csound / csound

Main repository for Csound

CSOUND

Version 6.18.0

A sound and music computing system.

Csound is copyright (c) 1991-2020 The Csound Developers, see CONTRIBUTORS

Csound is free software; you can redistribute them and/or modify them under the terms of the GNU Lesser General Public License as published by the Free Software Foundation; either version 2.1 of the License, or (at your option) any later version.

Csound is distributed in the hope that they will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more details.

You should have received a copy of the GNU Lesser General Public License along with this software; if not, write to the Free Software Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA

GETTING STARTED

This repository contains the code for the core Csound library, the interfaces library, and the command-line interface frontend. It also contains specific code for ports to various embedded, mobile and web platforms.

For general project information, please look at http://csound.com, where all details about Csound, what it does, its history, music made with it, and many other things can be found.

The Csound Reference Manual can be found online at http://docs.csound.com.

The Csound API application programming interface reference may be found online at http://csound.github.io/docs/api/index.html.

Information on how to build Csound on various platforms is given in the BUILD.md file at the top level directory of this repository.

Information about build pipelines with Azure and Github actions is given in DockerFiles/Readme.md.

CONTRIBUTORS

Csound contains contributions from musicians, scientists, and programmers from around the world. They include (but are not limited to):

• Allan Lee
• Andres Cabrera
• Anthony Kozar
• Barry Vercoe
• Bill Gardner
• Bill Verplank
• Dan Ellis
• David Macintyre
• Ed Costello
• Eli Breder
• Fabio P. Bertolotti
• Felipe Sataler
• François Pinot
• Gabriel Maldonado
• Greg Sullivan
• Hans Mikelson
• Henri Manson
• Ian McCurdy
• Istvan Varga
• Jean Piché
• Joachim Heintz
• John Ramsdell
• John ffitch
• Marc Resibois
• Mark Dolson
• Matt Ingalls
• Max Mathews
• Michael Casey
• Michael Clark
• Michael Gogins
• Mike Berry
• Nate Whetsell
• Paris Smaragdis
• Perry Cook
• Peter Neubäcker
• Peter Nix
• Rasmus Ekman
• Richard Dobson
• Richard Karpen
• Rob Shaw
• Robin Whittle
• Rory Walsh
• Sean Costello
• Stephen Kyne
• Steven Yi
• Tito Latini
• Tom Erbe
• Victor Lazzarini
• Ville Pulkki

csound's Issues

v4.23f07 sources flawed--bad check out [sf bugs #4]

Originally reported by Sourceforge User: ramsdell

The CSound v4.23f07 sources are flawed. Several files
contain lines with angle brackets and equal signs. For
example, version.h contains the lines:

define SUBVER (23)

<<<<<<< version.h
<<<<<<< version.h

# define BUGFIX (06)

define BUGFIX (05)

>>>>>>> 1.3

Possibly there was a problem during check out?

John

main.c patches [sf bugs #5]

Originally reported by Sourceforge User: ramsdell

The file main.c contains the wrong prototype for psignal
and fails to omit rtclose when RTAUDIO is defined.
Please patch it as follows.

--- csound-4.23f06/main.c Fri May 30 10:28:22 2003
+++ csound/main.c Sun Aug 31 16:58:32 2003
@@ -244,7 +244,7 @@
}
}

-void psignal(int sig, char *str)
+void psignal(int sig, const char *str)
{
err_printf( "%s: %s\n", str, signal_to_string(sig));
}
@@ -688,7 +688,9 @@
void tranRESET(void);
void zakRESET(void);

+#if defined RTAUDIO
rtclose(); /* In case need to reopen /
+#endif
printf("O=%p\n", O);
#if defined(HAVE_FLTK) && defined(never) / IV -
Nov 30 2002 /
void widgetRESET(void); / N.B. this is not used
yet, */

UDP support on PC [sf bugs #30]

*Originally reported by Sourceforge User: anonymous

Numerous problems with opcodes/sockrecv.c trying to build vers. 5.04 on PC, Mingw, Msys, with UDP support opt checked in Sconstruct.

reserved identifier violation [sf bugs #48]

Originally reported by Sourceforge User: elfring

I suggest to try the search pattern "_+[A-Z]+" on the source files. You will find a couple of places where names begin with an underscore and an uppercase letter.

Examples:

• _GUID
  http://csound.cvs.sourceforge.net/viewvc/csound/csound5/H/pvfileio.h?view=markup&revision=1.14

- __Brass_h
http://csound.cvs.sourceforge.net/viewvc/csound/csound5/Opcodes/brass.h?view=markup&revision=1.7

This does not fit to the expected naming conventions of the C language standard.
http://en.wikipedia.org/wiki/Reserved\_identifier

See also section "7.1.3 Reserved identifiers".
http://www.open-std.org/JTC1/SC22/WG14/www/docs/n1124.pdf

Buffers for temporary files are to small [sf bugs #49]

Originally reported by Sourceforge User: alsroot

In ONE_FILE_GLOBALS structure, sizes of orcname, sconame and midname are too small when TMPDIR envar in linux is not empty

How to reproduce:
TMPDIR=/more/then/20/chars/string csound foo.csd

lack of whitespace in if statement causes segfault [sf bugs #24]

*Originally reported by Sourceforge User: anonymous

For if statements, if there is not a space between the if and the evaluated
expression, csound (5.00, as well as 5.00rc2 and 5.00rc3 and possibly
MacCSound) crashes.

For example, the following code would work:

if (kthing1 == 0) then
printk 1, kthing1
endif

Whereas this code would cause a segmentation fault:

if(kthing1 == 0) then
printk 1, kthing1
endif

The same seems to apply to elseif statements.

attached is an example file, which should exhibit the problem.

Thanks,
Phil

URL typo in 'INSTALL' [sf bugs #36]

Originally reported by Sourceforge User: mkc

The URL for installation instructions given in the file ./INSTALL is incorrect. It should be

http://csounds.com/manual/html/BuildingCsound.html

(note case)

Check return codes everywhere [sf bugs #34]

Originally reported by Sourceforge User: elfring

Some checks for return codes are missing.

