+Q1 : Do you plan to support XYZ codec in libsndfile?
+Q2 : In version 0 the SF_INFO struct had a pcmbitwidth field
+ but version 1 does not. Why?
+Q3 : Compiling is really slow on MacOS X. Why?
+Q4 : When trying to compile libsndfile on Solaris I get a "bad
+ substitution" error during linking. What can I do to fix this?
+Q5 : Why doesn't libsndfile do interleaving/de-interleaving?
+Q6 : What's the best format for storing temporary files?
+Q7 : On Linux/Unix/MacOS X, what's the best way of detecting the
+ presence of libsndfile?
+Q8 : But I just want a simple Makefile! What do I do?
+Q9 : How about adding the ability to write/read sound files to/from
+ memory buffers?
+Q10 : Reading a 16 bit PCM file as normalised floats and then
+ writing them back changes some sample values. Why?
+Q11 : I'm having problems with u-law encoded WAV files generated by
+ libsndfile in Winamp. Why?
+Q12 : I'm looking at sf_read*. What are items? What are frames?
+Q13 : Why can't libsndfile open this Sound Designer II (SD2)
+ file?
+Q14 : I'd like to statically link libsndfile to my closed source
+ application. Can I buy a license so that this is possible?
+Q15 : My program is crashing during a call to a function in libsndfile.
+ Is this a bug in libsndfile?
+Q16 : Will you accept a fix for compiling libsndfile with compiler X?
+
+Q17 : Can libsndfile read/write files from/to UNIX pipes?
+
+Q18 : Is it possible to build a Universal Binary on Mac OS X?
+
+Q19 : I have project files for Visual Studio / XCode / Whatever. Why
+ don't you distribute them with libsndfile?
+
+Q20 : Why doesn't libsndfile support MP3? Lots of other Open Source
+ projects support it!
+
+Q21 : How do I use libsndfile in a closed source or commercial program
+ and comply with the license?
+
+Q22 : What versions of windows does libsndfile work on?
+
+Q23 : I'm cross compiling libsndfile for another platform. How can I
+ run the test suite?
+
+
+If source code for XYZ codec is available under a suitable license (LGPL, BSD, +MIT etc) then yes, I'd like to add it. +
++If suitable documentation is available on how to decode and encode the format +then maybe, depending on how much work is involved. +
++If XYZ is some proprietary codec where no source code or documentation is +available then no. +
++So if you want support for XYZ codec, first find existing source code or +documentation. +If you can't find either then the answer is no. +
+ + ++ This was dropped for a number of reasons: +
++As documented + here +there is now a well defined behaviour which ensures that no matter what the +bit width of the source file, the scaling always does something sensible. +This makes it safe to read 8, 16, 24 and 32 bit PCM files using sf_read_short() +and always have the optimal behaviour. +
+ + + ++When you configure and compile libsndfile, it uses the /bin/sh shell for a number +of tasks (ie configure script and libtool). +Older versions of OS X (10.2?) shipped a really crappy Bourne shell as /bin/sh +which resulted in really slow compiles. +Newer version of OS X ship GNU Bash as /bin/sh and this answer doesn't apply in that +case. +
++To fix this I suggest that you install the GNU Bash shell, rename /bin/sh to +/bin/sh.old and make a symlink from /bin/sh to the bash shell. +Bash is designed to behave as a Bourne shell when is is called as /bin/sh. +
++When I did this on my iBook running MacOS X, compile times dropped from 13 minutes +to 3 minutes. +
+ + + ++It seems that the Solaris Bourne shell disagrees with GNU libtool. +
++To fix this I suggest that you install the GNU Bash shell, rename /bin/sh to +/bin/sh.old and make a symlink from /bin/sh to the bash shell. +Bash is designed to behave as a Bourne shell when is is called as /bin/sh. +
+ + + ++This problem is bigger than it may seem at first. +
++For a stereo file, it is a pretty safe bet that a simple interleaving/de-interleaving +could satisfy most users. +However, for files with more than 2 channels this is unlikely to be the case. +If the user has a 4 channel file and want to play that file on a stereo output +sound card they either want the first 2 channels or they want some mixed combination +of the 4 channels. +
++When you add more channels, the combinations grow exponentially and it becomes +increasingly difficult to cover even a sensible subset of the possible combinations. +On top of that, coding any one style of interleaver/de-interleaver is trivial, while +coding one that can cover all combinations is far from trivial. +This means that this feature will not be added any time soon. +
