sampling a wav file

Erik de Castro Lopo wrote:
..
> 
> About a decade ago I suggested to John Eaton that libsndfile be used
> to read audio file. Since libsndfile now does FLAC and Ogg/Vorbis (not
> publically released yet), it might be time to try this again.
> 
..
> 
> Good luck adding FLAC and Ogg/Vorbis support in pure Matlab/Octave
> code :-).

Well, it's not such an unlikely thought; not least since matlab or 
octave may well have been used to develop the dsp for those formats. 
However, just having full WAVE and AIFF/AIFC PCM support would be very 
useful. Even AIFF becomes more moot every day, what with Macs now all on 
Intel, so that AIFF is no longer even the "native" format (and IMO has 
little else to commend it).

> 
> I really do thing its time to push for Octave to do audio file
> reading via libsndfile whereever libsndfile is available.
> 


I agree. the question is, how? Ostensibly it would have to be an 
app-level builtin function, not a m-file. How much format info to return 
to the caller, and in what way (I would want for example the WAVEX 
channel mask at the very least so I can process/create various 
multi-channel files, and PEAK chunk support). There is separate code for 
au files  ("au.m. auload.m etc), suggesting that the Matlab/octave way 
is to have separate commands for each primary format type. There is 
therefore an API to be designed that gets as much out of libsndfile as 
is practicable, into Octave-ese. Ineluctably, you may well be the person 
best placed to do this; and once done, who knows - maybe the Mathworks 
will be happy to negotiate a licence for Matlab, just as they do for 
fftw etc.

In the short term, there is still a need to know how to hack wavread and 
wavwrite (I had to do it just to get endianness handled properly on the 
PPC Mac), so discussing the perils of soundfile format handling is still 
an unavoidable activity, sometimes!


Richard Dobson

Richard Dobson <richarddobson@blueyonder.co.uk> writes:

> Erik de Castro Lopo wrote:
> ..
>> About a decade ago I suggested to John Eaton that libsndfile be used
>> to read audio file. Since libsndfile now does FLAC and Ogg/Vorbis (not
>> publically released yet), it might be time to try this again.
>>
> ..
>> Good luck adding FLAC and Ogg/Vorbis support in pure Matlab/Octave
>> code :-).
>
> Well, it's not such an unlikely thought; not least since matlab or
> octave may well have been used to develop the dsp for those
> formats. However, just having full WAVE and AIFF/AIFC PCM support
> would be very useful. Even AIFF becomes more moot every day, what with
> Macs now all on Intel, so that AIFF is no longer even the "native"
> format (and IMO has little else to commend it).
>
>> I really do thing its time to push for Octave to do audio file
>> reading via libsndfile whereever libsndfile is available.
>>
>
>
> I agree. the question is, how? Ostensibly it would have to be an
> app-level builtin function, not a m-file. How much format info to
> return to the caller, and in what way (I would want for example the
> WAVEX channel mask at the very least so I can process/create various
> multi-channel files, and PEAK chunk support). There is separate code
> for au files  ("au.m. auload.m etc), suggesting that the Matlab/octave
> way is to have separate commands for each primary format type. There
> is therefore an API to be designed that gets as much out of libsndfile
> as is practicable, into Octave-ese. Ineluctably, you may well be the
> person best placed to do this; and once done, who knows - maybe the
> Mathworks will be happy to negotiate a licence for Matlab, just as
> they do for fftw etc.
>
> In the short term, there is still a need to know how to hack wavread
> and wavwrite (I had to do it just to get endianness handled properly
> on the PPC Mac), so discussing the perils of soundfile format handling
> is still an unavoidable activity, sometimes!
>
>
> Richard Dobson

I hate to say it since the group is so dead, but there is an octave newsgroup:

  comp.soft-sys.octave 

--Randy


-- 
%  Randy Yates                  % "She has an IQ of 1001, she has a jumpsuit
%% Fuquay-Varina, NC            %            on, and she's also a telephone."
%%% 919-577-9882                % 
%%%% <yates@ieee.org>           %        'Yours Truly, 2095', *Time*, ELO   
http://home.earthlink.net/~yatescr

Randy Yates wrote:
..
> 
> I hate to say it since the group is so dead, but there is an octave newsgroup:
> 
>   comp.soft-sys.octave 
> 


It's so dead (or still-born) it doesn't show up at all on my ISP news 
server, though plenty of com.soft-sys ones are. There is the 
octave-maintainers list, with whom I see Eric is already in dialogue. I 
am too much of a novice with octave (and even more so with matlab!), to 
be in a position to contribute. Can test, of course.

Richard Dobson

Richard Dobson wrote:

> It's so dead (or still-born) it doesn't show up at all on my ISP news
> server, though plenty of com.soft-sys ones are. There is the
> octave-maintainers list, with whom I see Erik is already in dialogue.

Yep. It turns out that it is much easier to build external Octave
modules than it used to be when I last looked.

That means, that I can probably dump the three .m files I have in
the Octave directory of the libsndfile tarball now and replace it
with code to generate a binary octave module (.oct) to whatever
is necessary.

Then I just have to get all the Linux distro package maintainers to
make sure to build a sndfile-octave package.

> Can test, of course.

I should have something for you by next week (big hacking day planned
for Saturday).

Erik
-- 
-----------------------------------------------------------------
Erik de Castro Lopo
-----------------------------------------------------------------
"All great truths begin as blasphemies."
-- George Bernard Shaw

Erik de Castro Lopo <erikd@mega-nerd.com> writes:

> Richard Dobson wrote:
>
>> It's so dead (or still-born) it doesn't show up at all on my ISP news
>> server, though plenty of com.soft-sys ones are. There is the
>> octave-maintainers list, with whom I see Erik is already in dialogue.
>
> Yep. It turns out that it is much easier to build external Octave
> modules than it used to be when I last looked.
>
> That means, that I can probably dump the three .m files I have in
> the Octave directory of the libsndfile tarball now and replace it
> with code to generate a binary octave module (.oct) to whatever
> is necessary.
>
> Then I just have to get all the Linux distro package maintainers to
> make sure to build a sndfile-octave package.
>
>> Can test, of course.
>
> I should have something for you by next week (big hacking day planned
> for Saturday).
>
> Erik

Thank you for your contributions, Eric. I'm a heavy user of octave.
-- 
%  Randy Yates                  % "My Shangri-la has gone away, fading like 
%% Fuquay-Varina, NC            %  the Beatles on 'Hey Jude'" 
%%% 919-577-9882                %  
%%%% <yates@ieee.org>           % 'Shangri-La', *A New World Record*, ELO
http://home.earthlink.net/~yatescr

Erik de Castro Lopo wrote:
..
> 
>>Can test, of course.
> 
> 
> I should have something for you by next week (big hacking day planned
> for Saturday).
> 

No urgency here - I am at ICMC Copenhagen all next week anyway. Anyone 
on this list likely to be there?


Richard Dobson