Examples:
Would you like to add more error handling for return values from "fprintf" like in the function "het_export" and from "fseek" in the function "main"?
http://csound.cvs.sourceforge.net/csound/csound5/util/het\_export.c?revision=1.5&view=markup
http://csound.cvs.sourceforge.net/csound/csound5/fix\_format.c?revision=1.4&view=markup

Calling Csound VST method scoreEvent() from Python [sf bugs #19]

*Originally reported by Sourceforge User: anonymous

As requested by Michal Gogins:

"I suspect this method works in C++ but not in Python
since SWIG generates a binding to a special pointer
wrapper, and not to a list. I also think this could
easily be fixed. Could you please enter this as a bug
at SourceForge?"

This was his answer to my question about how to call
scoreEvent() from python.

csound.scoreEvent\("i", \[3, 0, 1, -2.2\]\)

gives me the following error:

File "CsoundVST.py", line 612, in scoreEvent
def scoreEvent(*args): return
_CsoundVST.CppSound_scoreEvent(*args)
TypeError: Expected a pointer

--with-fltk doesn't compile; --without-fltk does compile [sf bugs #11]

*Originally reported by Sourceforge User: anonymous

I checked out the latest source from CVS on June 8.
When I try to compile with "--with-fltk", I get this:

g++ -DHAVE_CONFIG_H -I. -I. -I. -fno-exceptions
-fno-rtti -g -O2 -Wall -I/usr/include/fltk-1.1
-I/usr/include/freetype2 -O2 -march=athlon-xp
-fomit-frame-pointer -I/usr/X11R6/include -c -o
widgets.o test -f widgets.cpp || echo './'widgets.cpp
In file included from
/usr/lib/gcc-lib/i686-pc-linux-gnu/3.3.3/include/g++-v3/backward/fstream.h:31,
from widgets.cpp:72:
/usr/lib/gcc-lib/i686-pc-linux-gnu/3.3.3/include/g++-v3/backward/backward_warning.h:32:2:
warning: #warning This file includes at least one
deprecated or antiquated header. Please consider using
one of the 32 headers found in section 17.4.1.2 of the
C++ standard. Examples include substituting the <X>
header for the <X.h> header for C++ includes, or
<sstream> instead of the deprecated header
<strstream.h>. To disable this warning use -Wno-deprecated.
widgets.cpp:77:23: strstream.h: No such file or directory
In file included from cs.h:36,
from widgets.cpp:89:
sysdep.h:43:29: warning: "/*" within comment
widgets.cpp: In function void load_snap(FLLOADSNAPS*)': widgets.cpp:1569: error: strstream' undeclared (first
use this function)
widgets.cpp:1569: error: (Each undeclared identifier is
reported only once for
each function it appears in.)
widgets.cpp:1569: error: parse error before ;' token widgets.cpp:1570: error: sbuf' undeclared (first use
this function)
widgets.cpp: In function void fl_slider_bank(FLSLIDERBANK*)': widgets.cpp:2764: error: parse error before ;' token
widgets.cpp: At top level:
widgets.cpp:1744: warning: void fltkKeybRun(void*)' defined but not used make[2]: *** [widgets.o] Error 1 make[2]: Leaving directory /home/znmeb/csound/csound'
make[1]: *** [all] Error 2
make[1]: Leaving directory `/home/znmeb/csound/csound'
make: *** [all-recursive] Error 1

The nice folks over at Gentoo have a bug filed against
this, but it doesn't seem to be here. Here's the URL to
the Gentoo discussion:

http://bugs.gentoo.org/show\_bug.cgi?id=45347

Ed Borasky
[email protected]

please cc: [email protected]

4.23f12 seg faults on amd64 linux [sf bugs #17]

*Originally reported by Sourceforge User: anonymous

2.6.10 64-bit kernel with 32-bit emulation on. Builds
fine after some monkeying around with the FL include
directories, and having to get autoheader.h out of CVS.
Spit out a bunch of warnings about the -malign-loops
cflag. Changed this to -falign-loops and all is well.
Using gcc-3.4.3. As I said, builds fine but seg faults
as soon as it starts. Can't even get a help or anything.

Compile of 5.06/7 fails in OOps/remote.c [sf bugs #41]

*Originally reported by Sourceforge User: anonymous

Compilation fails on Gentoo Linux (using the ebuild in 5.06) with:

OOps/remote.c: In function 'getIpAddress':
OOps/remote.c:70: error: storage size of 'ifr' isn't known
OOps/remote.c:70: warning: unused variable 'ifr'
OOps/remote.c: In function 'CLopen':
OOps/remote.c:230: warning: implicit declaration of function 'inet_aton'
scons: *** [OOps/remote.o] Error 1

It appears something in the ifreq structure in <linux/if.h> is sufficiently vague that the compiler refuses to give its size. On the other hand, a simple test file containing only the #include <linux/if.h> and struct ifreq ifr; lines compiles fine here...

5.06 (linux_d32) .so files depend on libresmgr.so (?) [sf bugs #37]

Originally reported by Sourceforge User: mkc

Several of the .so files in the 5.06 linux_d32 tarball depend on a shared library 'libresmgr.so'. This library seems not to exist in Debian or Ubuntu, for example. Is this dependency really needed? If not, it should be removed.

wxWindows + CSound = segmentation violation [sf bugs #2]

Originally reported by Sourceforge User: ramsdell

The program wxCSound is John D. Ramsdell's attempt to
contribute a
high-quality, vendor-neutral graphical user interface
for CSound. The
software is built on top of wxWindows, and is released
under the LGPL.
This prototype has been only tested on Red Hat Linux
7.1 with GTK.
The prototype demonstrates the validity of the
approach, as long as
the CSound bug with inconsistent AIFF sizes described
later can be
fixed.

Quick Start

Install the wxGTK version of wxWindows from
http://www.wxWindows.org.
If you do not install it as root, but sure to build
static libraries.

$ ./configure
$ make
$ cd src/wx
$ ./csoundwx knowing.csd

Design

The design calls for two threads, a worker thread used
by CSound to
render sound, and a main task used by the GUI.