+ + + ++When you want to store temporary data there are a number of requirements; +
++The format which best meets these requirements is AU, which allows data to be +stored in any one of short, int, float and double (among others) formats. +
++For instance, if an application uses float data internally, its temporary files +should use a format of (SF_ENDIAN_CPU | SF_FORMAT_AU | SF_FORMAT_FLOAT) which +will store big endian float data in big endian CPUs and little endian float data +on little endian CPUs. +Reading and writing this format will not require any conversions or byte swapping +regardless of the host CPU. +
+ + + + ++libsndfile uses the pkg-config (man pkg-config) method of registering itself with the +host system. +The best way of detecting its presence is using something like this in configure.ac +(or configure.in): +
+
+ PKG_CHECK_MODULES(SNDFILE, sndfile >= 1.0.2, ac_cv_sndfile=1, ac_cv_sndfile=0)
+
+ AC_DEFINE_UNQUOTED([HAVE_SNDFILE],${ac_cv_sndfile},
+ [Set to 1 if you have libsndfile.])
+
+ AC_SUBST(SNDFILE_CFLAGS)
+ AC_SUBST(SNDFILE_LIBS)
+
++This will automatically set the SNDFILE_CFLAGS and SNDFILE_LIBS +variables which can be used in Makefile.am like this: +
++ SNDFILE_CFLAGS = @SNDFILE_CFLAGS@ + SNDFILE_LIBS = @SNDFILE_LIBS@ ++
+If you install libsndfile from source, you will probably need to set the +PKG_CONFIG_PATH environment variable as suggested at the end of the +libsndfile configure process. For instance on my system I get this: +
++ -=-=-=-=-=-=-=-=-=-= Configuration Complete =-=-=-=-=-=-=-=-=-=- + + Configuration summary : + + Version : ..................... 1.0.5 + Experimental code : ........... no + + Tools : + + Compiler is GCC : ............. yes + GCC major version : ........... 3 + + Installation directories : + + Library directory : ........... /usr/local/lib + Program directory : ........... /usr/local/bin + Pkgconfig directory : ......... /usr/local/lib/pkgconfig + + Compiling some other packages against libsndfile may require + the addition of "/usr/local/lib/pkgconfig" to the + PKG_CONFIG_PATH environment variable. ++ + + + +
+The pkg-config program makes finding the correct compiler flag values and +library location far easier. +During the installation of libsndfile, a file named sndfile.pc is installed +in the directory ${libdir}/pkgconfig (ie if libsndfile is installed in +/usr/local/lib, sndfile.pc will be installed in +/usr/local/lib/pkgconfig/). +
++In order for pkg-config to find sndfile.pc it may be necessary to point the +environment variable PKG_CONFIG_PATH in the right direction. +
++ export PKG_CONFIG_PATH=/usr/local/lib/pkgconfig ++ +
+Then, to compile a C file into an object file, the command would be: +
++ gcc `pkg-config --cflags sndfile` -c somefile.c ++
+and to link a number of objects into an executable that links against libsndfile, +the command would be: +
++ gcc `pkg-config --libs sndfile` obj1.o obj2.o -o program ++ + + + +
+This has been added for version 1.0.13. +
+ + + + ++This is caused by the fact that the conversion from 16 bit short to float is +done by dividing by 32768 (0x8000 in hexadecimal) while the conversion from +float to 16 bit short is done by multiplying by 32767 (0x7FFF in hex). +So for instance, a value in a 16 bit PCM file of 20000 gets read as a floating +point number of 0.6103515625 (20000.0 / 0x8000). +Converting that back to a 16 bit short results in a value of 19999.3896484375 +(0.6103515625 * 0x7FFF) which then gets rounded down to 19999. +
++You will notice that for this particular case, the error is 1 in 20000 or +0.005%. +Interestingly, for values of less than 16369, dividing by 0x8000 followed +by multiplying by 0x7FFF and then rounding the result, gives back the +original value. +It turns out that as long as the host operating system supplies the 1999 ISO +C Standard functions lrintf and lrint (or a replacement has +been supplied) then the maximum possible error is 1 in 16369 or about 0.006%. +
++Regardless of the size of the error, the reason why this is done is rather +subtle. +
++In a file containing 16 bit PCM samples, the values are restricted to the range +[-32768, 32767] while we want floating point values in the range [-1.0, 1.0]. +The only way to do this conversion is to do a floating point division by a value +of 0x8000. +Converting the other way, the only way to ensure that floating point values in +the range [-1.0, 1.0] are within the valid range allowed by a 16 bit short is +to multiply by 0x7FFF. +
