Xiph.Org issueshttps://gitlab.xiph.org/groups/xiph/-/issues2008-05-30T21:39:16Zhttps://gitlab.xiph.org/xiph/vorbis-tools/-/issues/243Segmentation fault2008-05-30T21:39:16ZjanSegmentation fault```
Version 1.0
Using invalid characters for US-ASCII leads to a segmentation fault in
iconv.c:129 calling iconv(cd1, 0, 0, 0, 0). Commenting this line out fixes the
problem.
``````
Version 1.0
Using invalid characters for US-ASCII leads to a segmentation fault in
iconv.c:129 calling iconv(cd1, 0, 0, 0, 0). Commenting this line out fixes the
problem.
```Michael SmithMichael Smithhttps://gitlab.xiph.org/xiph/vorbis-tools/-/issues/242Assertion failed2006-06-12T11:48:20ZjanAssertion failed```
Then setting any of artist, album etc. through the command line options an
"assertion failed" ist generated:
oggenc: iconvert.c:111: iconvert: Zusicherung »(!k && !ibl) || (k ==
(size_t)(-1) && (*__errno_location ()) == 7 && ibl &...```
Then setting any of artist, album etc. through the command line options an
"assertion failed" ist generated:
oggenc: iconvert.c:111: iconvert: Zusicherung »(!k && !ibl) || (k ==
(size_t)(-1) && (*__errno_location ()) == 7 && ibl && obl < 6) || (k ==
(size_t)(-1) && ((*__errno_location ()) == 84 || (*__errno_location ()) == 22)
&& ibl)« nicht erfüllt
The same happens with vorbiscomment. I'm using a SuSE 6.3 (quite old),
nationalized to german. GCC is 2.95.3.
```Michael SmithMichael Smithhttps://gitlab.xiph.org/xiph/vorbis-tools/-/issues/241Assertion failed2006-06-12T11:33:57ZjanAssertion failed```
``````
```Michael SmithMichael Smithhttps://gitlab.xiph.org/xiph/vorbis-tools/-/issues/228oggenc segfaults on some .wav files2002-08-24T06:46:27Zjddoggenc segfaults on some .wav files```
Have found several .wav's that oggenc will core dump on. Platform is
Solaris 8 on UltraSparc running in 64-bit mode (compiling as 32-bit).
Example file: http://www.nnaf.net/~jdd/oggenc_core.wav
That file contains the first 10 seco...```
Have found several .wav's that oggenc will core dump on. Platform is
Solaris 8 on UltraSparc running in 64-bit mode (compiling as 32-bit).
Example file: http://www.nnaf.net/~jdd/oggenc_core.wav
That file contains the first 10 seconds or so (1 mb) of Ozzy Osbourne's
Mr. Tinkertrain.
Here is the stack trace:
(gdb) set args -o blah.ogg /tmp/oggenc_core.wav
(gdb) run
Program received signal SIGSEGV, Segmentation fault.
0xff24b7b4 in apsort (a=0xffbeef4c, b=0xffbeef54) at psy.c:953
953 if(fabs(**(float **)a)>fabs(**(float **)b))return -1;
(gdb) where
#0 0xff24b7b4 in apsort (a=0xffbeef4c, b=0xffbeef54) at psy.c:953
#1 0xff14b24c in qsort () from /usr/lib/libc.so.1
#2 0xff24bc6c in _vp_noise_normalize_sort (p=0x59708, magnitudes=0x93fe0,
sortedindex=0xffbef010) at psy.c:994
#3 0xff259220 in mapping0_forward (vb=0xffbef478) at mapping0.c:532
#4 0xff244068 in vorbis_analysis (vb=0xffbef478, op=0x0) at analysis.c:47
#5 0x1877c in oe_encode (opt=0xffbef7c0) at encode.c:283
#6 0x131e0 in main (argc=4, argv=0xffbef93c) at oggenc.c:337
```Michael SmithMichael Smithhttps://gitlab.xiph.org/xiph/vorbis-tools/-/issues/213cvs oggenc makes VBR files with upper/lower bitrate of 0.000000 kb/s2002-07-19T06:31:00ZKyungjoon Leecvs oggenc makes VBR files with upper/lower bitrate of 0.000000 kb/s```
I encoded somthing with cvs oggenc's VBR mode, and ran ogginfo on it.