Bugs

The current system does not handle csound.xmg files,
but the major
problem is this: the second time a rendering task runs,
the program
receives a segmentation fault. A transcript follows
along with the
code used to call CSound.

::::::::: TRANSCRIPT :::::::::::
Start this program with: ./csoundwx knowing.csd

Using default language
0dBFS level = 32767.0
Csound Version 4.23f04 (Aug 2 2003)
UnifiedCSD: knowing.csd
STARTING FILE

::::::::: CONTENT ELIDED :::::::::::

end of score. overall amps: 23497.4 40636.7
overall samples out of range: 0 388
0 errors in performance
peak CH 1: 23497.392578 (written: 0.717105) at 3150532
peak CH 2: 40636.710938 (written: 1.000000) at 1403251
4691 4096-byte soundblks of shorts written to
knowing.aiff (AIFF)
end of score. overall amps: 23497.4 40636.7
overall samples out of range: 0 388
0 errors in performance
O=(nil)
Using default language
0dBFS level = 32767.0
Csound Version 4.23f04 (Aug 2 2003)
UnifiedCSD: knowing.csd
STARTING FILE

::::::::: CONTENT ELIDED :::::::::::

inconsistent AIFF sizes

::::::::: END TRANSCRIPT :::::::::::

This message is printed by src/csound/aiff.c:182

::::::::: CONSOLE OUTPUT :::::::::::
Csound tidy up: Segmentation fault
Xlib: unexpected async reply (sequence 0x16ecb)!
::::::::: END CONSOLE OUTPUT :::::::::::

The file used to link wxCSound with CSound is
src/wx/jpff_goo.c, and
it contains:

::::::::: jpff_goo.c :::::::::::
#undef VERSION
#include &lt;csound.h&gt;
#include &lt;cs.h&gt;
#include &quot;jpff_goo.h&quot;

extern OPARMS O_;
extern GLOBALS cglob_;

int csoundTask(void *csound, int argc, char **argv)
{
int rc;
O = O_;
cglob = cglob_;
rc = csoundPerform(csound, argc, argv);
csoundCleanup(csound);
csoundReset(csound);
return rc;
}
::::::::: END jpff_goo.c :::::::::::

CSound was compiled with the following flags.

DEFS = -DWINDOWS -DSFIRCAM -DSYS5 -DPIPES
-DMACROS -DIV_SCHED -DFLTK_GUI

soundin causes csound (4.23f13) to segfault (w/ patch) [sf bugs #21]

Originally reported by Sourceforge User: tmoestl

Using soundin like this:
asig soundin "hellorcb.wav", iskiptime
causes a segfault in csound upon performance.
As far as I can tell, the reason is as follows: in r.
1.3 of soundio.h,
the additional parameter skipinit was added to soundin;
to link
it up, an additional optional parameter was added to
the spec
in entry1.c. However, there was already a parameter before
skipinit in the SOUNDIN structure, kfrqratio, that was not
specified in entry1.c, and completely unused. Thus,
kfrqratio
gets initialized with the value that ought to be in
skipinit,
and skipinit is not initalized, causing crashes as soon as
it is accessed.
The attached patch fixes this by removing kfrqratio
completely;
don't know if that might have some negative
side-effects (e.g.
on backward compatability), though

3 bugs ?? cs5 + winXP [sf bugs #16]

*Originally reported by Sourceforge User: anonymous

Hello,

Here 3 very simple csd giving problem with:
Windows XP Home and csound5 (last release from web
(5beta) )

;PROBLEM MISSING NOTE(S) ??
;****************************************
<CsoundSynthesizer>
<CsOptions>
-W
</CsOptions>
<CsInstruments>
sr = 44100
kr = 4410
ksmps = 10
nchnls = 1

instr 1
asig oscil 20000, 440, 1
out asig
endin
</CsInstruments>
<CsScore>
f1 0 8192 10 1

t 0 70 ;no more problem if this line is removed

i1 15 1
i1 16 1 ;WHERE IS THIS NOTE ?
i1 17 1
</CsScore>
</CsoundSynthesizer>

;BUG WITH FTGEN ??
;**************************
<CsoundSynthesizer>
<CsOptions>
-W
</CsOptions>
<CsInstruments>
sr = 44100
kr = 4410
ksmps = 10
nchnls = 1

gitab ftgen 0, 0, 8192, 10, 1
;give "Segmentation violation" message ON
CSOUND5

instr 1
asig oscil 20000, 440, gitab
out asig
endin
</CsInstruments>
<CsScore>
i1 0 1
</CsScore>
</CsoundSynthesizer>

;TOTAL PERF DURATION NOT EXACT ??
;*************************************************
<CsoundSynthesizer>
<CsOptions>
-W
</CsOptions>
<CsInstruments>
sr = 44100
kr = 4410
ksmps = 10
nchnls = 1

instr 1
asig oscil 20000, 55, 1
out asig
endin
</CsInstruments>
<CsScore>
f1 0 8192 10 1

i1 0 1

;Measure on the generated wave (with Sound Forge5 on
Windows XP)
;With csound5 : Duration = 0.871519 seconds
;With csoundAV and prévious csound : Duration =
1.000000 seconds
</CsScore>
</CsoundSynthesizer>

Segmentation fault (Version 4.24.1) [sf bugs #20]

*Originally reported by Sourceforge User: anonymous

I was trying out these files:
http://www.csounds.com/jmc/Articles/Woodwinds/clarinet\_macros.zip
which are found on this page:
http://www.csounds.com/jmc/Articles/Woodwinds/Csound+Woodwinds.htm
as example "Clarinet (improved .sco version)".

I applied this command line:
csound -C -W -o out.wav clarinethole3.orc clarinethole3.sco

After everything apparently went through fine, at the very end I
got this error message:
Csound tidy up: Segmentation fault

The result file out.wav was obviously not properly closed, as it
was not loadable with XMMS nor Audacity. I could however
import it as raw data in Audacity, and it seemed that all the
content had been produced correctly.

I am currently using
Csound Version 4.24.1 with 32 bit floats (Apr 6 2004)
with SuSE Linux 9.2 on a Dell Inspiron 7500 (Pentium III 600 MHz)
with 384 MB RAM.

