[integra-dev] integralive case study: "lichtweiss"

[Jamie Bullock]

On 14 Jul 2010, at 12:03, IOhannes m zmoelnig wrote:

> On 2010-07-14 12:40, Jamie Bullock wrote:
> 
>> Other than where I have specifically answered above, I think most of the problems of this piece can be solved with a well-designed SoundfilePlayer module. We are currently in a coding sprint in Birmingham, and I have set aside this afternoon for 3 developers to look at this.
>> 
> 
> i talked to peter about that yesterday, and i personally think that,
> while a SoundFilePlayer module is definitely important to have, it is
> not the solution for this piece (which would require a SoundFileRecorder)
> 

Yes absolutely, sorry I didn't make this clear. I'm thinking more about a SimpleSampler that incorporates record and playback with multiple buffers. Whether it is 'file' or not is as you say just an implementation issue.

> as i understand IL, it is meant to provide an abstraction layer
> independent of the used technology.
> whether the 4 moves of the piece are recorded on harddisk, into RAM or
> into an (admittedly very long) wire, is an implementation detail.
> when i will perform "lichtweiss" with my to-be-discovered backend
> "QuantuMusic 7" in 2307, i don't want to reimplement "SoundFilePlayer"
> just because somebody thought that the delayed playback of moves could
> nicely be done with files (or by carving the score into stone).
> 

Yes, that's the grand idea, or at least abstract 'as much as possible'. I'm sure there will be pieces where the composer provides a bunch of wav files and we need to load and play them, and we can't avoid implementation specifics of files etc, but this isn't the case with Lichtweiss. I guess the trick is to be pragmatic about matching the system ideals to specific musical demands.

> just my 2¢
> (i might be completely off the hook; it's pretty hot here)

Very welcome! (... and wishing it was so hot here)

Jamie