Nominal bitrate: 112.001000 kb/s
Upper bitrate: 0.000000 kb/s
Lower bitrate: 0.000000 kb/s
IMHO upper/lower should not be set when using VBR modes. Upper bitrate...```
I encoded somthing with cvs oggenc's VBR mode, and ran ogginfo on it.
Nominal bitrate: 112.001000 kb/s
Upper bitrate: 0.000000 kb/s
Lower bitrate: 0.000000 kb/s
IMHO upper/lower should not be set when using VBR modes. Upper bitrate being
equal to Lower bitrate seems awkward too..
```Michael SmithMichael Smithhttps://gitlab.xiph.org/xiph/vorbis-tools/-/issues/207oggenc man page still says RC3 in the title2002-07-19T06:25:55Zchaeseoggenc man page still says RC3 in the title```
The title of the oggenc man page still reads "Vorbis Tools release candidate 3"
You might want to change that before rolling out 1.0 :)
``````
The title of the oggenc man page still reads "Vorbis Tools release candidate 3"
You might want to change that before rolling out 1.0 :)
```Michael SmithMichael Smithhttps://gitlab.xiph.org/xiph/vorbis-tools/-/issues/205platform.c patch for Borland C2002-07-01T06:33:47Zdaveplatform.c patch for Borland C```
Small changes necessary for compiling with BCC in Windows. Diff against CVS.
``````
Small changes necessary for compiling with BCC in Windows. Diff against CVS.
```Michael SmithMichael Smithhttps://gitlab.xiph.org/xiph/vorbis-tools/-/issues/203vorbis-tools won't work unless used with autoconf-2.53 on Redhat 7.32002-11-07T14:15:59Zyusufgvorbis-tools won't work unless used with autoconf-2.53 on Redhat 7.3```
Hi, I freshly checked out from CVS HEAD the following three modules
ogg
vorbis
vorbis-tools
I am using Redhat Linux 7.3 with kernel 2.4.18-4. It has both autoconf (2.13)
and autoconf (2.53) installed on the system. autoconf 2.13 is ...```
Hi, I freshly checked out from CVS HEAD the following three modules
ogg
vorbis
vorbis-tools
I am using Redhat Linux 7.3 with kernel 2.4.18-4. It has both autoconf (2.13)
and autoconf (2.53) installed on the system. autoconf 2.13 is the default
I can autogen.sh inside the ogg and vorbis directly and install them into the
desired prefix. Whilst doing the same in vorbis-tools, I get the following error
/autogen.sh --help
Generating configuration files for vorbis-tools, please wait....
aclocal
libtoolize --automake
automake --add-missing
automake: configure.in: installing `./install-sh'
automake: configure.in: installing `./mkinstalldirs'
automake: configure.in: installing `./missing'
autoconf
./configure: line 217: syntax error near unexpected token `;'
./configure: line 217: ` ac_optarg= ;;'
Now, I setup my shell so that autoconf refers to autoconf-2.53, then I get no
failure though I get this warning
configure.in:29: warning: AC_CANONICAL_HOST invoked multiple times
```Michael SmithMichael Smithhttps://gitlab.xiph.org/xiph/vorbis-tools/-/issues/202change -t option to be consistent with -a, -l, etc...2009-04-19T21:17:41Zluke.usherwoodchange -t option to be consistent with -a, -l, etc...```
Currently, using the -t option has the side-effect of changeing the default filename and path,
whereas the options -a, -l etc. do not. The option '-n' can be used to provide the functionality to
generate a filename from the name s...```
Currently, using the -t option has the side-effect of changeing the default filename and path,
whereas the options -a, -l etc. do not. The option '-n' can be used to provide the functionality to
generate a filename from the name specified using '-t', therefore the extra functionality in the -t
option seems redundant.