Kind regards

Cannot compile sources and segmentation fault [sf bugs #1]

Originally reported by Sourceforge User: ramsdell

See transcript.

cant find python25.dll at startup [sf bugs #42]

Originally reported by Sourceforge User: alexkilias

csound installs python but l8r when u start the application this error appear!

cant find python25.dll...reinstall the application

Scons library checking flaw, because of missing libintl [sf bugs #45]

*Originally reported by Sourceforge User: anonymous

When libintl (gettext) is not installed, scons spuriously reports missing libsndfile instead.
Even when setting useGettext=0, library checks get still compiled with -lintl and fail.

Csound-5.10.1
OS X 10.4.11
scons-1.2.0
gcc 4.0.1

Tempest Bug [sf bugs #33]

*Originally reported by Sourceforge User: anonymous

Hello,

I've already submitted this to the Developers mailing list but thought i'd post it here as well.

When using tempest on mac 10.3.9 using Csound 5.04 from the command line it only returns 'nan' (when printing its value with printks).

Examples of tempest from online don't work which makes me think it's a bug with the new version.

Thanks,

Colin

scons install failure [sf bugs #23]

Originally reported by Sourceforge User: geemus

Fails when trying to install csound5-20051124.tar.bz2
from csound.sourceforge.net

changed line 27 in Opcodes/dssi4cs/src/dssi.h
from:

#include <ladspa.h>

to:

#include "ladspa.h"

and it works fine.

thanks,
wes

line opcode only takes three args [sf bugs #43]

*Originally reported by Sourceforge User: anonymous

According to the manual, the line generated can pass through 2 or more specified points. But attempting to create more than two points gives an error: too many input args, line 12:
kcps line 880, p3, 220, 1, 110

In particular, the example program will fail because of this.

tar file extracts to current directory [sf bugs #9]

Originally reported by Sourceforge User: ivan_ganza

This seems to have this way for a long time now.

When you untar the distro (on unix), I suspect its the
same on windows, all the files are placed in the
current directoy you are in. This is not very good,
usually the user would expect the extraction to go into
a seperate dir, perhaps &quot;csound&quot; or some such.

Its annoying to extract a tarfile then have to look and
delete all the files since you probably didn't mean for
all the files to be extracted to your current working
directory.

CsoundVST appears as audio plugin not VSTi [sf bugs #8]

Originally reported by Sourceforge User: analoghell

CsoundVST appears as an audio plugin not a VSTi in
Cubase SX 2.01?

input files crashing cs5/ soundin,diskin [sf bugs #15]

*Originally reported by Sourceforge User: anonymous

I noticed that when using -i soundfile.wav I get a
crash from csound5 (I'm running aug12 version).
Also soundin does not seem to work; diskin does, but
seems to cut after a little while (2 secs).

[email protected]

Critical backwards compatibility issue [sf bugs #28]

Originally reported by Sourceforge User: dakbari

Since I've posted this topic to the Csound and Csound-devel lists and still haven't received a response that leads to a solution to this issue of vital importance, I turn now to this outlet in the hopes that the idea will persist after the thread on the mailing lists has died.

Basically getting right to the point, Csound version 5.03.1 breaks a piece that was written with 5.02.0.

A backwards incompatible change has to have been made to one or more of the following opcodes

* pgmassign

• massign
• ctrlinit
• midiin
• sensekey
• changed
• turnoff2
• event
• diskin2
• filenchnls
• filelen

After countless hours spent grepping through source code and the like, I can see that the Csound file does indeed work correctly when used with version tagged 5.02 available from this very website, however it stops working when a user gets any of the following recent versions: 5.03.0, 5.03.1, 5.04.0.

Here is the example file that works with 5.02 but not with 5.03

http://www.csounds.com/akbari/akbari\_olpc2.zip

I will leave this up as long as this bug remains unfixed.

I have also attached the verbose output in plain text (using -v in <CsOptions>) of both the correctly working version 5.02 and the anomalous version 5.03. This behavior is consistent on at least GNU/Linux 2.6.12 kernel, Mac OS 10.3.9, and Windows XP (MinGW) all built from sources, tagged as official releases on this website.

Since it looks like it's only going to let me upload one file, here are links to both outputs

This one works but is the old version of Csound5
http://www.csounds.com/akbari/cs502out.txt

This one doesn't work but is the new version of Csound5
http://www.csounds.com/akbari/cs503out.txt

Completion of error handling [sf bugs #46]

Originally reported by Sourceforge User: elfring

I would like to continue the previous bug report "Check return codes everywhere".
https://sourceforge.net/tracker/?func=detail&atid=564599&aid=1680767&group\_id=81968

A couple of source files contain still unchecked function calls.

Examples:

• ftload() ⇒ fread(), fgets()
  ftsave() ⇒ fwrite(), fprintf()
  http://csound.cvs.sourceforge.net/viewvc/csound/csound5/Opcodes/ftgen.c?revision=1.16&view=markup

- RTclose() ⇒ fclose(), fcntl()
http://csound.cvs.sourceforge.net/viewvc/csound/csound5/Engine/linevent.c?revision=1.44&view=markup

- Conversions::fill() ⇒ strdup()
http://csound.cvs.sourceforge.net/viewvc/csound/csound5/frontends/CsoundAC/Conversions.cpp?revision=1.8&view=markup

Would you like to reduce the efforts for error code checking by an exception class hierarchy?
http://dietmar-kuehl.de/mirror/c++-faq/exceptions.html\#faq-17.1
http://cexcept.sourceforge.net/

Illegal instruction error message in csound5gui 5.03 &others [sf bugs #27]

Originally reported by Sourceforge User: gomesbascoy

When I execute csound5gui, csoundVST or winsound I get
this error "Application not initialized correctly
(0xc000001d)".
I think that the last release of csound it wasn't
compiled for my processor architecture (AMD K6-II 550Mhz)..

[FIX] Heap Overflow in Marker Parsing [sf bugs #32]

Originally reported by Sourceforge User: jcmdev0

Creating a mark label longer than the number of existing labels causes heap corruption (exhibited by a glibc doubly linked list corruption detection in newer libc). Looking at the code, this is because i is used to determine the length of the label, then used to index into the label list, then reference again as the length of the label.