++Some people would say that this is a severe short-coming of libsndfile. +I would counter that anybody who is constantly converting back and forth +between 16 bit shorts and normalised floats is going to suffer other losses +in audio quality that they should also be concerned about. +
++Since this problem only occurs when converting between integer data on disk and +normalized floats in the application, it can be avoided by using something +other than normalized floats in the application. +Alternatives to normalized floats are the short and int data +types (ie using sf_read_short or sf_read_int) or using un-normalized floats +(see + + SFC_SET_NORM_FLOAT). +
++Another way to deal with this problem is to consider 16 bit short data as a +final destination format only, not as an intermediate storage format. +All intermediate data (ie which is going to be processed further) should be +stored in floating point format which is supported by all of the most common +file formats. +If floating point files are considered too large (2 times the size of a 16 bit +PCM file), it would also be possible to use 24 bit PCM as an intermediate +storage format (and which is also supported by most common file types). +
+ + + + ++This is actually a Winamp problem. +The official Microsoft spec suggests that the 'fmt ' chunk should be 18 bytes. +Unfortunately at least one of Microsoft's own applications (Sound Recorder on +Win98 I believe) did not accept 18 bytes 'fmt ' chunks. +
++Michael Lee did some experimenting and found that: +
++ I have checked that Windows Media Player 9, QuickTime Player 6.4, + RealOne Player 2.0 and GoldWave 5.06 can all play u-law files with + 16-byte or 18-byte 'fmt ' chunk. Only Winamp (2.91) and foobar2000 + are unable to play u-law files with 16-byte 'fmt ' chunk. ++ +
+Even this is a very small sampling of all the players out there. +For that reason it is probably not a good idea to change this now because there +is the risk of breaking something that currently works. +
+ + + + ++An itemtt> is a single sample of the data type you are reading; ie a +single short value for sf_read_short or a single float +for sf_read_float. +
+ +For a sound file with only one channel, a frame is the same as a item (ie a +single sample) while for multi channel sound files, a single frame contains a +single item for each channel. + + ++Here are two simple, correct examples, both of which are assumed to be working +on a stereo file, first using items: +
+ ++ #define CHANNELS 2 + short data [CHANNELS * 100] ; + sf_count items_read = sf_read_short (file, data, 200) ; + assert (items_read == 200) ; ++ +
+and now readng the exact same amount of data using frames: +
+ ++ #define CHANNELS 2 + short data [CHANNELS * 100] ; + sf_count frames_read = sf_readf_short (file, data, 100) ; + assert (frames_read == 100) ; ++ + + + +
+This is somewhat complicated. +First some background. +
+ ++SD2 files are native to the Apple Macintosh platform and use features of +the Mac filesystem (file resource forks) to store the file's sample rate, +number of channels, sample width and more. +When you look at a file and its resource fork on Mac OS X it looks like +this: +
+ ++ -rw-r--r-- 1 erikd erikd 46512 Oct 18 22:57 file.sd2 + -rw-r--r-- 1 erikd erikd 538 Oct 18 22:57 file.sd2/rsrc ++ +
+Notice how the file itself looks like a directory containing a single file +named rsrc. +When libsndfile is compiled for MacOS X, it should open (for write and read) +SD2 file with resource forks like this without any problems. +It will also handle files with the resource fork in a separate file as +described below. +
+ ++When SD2 files are moved to other platforms, the resource fork of the file +can sometimes be dropped altogether. +All that remains is the raw audio data and no information about the number +of channels, sample rate or bit width which makes it a little difficult for +libsndfile to open the file. +
+ ++However, it is possible to safely move an SD2 file to a Linux or Windows +machine. +For instance, when an SD2 file is copied from inside MacOS X to a windows +shared directory or a Samba share (ie Linux), MacOS X is clever enough to +store the resource fork of the file in a separate hidden file in the +same directory like this: +
++ -rw-r--r-- 1 erikd erikd 538 Oct 18 22:57 ._file.sd2 + -rw-r--r-- 1 erikd erikd 46512 Oct 18 22:57 file.sd2 ++ +
+Regardless of what platform it is running on, when libsndfile is asked to +open a file named "foo" and it can't recognize the file type from +the data in the file, it will attempt to open the resource fork and if +that fails, it then tries to open a file named "._foo" to see if +the file has a valid resource fork. +This is the same regardless of whether the file is being opened for read +or write. +