It would seem more logical to just have the '-t' perform just the 1 task
of setting the title tag, and leave the '-o' or '-n' options to specify the output filename.
```Michael SmithMichael Smithhttps://gitlab.xiph.org/xiph/vorbis-tools/-/issues/197ogg tools can't compile without libcurl... Provide more modularity2006-06-12T10:27:32Ztsvetkovogg tools can't compile without libcurl... Provide more modularity```
Vorbis tools do NOT need CURL to provide basic functionality! Please provide
posibility for more modular compilation!
Also libao is too much platform dependent and is NOT needed for encoder to work!
For example: I can't _encode_ ogg ...```
Vorbis tools do NOT need CURL to provide basic functionality! Please provide
posibility for more modular compilation!
Also libao is too much platform dependent and is NOT needed for encoder to work!
For example: I can't _encode_ ogg files on HP pa-risk machine, only because its
not supported by libao!
Can I have some realy simple and compilable encoder which is realy portable?
```Jack MoffittJack Moffitthttps://gitlab.xiph.org/xiph/vorbis-tools/-/issues/196cannot compile: In `oe_encode': encode.o: undefined references to `vorbis_enc...2002-05-26T13:03:38Zmarvincannot compile: In `oe_encode': encode.o: undefined references to `vorbis_encode_init_vbr' `vorbis_bitrate_addblock' `vorbis_bitrate_flushpacket'```
using Caldera OpenLinux 3.1 ( and replacing it as soon as possible -- they went
institutional on us ), with gcc 2.95.2 Caldera-patched, installed the latest
cURL because vorbis .configure demanded it, ran ldconfig because cURL instal...```
using Caldera OpenLinux 3.1 ( and replacing it as soon as possible -- they went
institutional on us ), with gcc 2.95.2 Caldera-patched, installed the latest
cURL because vorbis .configure demanded it, ran ldconfig because cURL install
didn't do it, ran "rm config.cache -f ; ./configure && make" and got dumped.
p.s. I run /tmp noexec, and that's created problems with installing
OpenOffice, and maybe with other things, too. If this ( or any other
filesystem limitations designed to limit damage/hostile-action is creating it,
then please put something in the README about it, eh? )
Here's the text of the failure
===============================================
make[2]: Leaving directory
`/mnt/md1/marvin/Linux.MultiMedia/Sound.OggVorbis/vorbis-tools-1.0rc3/oggenc/man'
make[2]: Entering directory
`/mnt/md1/marvin/Linux.MultiMedia/Sound.OggVorbis/vorbis-tools-1.0rc3/oggenc'
/bin/sh ../libtool --mode=link gcc -O20 -ffast-math -fsigned-char -o oggenc
oggenc.o audio.o encode.o platform.o -lvorbisenc -lvorbis -lm -logg
../share/libutf8.a ../share/libgetopt.a
gcc -O20 -ffast-math -fsigned-char -o oggenc oggenc.o audio.o encode.o
platform.o /usr//lib/libvorbisenc.so /usr//lib/libvorbis.so -lm
/usr//lib/libogg.so ../share/libutf8.a ../share/libgetopt.a -Wl,--rpath
-Wl,/usr//lib -Wl,--rpath -Wl,/usr//lib
encode.o: In function `oe_encode':
encode.o(.text+0x8e): undefined reference to `vorbis_encode_init_vbr'
encode.o(.text+0x35c): undefined reference to `vorbis_bitrate_addblock'
encode.o(.text+0x404): undefined reference to `vorbis_bitrate_flushpacket'
make[2]: *** [oggenc] Error 1
make[2]: Leaving directory
`/mnt/md1/marvin/Linux.MultiMedia/Sound.OggVorbis/vorbis-tools-1.0rc3/oggenc'
make[1]: *** [all-recursive] Error 1
make[1]: Leaving directory
`/mnt/md1/marvin/Linux.MultiMedia/Sound.OggVorbis/vorbis-tools-1.0rc3/oggenc'
make: *** [all-recursive] Error 1
========================
That's all, folks. . .