I've attached a patch that adds a j index for the label list.

This appears to still be an issue in the cvs csound5 version.

[Having not quite gotten things testably compiling, so I can't verify]

Prevent building _csnd with libcsound soname [sf bugs #47]

Originally reported by Sourceforge User: alsroot

thats a python '+=' issue in SConstruct
it leads to generate _csnd.so(at least) with soname "libscound.so..."

CSound sf home page [sf bugs #6]

Originally reported by Sourceforge User: ramsdell

Please insert the enclosed index.html file as your
SourceForge home page. It redirects to
www.csounds.com. Edit the URLs if you prefer
something else. You change your home page by
changing the contents of the htdocs module in your CVS
repository.

Csound-5.07.0.tar.gz unpacks into worng dir [sf bugs #40]

Originally reported by Sourceforge User: rwst

the distributed tarball has all files in the csound5 subdirectory which messes up the Gentoo ebuild. (The previous version had it all in Csound5.06.0/) Don't you guys have standardized procedures for making the tarball like 'make dist' or 'make tarball'?

rtaudio fails on Gentoo Linux build [sf bugs #12]

Originally reported by Sourceforge User: znmeb

I'm trying to build CSound 4.23 from CVS source on a
Gentoo Linux system. If I use the default,
"--enable-rtaudio", it crashes thusly:

gcc -DHAVE_CONFIG_H -I. -I. -I. -g -O2 -Wall
-falign-loops=4 -falign-jumps=4 -ffast-math
-fomit-frame-pointer -finline-functions -funroll-loops
-I/usr/X11R6/include -c winX11.c -MT winX11.lo -MD -MP
-MF .deps/winX11.TPlo -o winX11.o >/dev/null 2>&1
/bin/sh ../libtool --mode=link g++ -fno-exceptions
-fno-rtti -g -O2 -Wall -L/usr/X11R6/lib -o
libcsound.la -rpath /usr/local/lib -version-info 0:0:0
-no-undefined libcsound_la-main.lo
libcsound_la-argdecode.lo libcsound_la-musmon.lo
libcsound_la-otran.lo libcsound_la-oload.lo
libcsound_la-rdorch.lo libcsound_la-express.lo
libcsound_la-rdscor.lo libcsound_la-linevent.lo
libcsound_la-cscore.lo libcsound_la-cscorfns.lo
libcsound_la-insert.lo libcsound_la-entry1.lo
libcsound_la-entry2.lo libcsound_la-aops.lo
libcsound_la-midiops.lo libcsound_la-midirecv.lo
libcsound_la-fgens.lo libcsound_la-ugens1.lo
libcsound_la-ugens2.lo libcsound_la-ugens3.lo
libcsound_la-ugens4.lo libcsound_la-ugens5.lo
libcsound_la-ugens6.lo libcsound_la-ugens7.lo
libcsound_la-ugens8.lo libcsound_la-ugens9.lo
libcsound_la-spectra.lo libcsound_la-disprep.lo
libcsound_la-window.lo libcsound_la-winascii.lo
libcsound_la-hetro.lo libcsound_la-lpanal.lo
libcsound_la-lptrkfns.lo libcsound_la-pvanal.lo
libcsound_la-pvoc.lo libcsound_la-fft.lo
libcsound_la-dsputil.lo libcsound_la-sndinfo.lo
libcsound_la-sfheader.lo libcsound_la-soundin.lo
libcsound_la-soundio.lo libcsound_la-ulaw.lo
libcsound_la-aiff.lo libcsound_la-wave.lo
libcsound_la-ieee80.lo libcsound_la-filopen.lo
libcsound_la-memalloc.lo libcsound_la-memfiles.lo
libcsound_la-auxfd.lo libcsound_la-windin.lo
libcsound_la-dumpf.lo libcsound_la-scot.lo
libcsound_la-scsort.lo libcsound_la-scxtract.lo
libcsound_la-sread.lo libcsound_la-sort.lo
libcsound_la-extract.lo libcsound_la-twarp.lo
libcsound_la-swrite.lo libcsound_la-butter.lo
libcsound_la-grain.lo libcsound_la-vdelay.lo
libcsound_la-cmath.lo libcsound_la-follow.lo
libcsound_la-cvanal.lo libcsound_la-grain4.lo
libcsound_la-pvinterp.lo libcsound_la-pvread.lo
libcsound_la-sndwarp.lo libcsound_la-vpvoc.lo
libcsound_la-hrtferX.lo libcsound_la-winEPS.lo
libcsound_la-nlfilt.lo libcsound_la-opcode.lo
libcsound_la-ugrw1.lo libcsound_la-ugrw2.lo
libcsound_la-ugensa.lo libcsound_la-cross2.lo
libcsound_la-dam.lo libcsound_la-midiout.lo
libcsound_la-midiops2.lo libcsound_la-midisend.lo
libcsound_la-wavegde.lo libcsound_la-filter.lo
libcsound_la-complex.lo libcsound_la-repluck.lo
libcsound_la-pluck.lo libcsound_la-physutil.lo
libcsound_la-physmod.lo libcsound_la-modal4.lo
libcsound_la-shaker.lo libcsound_la-fm4op.lo
libcsound_la-diskin.lo libcsound_la-pvadd.lo
libcsound_la-moog1.lo libcsound_la-singwave.lo
libcsound_la-mandolin.lo libcsound_la-space.lo
libcsound_la-locsig.lo libcsound_la-dcblockr.lo
libcsound_la-pvocext.lo libcsound_la-flanger.lo
libcsound_la-aifc.lo libcsound_la-lowpassr.lo
libcsound_la-uggab.lo libcsound_la-biquad.lo
libcsound_la-one_file.lo libcsound_la-schedule.lo
libcsound_la-midiops3.lo libcsound_la-pitch.lo
libcsound_la-getstring.lo libcsound_la-ugsc.lo
libcsound_la-fout.lo libcsound_la-sndinfUG.lo
libcsound_la-pvlook.lo libcsound_la-scansyn.lo
libcsound_la-sfont.lo libcsound_la-vbap.lo
libcsound_la-vbap_four.lo libcsound_la-vbap_eight.lo
libcsound_la-vbap_sixteen.lo libcsound_la-vbap_zak.lo
libcsound_la-control.lo libcsound_la-bowedbar.lo
libcsound_la-phisem.lo libcsound_la-sdif.lo
libcsound_la-sdif-mem.lo libcsound_la-babo.lo
libcsound_la-ugmoss.lo libcsound_la-bbcut.lo
libcsound_la-spat3d.lo libcsound_la-pvxanal.lo
libcsound_la-mxfft.lo libcsound_la-pvsanal.lo
libcsound_la-pstream.lo libcsound_la-pvfileio.lo
libcsound_la-oscbnk.lo libcsound_la-scansynx.lo
libcsound_la-oscils.lo libcsound_la-wave-terrain.lo
libcsound_la-dnoise.lo libcsound_la-dnfft.lo
libcsound_la-mididevice.lo libcsound_la-clfilt.lo
libcsound_la-midiinterop.lo libcsound_la-csound.lo
libcsound_la-load_opcodes.lo libcsound_la-namedins.lo
libcsound_la-cmsg.lo rtlinux.lo LINUXaudio.lo
winX11.lo -lportaudio -lasound -lX11 -ldl -lm
libtool: link: rtlinux.lo' is not a valid libtool object make[2]: *** [libcsound.la] Error 1 make[2]: Leaving directory /home/znmeb/csound/csound'
make[1]: *** [all] Error 2
make[1]: Leaving directory `/home/znmeb/csound/csound'
make: *** [all-recursive] Error 1

