this because I ignore if it is relevant that I report bugs from v5.02 (the > stable Debian repos must be months behind, at the least). > > If sending bugs from this version is relevant, I also found a bug in > ATSsinnoi: the ipartialincr parameter gives a segmentation fault if set > to anything but 1. If it is useful, I'll send a simple .orc and .sco that > reproduce the bug. > > Best, > > Ernesto > Just a small reminder; there is a fresh build from GIT for Debian intel 32bit processr) every (UK) night availble from htto://codemist.co.uk/jpff/Debian_Snapshot.tgz ==John ff 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 | 2011-06-29 15:04 |
From | Ernesto Illescas |
Subject | Re: [Csnd] Probable bug in ATSadd |
Thanks for the tip, John! I didn't know about the nightly builds. I guess I can force the 32-bit into my architecture with dpkg. Regards, Ernesto De: "jpff@cs.bath.ac.uk" <jpff@cs.bath.ac.uk> Para: csound@lists.bath.ac.uk Enviado: martes, 28 de junio, 2011 23:35:52 Asunto: Re: [Csnd] Probable bug in ATSadd > Dear Victor, > > Thanks for your fast response. I guess I'll have to download the sources > from Guitorius, and use the development version. With my skills, this > seems like a free-weekend project. One thing wasn't clear, though. Was the > bug already fixed, or was it a simple correction that you just made? I ask > this because I ignore if it is relevant that I report bugs from v5.02 (the > stable Debian repos must be months behind, at the least). > > If sending bugs from this version is relevant, I also found a bug in > ATSsinnoi: the ipartialincr parameter gives a segmentation fault if set > to anything but 1. If it is useful, I'll send a simple .orc and .sco that > reproduce the bug. > > Best, > > Ernesto > Just a small reminder; there is a fresh build from GIT for Debian intel 32bit processr) every (UK) night availble from htto://codemist.co.uk/jpff/Debian_Snapshot.tgz ==John ff 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 | 2011-06-29 16:06 |
From | Victor Lazzarini |
Subject | Re: [Csnd] Probable bug in ATSadd |
Please send bugs, some of them have been fixed, but others (like the one you reported) have not. Victor On 29 Jun 2011, at 02:06, Ernesto Illescas wrote:
Dr Victor Lazzarini Senior Lecturer Dept. of Music NUI Maynooth Ireland tel.: +353 1 708 3545 Victor dot Lazzarini AT nuim dot ie |
Date | 2011-07-21 01:22 |
From | Ernesto Illescas |
Subject | Re: [Csnd] Probable bug in ATSadd |
Dear Victor, I was finally able to build the development version and test the fix. I'll post information that could be useful for Debianers in another thread. The thing that might be relevant for this post is that I'm now on testing (I just could not build without errors on stable). I ran the example score I sent in the development version, and it does run, however I'm still observing strange behaviour. The problem now is with the amplitudes. The first note sounds alright, but at the beginning of the second and third notes the amplitude goes to the moon for 10 samples. here's the output from the terminal: ernesto@studio:~$ csound test.orc test.sco time resolution is 1000.000 ns 0dBFS level = 32768.0 Csound version 5.13 beta (double samples) Jul 20 2011 libsndfile-1.0.24 Reading options from $CSOUNDRC: /home/ernesto/nube9/csound/.csoundrc orchname: test.orc scorename: test.sco rtaudio: ALSA module enabled graph init orch compiler: Macro definition for AtsNumPartials Macro definition for AtsDur 30 (31) lines read instr 1 Elapsed time at end of orchestra compile: real: 0.002s, CPU: 0.000s sorting score ... ... done Elapsed time at end of score sort: real: 0.002s, CPU: 0.000s Csound version 5.13 beta (double samples) Jul 20 2011 0dBFS level = 1.0 orch now loaded audio buffered in 256 sample-frame blocks writing 512-byte blks of shorts to /home/ernesto/nube9/csound/soundfiles/test.wav (WAV) SECTION 1: ftable 1: new alloc for instr 1: file original.ats (791640 bytes) loaded into memory instr 1: iAtsNumPartials = 77.000 B 0.000 .. 2.482 T 2.482 TT 2.482 M: 0.85000 instr 1: iAtsNumPartials = 77.000 B 2.482 .. 4.964 T 4.964 TT 4.964 M:171150994.93107 number of samples out of range: 10 instr 1: iAtsNumPartials = 77.000 B 4.964 .. 7.446 T 7.446 TT 7.446 M:123639702.91078 number of samples out of range: 10 Score finished in csoundPerform(). inactive allocs returned to freespace end of score. overall amps:171150994.93107 overall samples out of range: 20 0 errors in performance Elapsed time at end of performance: real: 0.190s, CPU: 0.150s 1283 512-byte soundblks of shorts written to /home/ernesto/nube9/csound/soundfiles/test.wav (WAV) Removing temporary file /tmp/csound-mcXeYN.srt ... Best, Ernesto De: Victor Lazzarini <Victor.Lazzarini@nuim.ie> Para: csound@lists.bath.ac.uk Enviado: miércoles, 29 de junio, 2011 10:06:12 Asunto: Re: [Csnd] Probable bug in ATSadd Please send bugs, some of them have been fixed, but others (like the one you reported) have not. Victor On 29 Jun 2011, at 02:06, Ernesto Illescas wrote:
Dr Victor Lazzarini Senior Lecturer Dept. of Music NUI Maynooth Ireland tel.: +353 1 708 3545 Victor dot Lazzarini AT nuim dot ie |