```Michael SmithMichael Smithhttps://gitlab.xiph.org/xiph/vorbis-tools/-/issues/186encoder stops when encoded file gets larger than RAM memory2002-04-22T08:56:57ZGitlab Botencoder stops when encoded file gets larger than RAM memory```
Hi,
i sometimes real-time-encode long radio streams by piping the wav-stream to
oggenc's standard input.
but for very long sessions this does not work, because the encoded file cannot
get larger than my physical RAM memory (in my c...```
Hi,
i sometimes real-time-encode long radio streams by piping the wav-stream to
oggenc's standard input.
but for very long sessions this does not work, because the encoded file cannot
get larger than my physical RAM memory (in my case 192MB ram, whitch is about
220 minutes of music).
it would be nice, if one could encode music of about 24 hours without
having 1,4 Gig of memory available.
```Michael SmithMichael Smithhttps://gitlab.xiph.org/xiph/vorbis-tools/-/issues/183memory leak2002-04-15T00:48:22ZGitlab Botmemory leak```
OggEnc v09.L seems to have a memory leak, when I encoded a 13 min
44Khz track it gradually consumed all my free memory.
I'm under Win2K SP2.
``````
OggEnc v09.L seems to have a memory leak, when I encoded a 13 min
44Khz track it gradually consumed all my free memory.
I'm under Win2K SP2.
```Michael SmithMichael Smithhttps://gitlab.xiph.org/xiph/vorbis-tools/-/issues/181short files are encoded into empty ogg files2002-04-22T08:54:18Zretoshort files are encoded into empty ogg files```
I have a short wav file and use oggenc with no flags to encode it. It works
perfectly, except for files that are shorter than ~0.5 seconds. These will just
be blank. I am aware it may not be space effective to encode tiny sounds, but...```
I have a short wav file and use oggenc with no flags to encode it. It works
perfectly, except for files that are shorter than ~0.5 seconds. These will just
be blank. I am aware it may not be space effective to encode tiny sounds, but
want to do it for consistency
```Michael SmithMichael Smithhttps://gitlab.xiph.org/xiph/vorbis-tools/-/issues/177oggenc >FILE creates a wrong file2019-07-01T08:51:49Zfabrice.bauzacoggenc >FILE creates a wrong file```
Let me reproduce the bugreport I sent to the Debian bug-tracking system:
Subject: oggenc: --output=FILE and >FILE have different results
Package: vorbis-tools
Version: 1.0rc3-1
Severity: important
oggenc seems to create different f...```
Let me reproduce the bugreport I sent to the Debian bug-tracking system:
Subject: oggenc: --output=FILE and >FILE have different results
Package: vorbis-tools
Version: 1.0rc3-1
Severity: important
oggenc seems to create different files depending on the style of the
commandline, whether it is
$ cat test.raw | oggenc --raw - --output=test-inputpipe.ogg
or
$ cat test.raw | oggenc --raw - > test-pipe.ogg
The big trouble is that the latter method (> FILE) produces a file
that ogg123 can read without any visible error/warning, but ogginfo
says that the integrity test fails, and vorbiscomment doesn't even
recognize the file as a Ogg file.
An example should make this clearer:
================================================================
*** First, we create the files.
$ oggenc --raw test.raw --output=test.ogg
Encoding "test.raw" to
"test.ogg" at quality 3,00
Encoding with VBR
Encoding [ 1m39s so far] -
Done encoding file "test.ogg"
File length: 1m 01,0s
Elapsed time: 1m 40,2s
Rate: 0,6183
Average bitrate: 84,1 kb/s
$ cat test.raw | oggenc --raw - > test-pipe.ogg
Encoding standard input to
standard output at quality 3,00
Encoding [ 1m48s so far] -
Done encoding.