Is "rtlinux.lo" part of CSound, or is it something I
have to get somewhere else? And, if somewhere else, where?

Ed Borasky
[email protected]
http://AlgoCompSynth.com

Please cc: [email protected] and reference bug 45347

http://bugs.gentoo.org/show\_bug.cgi?id=45347

sndinfo works but gives seg violation on x86 [sf bugs #39]

Originally reported by Sourceforge User: tims45

Csound 5.06(d) on a 2GHz Celeron running Windows XP SP2. sndinfo (or csound -U sndinfo) does work in terms of reporting file into, but always ends with "Cannot tidy up: Segmentation violation". Doesn't seem to matter what the format of the file is.

Not a serious error, at least as far as my use is concerned...

Typical example follows

$ sndinfo A440.aif
Localisation of messages is disabled, using default language.
WARNING: 'C:\Program Files\Csound\plugins64\csladspa.dll' is not a Csound plugin
library
WARNING: 'C:\Program Files\Csound\plugins64\csnd.dll' is not a Csound plugin lib
rary
PortMIDI real time MIDI plugin for Csound
PortAudio real-time audio module for Csound
Windows MME real time audio and MIDI module for Csound by Istvan Varga
WARNING: 'C:\Program Files\Csound\plugins64\scoregen.dll' is not a Csound plugin
library
virtual_keyboard real time MIDI plugin for Csound
util sndinfo:
A440.aif:
srate 44100, monaural, 16 bit AIFF, 1.485 seconds
(65489 sample frames)
Csound tidy up: Segmentation violation

$

csnd.jar bug, chnget doesn't work properly [sf bugs #31]

*Originally reported by Sourceforge User: anonymous

When using csnd.jar in java, the chnget opcode does not work on Windows PCs. This makes real-time use of java and Csound impossible. The following compilation error occurs when running the sliderdemo example from the csound journal:

>SliderDemo.java:180:
>GetChannelPtr(csnd.SWIGTYPE_p_float,java.lang.String,int) i
>n csnd.Csound cannot be applied to
>(csnd.SWIGTYPE_p_p_float,java.lang.String,int
>)if(csound.GetChannelPtr(myfltarray.GetPtr(), "pitch",
> ^
>SliderDemo.java:217:
>GetChannelPtr(csnd.SWIGTYPE_p_float,java.lang.String,int) i
>n csnd.Csound cannot be applied to
>(csnd.SWIGTYPE_p_p_float,java.lang.String,int
>)if(csound.GetChannelPtr(myfltarray.GetPtr(), "pitch",
> ^
>2 errors

Other similar examples result in a similar compilation error. Jim Hearon and I have tried to trace the problem, here is what he said:

---

Hi Casey,
I did a little more work on this and found a couple more things. As I
mentioned before the wrappers are very delicate and sensitive to changes in
the c/c++ source code.

The GetChannelPtr wrapper method changed somewhere along the way from:

public int GetChannelPtr(SWIGTYPE_p_p_float p, String name, int type) {
return csndJNI.Csound_GetChannelPtr(swigCPtr,
SWIGTYPE_p_p_float.getCPtr(p), name, type);
}

to this:

public int GetChannelPtr(SWIGTYPE_p_float p, String name, int type) {
return csndJNI.Csound_GetChannelPtr(swigCPtr, this,
SWIGTYPE_p_float.getCPtr(p), name, type);
}

but the odd thing is the api (csound.h) method hasn't changed, so I'm not
sure why it's being wrapped differently and causing errors.

which throws problematic frame error in lib
[_jcsound.dll+0xba91]

the other thing I noticed in the java code for the SliderDemo.java is that
the
CsoundMYFLTArray.java, GetPtr() method quit working for some reason, and
now needs the CsoundMYFLTArray.java, GetPtr(int indx) method. So something
changed
there too.

if(csound.GetChannelPtr(myfltarray.GetPtr(1),
"pitch", csndConstants.CSOUND_INPUT_CHANNEL
|csndConstants.CSOUND_CONTROL_CHANNEL ) == 0)
myfltarray.SetValue(0, (double)PitchNum);

Regards,
Jim Hearon

Thanks for your help,

Casey Mongoven
[email protected]

undefined symbol in lib_jcsound.so [sf bugs #50]

*Originally reported by Sourceforge User: anonymous

when attempting to use Java API following error occurred:

java: symbol lookup error: /usr/lib/jni/lib_jcsound.so: undefined symbol: csoundCreate

autoheader.h missing from csound4.23f12 [sf bugs #14]

Originally reported by Sourceforge User: bencollver

My build fails when cs.h tries to include autoheader.h,
which is not in the source distribution. I don't see any
provision such as a configure script or Makefile target to
generate autoheader.h. The error follows:

cc -Wall -DSFIRCAM -DSYS5 -DLINUX -DPIPES -
DMACROS -DPREFIX="/usr/pkg" -DRTAUDIO -
DNETBSDAUDIO -c main.c -o main.o
In file included from cs.h:35,
from main.c:24:
sysdep.h:24: autoheader.h: No such file or directory

csound5gui crashes [sf bugs #35]

*Originally reported by Sourceforge User: anonymous

Installed csound 5.05 on a WinXP SP2 PC, SoundMax built in audio.