+ ++In short, libsndfile should open SD2 files with a valid resource fork on +all of the platforms that libsndfile supports. +If a file has lost its resource fork, the only option is the open the file +using the SF_FORMAT_RAW option and guessing its sample rate, channel count +and bit width. +
+ ++Occasionally, when SD2 files are moved to other systems, the file is + BinHexed +which wraps the resource fork and the data fork together. +For these files, it would be possible to write a BinHex parser but +there is not a lot to gain considering how rare these BinHexed SD2 +files are. +
+ + + ++Unfortunately no. +libsndfile contains code written by other people who have agreed that their +code be used under the GNU LGPL but no more. +Even if they were to agree, there would be significant difficulties in +dividing up the payments fairly. +
+ ++The only way you can legally use libsndfile as a statically linked +library is if your application is released under the GNU GPL or LGPL. +
+ + + ++libsndfile is being used by large numbers of people all over the world +without any problems like this. That means that it is much more likely +that your code has a bug than libsndfile. However, it is still possible +that there is a bug in libsndfile. +
++To figure out whether it is your code or libsndfile you should do the +following: +
++If compiler X is a C++ compiler then no. +C and C++ are different enough to make writing code that compiles as valid C +and valid C++ too difficult. +I would rather spend my time fixing bugs and adding features. +
+ ++If compiler X is a C compiler then I will do what I can as long as that does +not hamper the correctness, portability and maintainability of the existing +code. +It should be noted however that libsndfile uses features specified by the 1999 +ISO C Standard. +This can make compiling libsndfile with some older compilers difficult. +
+ + + ++Yes, libsndfile can read files from pipes. +Unfortunately, the write case is much more complicated. +
+ ++File formats like AIFF and WAV have information at the start of the file (the +file header) which states the length of the file, the number of sample frames +etc. +This information must be filled in correctly when the file header is written, +but this information is not reliably known until the file is closed. +This means that libsndfile cannot write AIFF, WAV and many other file types +to a pipe. +
+ ++However, there is at least one file format (AU) which is specifically designed +to be written to a pipe. +Like AIFF and WAV, AU has a header with a sample frames field, but it is +specifically allowable to set that frames field to 0x7FFFFFFF if the file +length is not known when the header is written. +The AU file format can also hold data in many of the standard formats (ie +SF_FORMAT_PCM_16, SF_FORMAT_PCM_24, SF_FORMAT_FLOAT etc) as well as allowing +data in both big and little endian format. +
+ ++See also FAQ Q6. +
+ + + ++Yes, but you must do two separate configure/build/test runs; one on PowerPC +and one on Intel. +It is then possible to merge the binaries into a single universal binary using +one of the programs in the Apple tool chain. +
+ ++It is not possible to build a working universal binary via a single +compile/build run on a single CPU. +
+ ++The problem is that the libsndfile build process detects features of the CPU its +being built for during the configure process and when building a universal binary, +configure is only run once and that data is then used for both CPUs. +That configure data will be wrong for one of those CPUs. +You will still be able to compile libsndfile, and the test suite will pass on +the machine you compiled it on. +However, if you take the universal binary test suite programs compiled on one +CPU and run them on the other, the test suite will fail. +
+ ++Part of the problem is the the CPU endian-ness is detected at configure time. +Yes, I know the Apple compiler defines one of the macros __LITTLE_ENDIAN__ +and __BIG_ENDIAN__, but those macros are not part of the 1999 ISO C Standard +and they are not portable. +
+ ++Endian issues are not the only reason why the cross compiled binary will fail. +The configure script also detects other CPU specific idiosyncrasies to provide +more optimized code. +
+ ++Finally, the real show stopper problem with universal binaries is the problem +with the test suite. +libsndfile contains a huge, comprehensive test suite. +When you compile a universal binary and run the test suite, you only test the +native compile. +The cross compiled binary (the one with the much higher chance of having +problems) cannot be tested. +