File length: 1m 01,0s
Elapsed time: 1m 49,1s
Rate: 0,5675
Average bitrate: 84,1 kb/s
$ cat test.raw | oggenc --raw - --output=test-inputpipe.ogg
Encoding standard input to
"test-inputpipe.ogg" at quality 3,00
Encoding with VBR
Encoding [ 1m40s so far] -
Done encoding file "test-inputpipe.ogg"
File length: 1m 01,0s
Elapsed time: 1m 41,0s
Rate: 0,6132
Average bitrate: 84,1 kb/s
*** Now, we try these files:
$ ogginfo test.ogg
filename=test.ogg
serial=1963857601
header_integrity=pass
vendor=Xiphophorus libVorbis I 20011231
version=0
channels=2
rate=44100
bitrate_upper=none
bitrate_nominal=112015
bitrate_lower=none
stream_integrity=pass
bitrate_average=83060
length=61.936327
playtime=1:01
stream_truncated=false
total_length=61.936327
total_playtime=1:01
*** No problem for test.ogg.
$ ogginfo test-pipe.ogg
filename=test-pipe.ogg
header_integrity=fail
stream_integrity=fail
stream_truncated=true
serial=1478540987
header_integrity=pass
vendor=Xiphophorus libVorbis I 20011231
version=0
channels=2
rate=44100
bitrate_upper=none
bitrate_nominal=112015
bitrate_lower=none
stream_integrity=pass
bitrate_average=82533
length=61.936327
playtime=1:01
stream_truncated=false
total_length=61.936327
total_playtime=1:01
*** Trouble!
$ ogginfo test-inputpipe.ogg
filename=test-inputpipe.ogg
serial=33960753
header_integrity=pass
vendor=Xiphophorus libVorbis I 20011231
version=0
channels=2
rate=44100
bitrate_upper=none
bitrate_nominal=112015
bitrate_lower=none
stream_integrity=pass
bitrate_average=83060
length=61.936327
playtime=1:01
stream_truncated=false
total_length=61.936327
total_playtime=1:01
*** No problem. Now let's try vorbiscomment:
$ vorbiscomment test.ogg
$ vorbiscomment test-pipe.ogg
Failed to open file as vorbis: Input is not an Ogg bitstream.
$ vorbiscomment test-inputpipe.ogg
$
================================================================
-- System Information
Debian Release: 3.0
Architecture: i386
Kernel: Linux noon.dnsalias.net 2.4.9 #1 sam mar 2 20:48:32 CET 2002 i686
Locale: LANG=fr_FR, LC_CTYPE=fr_FR
Versions of packages vorbis-tools depends on:
ii libao2 0.8.2-1 Cross Platform Audio Output Librar
ii libc6 2.2.5-4 GNU C Library: Shared libraries an
ii libcurl2-ssl 7.9.5-2 Multi-protocol file transfer libra
ii libogg0 1.0rc3-1 Ogg Bitstream Library
ii libvorbis0 1.0rc3-1 The Vorbis General Audio Compressi
```Michael SmithMichael Smithhttps://gitlab.xiph.org/xiph/vorbis-tools/-/issues/173oggenc ignores bitrate setting2002-03-27T06:54:03Zflifsonoggenc ignores bitrate setting```
oggenc seems to ignore the bitrate setting. I get huge .ogg files with bitrates
in the 300+ area when I specified a much lower bit rate. Here's an example.
[flifson@asgard london_calling]$ oggenc -b 128 11_wrong_em_boyo.wav
Opening...```
oggenc seems to ignore the bitrate setting. I get huge .ogg files with bitrates
in the 300+ area when I specified a much lower bit rate. Here's an example.
[flifson@asgard london_calling]$ oggenc -b 128 11_wrong_em_boyo.wav
Opening with wav module: WAV file reader
Encoding "11_wrong_em_boyo.ogg" [ 99.9%] [ 0m00s remaining] \
Done encoding file "11_wrong_em_boyo.ogg"
File length: 3m 13.0s
Elapsed time: 0m 56.9s
Rate: 3.3988
Average bitrate: 347.9 kb/s
[flifson@asgard london_calling]$ oggenc -v
OggEnc v0.8 (libvorbis rc2)
I'm using 0.8rc2. Does this not have bitrate limits enabled?