Using csound5gui, I could successfully play the example trapped.csd

However, the following actions cause immediate exit of csound5gui:

- clicking "Configure" in "Output file/device" block
(row of controls are all dimmed)

- selecting "Options>Csound" in bottom right block

(Selecting "Options>GUI" does not cause exit.)

tim scott
[email protected]

4.23f09 tarball broken [sf bugs #10]

Originally reported by Sourceforge User: fugalh

$ tar tvzf Csound_423f09.tar.gz
drwxrwxr-x jpff/jpff 0 2004-01-07 01:56:14
csound_4.23f09/
-rwxr-xr-x jpff/jpff 1730 2003-05-30 08:28:12
csound_4.23f09/3Dug.h
tar: Skipping to next header
tar: Archive contains obsolescent base-64 headers

gzip: stdin: invalid compressed data--format violated
tar: Child returned status 1
tar: Error exit delayed from previous errors

$ gunzip Csound_423f09.tar.gz

gunzip: Csound_423f09.tar.gz: invalid compressed
data--format violated

Environment variable handling problems [sf bugs #38]

Originally reported by Sourceforge User: tims45

The environment is Win XP SP2, Csound 5.06(d)

I am attempting to set SFDIR and/or SSDIR in the <CsOptions> clause of a CSD.

First of all, attempting to use an argument such as:

--env:SSDIR=C:\Documents and Settings\zorg\My Documents\Personal\Studio stuff\Csound files 2.0GB\CD 3 Csound Catalog\instruments and compositions

(with or without " or ' quotes around the pathname) fails as the parser sees something like \D as an escaped D.

So I changed all \ to / and then Csound appears to think that the first space indicates the end of the first argument. Attempting to "escape" spaces with something like {space} didn't work either.

So then I attempted to replace the spaces with %20

--env:SSDIR="C:/Documents%20and%20Settings/zorg/My%20Documents/Personal/Studio%20stuff/Csound%20files%202.0GB/CD%203%20Csound%20Catalog/catalog_samples/"

which also does not work. I don't see an error message, but any files required by the CSD that live in that directory are not found.

However, if I use a short path, even relative, such as

--env:SSDIR=../catalog_samples

that works (assuming the file you're running is in the proper relation to that directory.)

So my guess is that the parser may stop reading the string after some number of characters, sets SSDIR to some unknown quantity, and thus fails to find the desired file.

If I could figure out a way for Csound to print what it understands SSDIR to be from within a CSD file, I could probably confirm this...

Release file names must include a version number [sf bugs #7]

Originally reported by Sourceforge User: ramsdell

Currently, source releases are all named Csound.tar.gz.
The result is that requests for previous releases
return the current release. In other words, all the
previous releases have been overwritten. Please name
releases using the pattern csound-VERSION.tar.gz, were
VERSION is the version of the release. The current
release should be named.

csound-4.23f07.tar.gz

tarball broken [sf bugs #3]

Originally reported by Sourceforge User: fugalh

The Csound.tar.gz source tarball for 4.23f06 is broken:

$ tar xzf ../Csound.tar.gz
incomplete literal tree

gzip: stdin: invalid compressed data--format violated
tar: Child returned status 1
tar: Error exit delayed from previous errors

ZIW opcode does not work [sf bugs #26]

*Originally reported by Sourceforge User: anonymous

Hi,

in the ziw() function (ugrw1.c) the following code at
the beginning prevents ZIW from working normally:

if (zkset(csound,(ZKR*)p) == 0) return NOTOK;

zkset() returns zero if ZAK space is allocated; so in
the normal conditions the rest of ziw() function is
not executed.

---

Thank you

Many params in ftgen causes problem [sf bugs #29]

Originally reported by Sourceforge User: mbechard

System: Win XP Pro SP 2, Csound 5.04, 2ghz CPU, 2ghz RAM

Synopsis: usage of ftgen using GEN02 with many (256) parameters causes strange problem (ftgen used outside instrument declaration).

Compile the attached CSD; should give following error:

INIT ERROR in instr 1: Invalid ftable no. 0.000000
kamp envlpx 1 .01 p3 #i0 gitbatk_1 .85 .001
0
B 0.000 - note deleted. i1 had 1 init errors

Comment out "gitbstif_1 ftgen..." and it should compile fine.

More soundin breakage (4.23f13): GEN01, cvanal etc. segfault [sf bugs #22]

Originally reported by Sourceforge User: tmoestl

(Please see report # 1351582 for an introduction to the
problem)

More problems due to the addition of the skipinit parameter
to soundin: in two places where soundin is used internally,
skipinit was not initialized; the first is GEN01, the
second
is SAsndgetset(), which is used for various stand-alone
utilities, e.g. cvanal.
The attached patch should fix that, and also includes the
diff from report # 1351582.

Potential integer overflow in pvlook() function [sf bugs #44]

Originally reported by Sourceforge User: wtl-icst

Pvlook doesn't check input file rightly. A crafted input file could cause integer overflow.

PVOCDATA data;
......
if ((fp = csound->PVOC_OpenFile(csound, argv[argc - 1], &data, &fmt)) < 0)
......

framesize = data.nAnalysisBins * 2 * sizeof(float);
frames = (float*) csound->Malloc(csound, framesize * numframes);

PVOC_Openfile() fills PVOCDATA "data" with data from input file by invoking pvoc_readheader() function. data.nAnalysisBins is a tainted value, however,PVlook trusts the value.
data.nAnalysisBins * 2 * sizeof(float)* numframes is easily overflowed by a huge data.nAnalysisBins.