+ ++Now, if you have read this far you're probably thinking there must be a way +to fix this and there probably is. +The problem is that its a hell of a lot of work and would require significant +changes to the configure process, the internal code and the test suite. +In addition, these changes must not break compilation on any of the platforms +libsndfile is currently working on. +
+ + + + ++There's a very good reason for this. +I will only distribute things that I actually have an ability to test and +maintain. +Project files for a bunch of different compilers and Integrated Development +Environments are simply too difficult to maintain. +
+ ++The problem is that every time I add a new file to libsndfile or rename an +existing file I would have to modify all the project files and then test that +libsndfile still built with all the different compilers. +
+ ++Maintaining these project files is also rather difficult if I don't have access +to the required compiler/IDE. +If I just edit the project files without testing them I will almost certainly +get it wrong. +If I release a version of libsndfile with broken project files, I'll get a bunch +of emails from people complaining about it not building and have no way of +fixing or even testing it. +
+ ++I currently release sources that I personally test on Win32, Linux and +MacOS X (PowerPC) using the compiler I trust (GNU GCC). +Supporting one compiler on three (actually much more because GCC is available +almost everywhere) platforms is doable without too much pain. +I also release binaries for Win32 with instructions on how to use those +binaries with Visual Studio. +As a guy who is mainly interested in Linux, I'm not to keen to jump through +a bunch of hoops to support compilers and operating systems I don't use. +
+ ++So, I hear you want to volunteer to maintain the project files for Some Crappy +Compiler 2007? +Well sorry, that won't work either. +I have had numerous people over the years offer to maintaining the project +files for Microsoft's Visual Studio. +Every single time that happened, they maintained it for a release or two and +then disappeared off the face of the earth. +Hence, I'm not willing to enter into an arrangement like that again. +
+ + + ++MP3 is not supported for one very good reason; doing so requires the payment +of licensing fees. +As can be seen from + + mp3licensing.com +the required royalty payments are not cheap. +
+ ++Yes, I know other libraries ignore the licensing requirements, but their legal +status is extremely dubious. +At any time, the body selling the licenses could go after the authors of those +libraries. +Some of those authors may be students and hence wouldn't be worth pursuing. +
+ ++However, libsndfile is released under the name of a company, Mega Nerd Pty Ltd; +a company which has income from from libsamplerate licensing, libsndfile based +consulting income and other unrelated consulting income. +Adding MP3 support to libsndfile could place that income would be under legal +threat. +
+ ++Fortunately, Ogg Vorbis exists as an alternative to MP3. +Support for Ogg Vorbis was added to libsndfile (mostly due to the efforts of +John ffitch of the Csound project) in version 1.0.18. +
+ + + + ++Here is a checklist of things you need to do to make sure your use of libsndfile +in a closed source or commercial project complies with the license libsndfile is +released under, the GNU Lesser General Public License (LGPL): +
+ ++Currently the precompiled windows binaries are thoroughly tested on Windows XP. +As such, they should also work on Win2k and Windows Vista. +They may also work on earlier versions of Windows. +
+ ++Since version 0.1.18 I have also been releasing precompiled binaries for Win64, +the 64 bit version of Windows. +These binaries have received much less testing than the 32 bit versions, but +should work as expected. +I'd be very interested in receiving feedback on these binaries. +
+ + + ++
+ ++Since version 1.0.21 the top level Makefile has an extra make target, +'test-tarball'. +Building this target creates a tarball called called: +
+ ++in the top level directory. +This tarball can then be copied to the target platform. +Once untarred and test script test_wrapper.sh can be run from +the top level of the extracted tarball. +
+ + +
+ The libsndfile home page is here :
+
+ http://www.mega-nerd.com/libsndfile/.
+
+Version : 1.0.25
+