```Michael SmithMichael Smithhttps://gitlab.xiph.org/xiph/vorbis-tools/-/issues/172Oggenc RC3 Cannot handle Very Long Pathnames2002-04-07T21:14:58ZkoranskyOggenc RC3 Cannot handle Very Long Pathnames```
I have found that Oggenc.exe does not run if the pathnames to the files are
very long. I have double checked with LAME and MPEG Plus, they run with no
problem. Very long pathnames are needed with my music collection for detailed
d...```
I have found that Oggenc.exe does not run if the pathnames to the files are
very long. I have double checked with LAME and MPEG Plus, they run with no
problem. Very long pathnames are needed with my music collection for detailed
descriptions.
```Michael SmithMichael Smithhttps://gitlab.xiph.org/xiph/vorbis-tools/-/issues/168%n to -n should be more flexible2002-03-22T14:45:39Zjdahlin%n to -n should be more flexible```
%n should support something like this:
%2n -> ' 1', ' 2', ..., '10', '11' (always 2 in width)
%02n -> '01', '02', ..., '10', '11' (always 2 in width, pad with zeros)
%-2n -> '1 ', '2 ', ..., '10,, '11' (always 2 in width, align lef...```
%n should support something like this:
%2n -> ' 1', ' 2', ..., '10', '11' (always 2 in width)
%02n -> '01', '02', ..., '10', '11' (always 2 in width, pad with zeros)
%-2n -> '1 ', '2 ', ..., '10,, '11' (always 2 in width, align left)
I feel that %0xN is most important of the three.
```Michael SmithMichael Smithhttps://gitlab.xiph.org/xiph/vorbis-tools/-/issues/160Win32 Vorbis Tools is missing documentation2002-08-09T16:22:14ZtoddWin32 Vorbis Tools is missing documentation```
I didn't even know there was any documentation for these, but I heard from
Segher Boessenkool that there are supposed to at least be man pages for each of
the tools. However, in the 1.0rc3 ZIP file, there are none. Just the
exec...```
I didn't even know there was any documentation for these, but I heard from
Segher Boessenkool that there are supposed to at least be man pages for each of
the tools. However, in the 1.0rc3 ZIP file, there are none. Just the
executables.
```Jack MoffittJack Moffitthttps://gitlab.xiph.org/xiph/vorbis-tools/-/issues/147Can't Use full filenames in oggenc since Vakor started committing afresh2006-06-12T11:37:34ZnisharfiCan't Use full filenames in oggenc since Vakor started committing afresh```
C:\>oggenc.postrc3 "D:\share\src\First Time\wav\track15.wav" --
output="D:\share\src\First Time\ogg\Battle Royal.ogg"
Opening with wav module: WAV file reader
Couldn't create directory "D:": Permission denied
ERROR: Could not creat...```
C:\>oggenc.postrc3 "D:\share\src\First Time\wav\track15.wav" --
output="D:\share\src\First Time\ogg\Battle Royal.ogg"
Opening with wav module: WAV file reader
Couldn't create directory "D:": Permission denied
ERROR: Could not create required subdirectories for output
filename "D:\share\src\First Time\ogg\Battle Royal.ogg"
the same error happens when I try to do the same at D:\ . However, when I run
the same command line from D:\cvs (an unrelated directory), I get the error
D:\cvs>oggenc.postrc3 "D:\share\src\First Time\wav\track15.wav" --
output="D:\share\src\First Time\ogg\Battle Royal.ogg"
Opening with wav module: WAV file reader
Couldn't create directory "D:": File exists
ERROR: Could not create required subdirectories for output
filename "D:\share\src\First Time\ogg\Battle Royal.ogg"
I'd bet this has something to do with the autocreate directory code that
Michael added; could this be tweaked to not try and overwrite the root
directory when using absolute paths?
```Michael SmithMichael Smith