My thoughts:
1. You'll get the audio data in buffers. So this means the file modules
will need to handle a stdin stream. I could very easily do this. Another
option is to setup the file modules to accept a data buffer and read from
that rather then opening a file.
2. Output is much more simple. You open a aflibAudioFile with the input as
parent after setting the config to what you want. You should be able to
choose what format you want too. Write it to /dev/null. Currently we don't
have a raw 8kHz ULAW format. The sox module can handle it but it's not
currently listed in the exported formats. I'll fix that.
so
aflibAudioFile* output = new aflibAudioFile( input, "8kHz U-LAW",
"/dev/null", config, status);
then:
aflibData* data;
while( no_samples){
data = output->process(status,-1,no_samples,FALSE);
do_something_with_data( data );
}
On Monday 20 August 2001 10:52, Anthony Zawacki wrote:
> Hello,
>
> I'm seeking a bit of clarification on how to use OSALP in my application.
>
> A brief background:
> My module will receive audio data in buffers, read from an HTTP source
> (though I am not responsible for the reading.) This data must then be
> converted to 8kHz AU-ULAW with the header stripped. What I was planning to
> do is build an audio device that requires 8khz AU-ULAW. It would then
> start by stripping off the header, and write the data out to a set of
> buffers. Other portions of the application will read those buffers and do
> the actual writing to the audio device through it's API. I thought that an
> aflibAudioMemoryInput class would do the trick, but as far as I can
> understand it, the data must already be in the "correct" format as noted in
> section 4.2.10 of the Audio Library Programmers Manual. It's not clear to
> me what is the "correct" format.
>
> My input data could be one of:
> Raw (headerless) 8kHz 8-bit mu-law [PCM] single channel*
> Raw (headerless) 8kHz 8-bit a-law [PCM] single channel*
> WAV (RIFF header) 8kHz 8-bit mu-law [PCM] single channel*
> WAV (RIFF header) 8kHz 8-bit a-law [PCM] single channel*
> Others we specify as supported(MP3, ...)
>
> The item's marked by an (*) are required. I need to convert these formats
> to 8kHz Raw (headerless) 8-bit mu-law [PCM] single channel, which is what
> our audio device understands. The input data format I was hoping for is
> all of the required items, plus additional formats as supported by OSALP
> (plus any additional we may want to create supporting our audio device.)
>
> In general, I know which of the input types I will be getting from the
> content-type in the header. The question is, after I recognize that the
> file is MP3, how do I put an MP3 decoder into the audio chain so that the
> device writer can get the correct input.
>
> If anything is not clear, let me know, and I'll try to explain better.
>
> Thanks,
> Anthony
>
>
> _______________________________________________
> Osalp-dev mailing list
> Osa...@li...
> http://lists.sourceforge.net/lists/listinfo/osalp-dev
|
