This interface provides a means for the MovieMaker to interact with the
current sound-producing model to be recorded. All models that produce sound
should implement this interface if the audio is to be automatically merged
with the movie. *
In the current implementation, the audio model is expected to produce a raw
audio file. The file path will be saved to a MovieOptions data structure
stored in myMain.getMovieOptions() The raw file will be converted by the
MovieMakerDialog and then merged with the movie.
The setRenderAudioToFile is called right before movie recording starts. This
method should initialize the audio model to produce a raw audio file
according to the path set in the MovieOptions. Then, if necessary, during the
onStop method it should do any post processing of the newly created audio
file (e.g. normalization). The MovieMaker will then convert the raw file into
a .wav, and merge it with the movie.
After the movie is made, setRenderAudioToFile can be called again with
'false', so that the model goes back to producing audio the normal way (i.e.
through the Soundcard, etc.), and no new output files are produced when the
model is run normally.
As a reference, see the implementation of HasAudio by the VTNTDemo in
If future models will produce Wave (or other) files directly, the
MovieOptions class should be expanded to allow for these, and no file
conversion will be necessary during the movie making process
Additionally, an option to write out the audio as a text file could also be
set by the MovieMakerDialog. This option expects the audio model to output an
ASCII data file of the audio samples.
TODO: implement adherence to specifications such as sample/bit rates, etc, if
necessary. It all depends on how acoustic models produce the audio output.