Re: [Ayatana] Fwd: CHI Day 4 continued

[Jim Rorie <jfrorie@xxxxxxxxx>]

On Thu, 2010-04-15 at 16:40 +0100, Conor Curran wrote: 

> >State type information that Charlene is referring to would be something
> >novel like the musical pitch of the entire sound theme could being
> >lowered by an octave as battery power drains to a critical level.
> >Something easily done with a DFT audio processing routine. 
> 
> This would be very nice indeed but are you implying we should schedule
> real-time signal processing of a sound event on a machine that is
> already dying because of lack of power :)

I should think it wouldn't cause too much overhead, as the duty cycle
should be pretty low.   Anyone generating continuous sound events has
got bigger problems than a dying battery. IIRC, a good DFT transform for
octave shift isn't too expensive, computationally.

> 
> It is funny you mention this since we have been talking about the
> possibility using real-time processing on sound events using simple
> Gstreamer pipelines or Faust DSPs.

Another approach would be to preprocess the media before packaging, but
with the number of combinations, this might be prohibitive.

I'm not sure I would leave it to the designer.  The free desktop spec is
pretty big to begin with.  But, opinions might vary.

> We came to a point during the week where we realized we needed more
> community feedback before we committed to a brief hence the urge to
> organise a session at UDS about sound themes. 
> More info on session to follow. 
> 
>