Right? Waiting for your reply, thanks.

Tielei Wang.

Build problem [sf bugs #18]

*Originally reported by Sourceforge User: anonymous

csound-4.23f12gbs.9 does not build here because of line
6 in csound/ccsound.cpp.

I changed <csound/csound.h> to "csound.h", which fixed
the problem.

Regards,
Matthias <[email protected]>

Build of csound5 fails in rtpa.c with Gentoo Linux [sf bugs #13]

Originally reported by Sourceforge User: znmeb

I'm running Gentoo Linux (up to date) with gcc=3.3.3:

bash-2.05b$ gcc --version
gcc (GCC) 3.3.3 20040412 (Gentoo Linux 3.3.3-r6,
ssp-3.3.2-2, pie-8.7.6)

I checked CSound5 out of CVS today (2004-06-12). The
build croaked on rtpa.c:

gcc -DCSOUND_WITH_API -g -O2 -DLINUX -Wall -DPIPES
-DHAVE_FCNTL_H -DHAVE_UNISTD_H -DHAVE_MALLOC_H
-DHAVE_SGTTY_H -DHAVE_SYS_IOCTL_H -DHAVE_SYS_TIME_H
-DHAVE_SYS_TYPES_H -DHAVE_TERMIOS_H -DHAVE_STRING_H
-DHAVE_STRINGS_H -DHAVE_DIRENT_H -DRTAUDIO -DBETA -I.
-IH -I/usr/local/include -I/usr/include -c -o
InOut/rtpa.o InOut/rtpa.c
InOut/rtpa.c: In function recopen_': InOut/rtpa.c:69: error: storage size of paStreamParameters_' isn't known
InOut/rtpa.c:84: error: paALSA' undeclared (first use in this function) InOut/rtpa.c:84: error: (Each undeclared identifier is reported only once InOut/rtpa.c:84: error: for each function it appears in.) InOut/rtpa.c:87: error: paUseHostApiSpecificDeviceSpecification' undeclared
(first use in this function)
InOut/rtpa.c:119: warning: passing arg 3 of
Pa_OpenStream' makes integer from pointer without a cast InOut/rtpa.c:119: warning: passing arg 5 of Pa_OpenStream' makes pointer from integer without a cast
InOut/rtpa.c:119: warning: passing arg 7 of
Pa_OpenStream' makes integer from pointer without a cast InOut/rtpa.c:119: warning: passing arg 8 of Pa_OpenStream' makes integer from pointer without a cast
InOut/rtpa.c:119: error: too few arguments to function
Pa_OpenStream' InOut/rtpa.c:69: warning: unused variable paStreamParameters_'
InOut/rtpa.c: In function listPortAudioDevices': InOut/rtpa.c:130: error: PaDeviceIndex' undeclared
(first use in this function)
InOut/rtpa.c:130: error: syntax error before "deviceIndex"
InOut/rtpa.c:134: error: deviceCount' undeclared (first use in this function) InOut/rtpa.c:134: warning: implicit declaration of function Pa_GetDeviceCount'
InOut/rtpa.c:136: error: deviceIndex' undeclared (first use in this function) InOut/rtpa.c:146: error: structure has no member named defaultSampleRate'
InOut/rtpa.c: In function playopen_': InOut/rtpa.c:154: error: storage size of paStreamParameters_' isn't known
InOut/rtpa.c:173: error: paALSA' undeclared (first use in this function) InOut/rtpa.c:176: error: paUseHostApiSpecificDeviceSpecification' undeclared
(first use in this function)
InOut/rtpa.c:207: warning: passing arg 2 of
Pa_OpenStream' makes integer from pointer without a cast InOut/rtpa.c:207: warning: passing arg 5 of Pa_OpenStream' makes pointer from integer without a cast
InOut/rtpa.c:207: warning: passing arg 7 of
Pa_OpenStream' makes integer from pointer without a cast InOut/rtpa.c:207: warning: passing arg 8 of Pa_OpenStream' makes integer from pointer without a cast
InOut/rtpa.c:207: error: too few arguments to function
Pa_OpenStream' InOut/rtpa.c:154: warning: unused variable paStreamParameters_'
InOut/rtpa.c: In function rtrecord_': InOut/rtpa.c:233: warning: implicit declaration of function Pa_ReadStream'
InOut/rtpa.c: In function rtplay_': InOut/rtpa.c:261: warning: implicit declaration of function Pa_WriteStream'
scons: *** [InOut/rtpa.o] Error 1
scons: building terminated because of errors.

Taking a hint from previous attempts with CSound 4.23,
I disabled the "PortAudio" flag in the SConstruct file,
and it got past this. However, it still doesn't finish
... it gets into trouble in "libbabo":

gcc "-Wl,-Bdynamic -shared" -o libbabo.so
Opcodes/babo.os -L. -L. -L. -L/usr/lib -L/usr/local/lib
/usr/lib/gcc-lib/i686-pc-linux-gnu/3.3.3/../../../../i686-pc-linux-gnu/bin/ld:
unrecognized option '-Bdynamic -shared'
/usr/lib/gcc-lib/i686-pc-linux-gnu/3.3.3/../../../../i686-pc-linux-gnu/bin/ld:
use the --help option for usage information
collect2: ld returned 1 exit status
scons: *** [libbabo.so] Error 1
scons: building terminated because of errors.

Ed Borasky
[email protected]
http://AlgoCompSynth.com/

ZIR opcode does not work [sf bugs #25]

*Originally reported by Sourceforge User: anonymous

Hi all,

in the zir() function (ugrw1.c) the following code at
the beginning prevents ZIR from working normally:

if (zkset(csound,p) == 0) return OK;

zkset() returns zero if ZAK space allocated; so, in
the normal conditions, the rest of zir() function is
not executed.

I suspect that this line should look like:

if (zkset(csound,p) != OK) return OK;

---

May the Force be with you