Rich,
The kernel provides the floating point in software.
Regards,
Shane

On 3/3/07, Rich Rattanni <[EMAIL PROTECTED]> wrote:
> Claudio:
>    When we compiled FusionSound with IEEE floats enabled, all we got was a
> single pop on the speakers when the audio started to play.  Both my app and
> music.c produced these results.  I know there is no floating point unit in
> the ARM, I assume it is the C library to pick up the slack (assumption on my
> part).
>
> >>Another interesting test could be enabling IEEE floats and watch what
> >>happens.
> >>You can also use the example program called "music" to do such a test
> >>(use v/V key to decrease/increase the volume level).
>
>
>
>
> On 3/3/07, [EMAIL PROTECTED] <[EMAIL PROTECTED]> wrote:
> > Denis, (Sorry the last email got sent before I was complete do to
> > firefox crashing!) so I will redo it below:
> >
> > The target architecture is a Xscale (Intel's ARM) (PXA270).  The
> > compiler is GCC 4.1.1, we (Rich and I) tried with floats enabled and
> > the audio quality got even worse to the point were there was NO
> > discernible sound just noise.
> >
> > We have compiled many packages with our cross-compiler and have had no
> > issues.  I'm not sure what could be incorrect with the compiler but if
> > you have some suggestions or tests we would certainly be willing to
> > try them.
> > Regards,
> > Shane
> >
> > - Show quoted
> >
> >
> > On 3/3/07, [EMAIL PROTECTED] <[EMAIL PROTECTED]> wrote:
> > > Denis,
> > > The target architecture is a Xscale (Intel's ARM) (PXA270).  The
> > > compiler is GCC 4.1.1, we (Rich and I) tried with floats enabled and
> > > the audio q
> > >
> > > On 3/3/07, Denis Oliver Kropp < [EMAIL PROTECTED]> wrote:
> > > > Claudio Ciccani wrote:
> > > > > Rich Rattanni wrote:
> > > > >> Claudio:
> > > > >>      I first tried to adjust the volume while playing other audio
> types
> > > > >> (MP3 specifically) and there was no noticable improvement in
> quality.
> > > > >> As per your recommendation, the latest CVS snapshot of FusionSound
> was
> > > > >> pulled and compiled with --disable-precision
> > > > >> and --disable-dithering, and the system exhibited the same problem.
> > > > >> Terrible audio quality (output sounds clipped and other strange
> > > > >> artifacts) when volume was adjusted to anything besides 1.0.  What
> is
> > > > >> strange is that the volume value does not seem to matter.  0.5 or
> 60.0
> > > > >> both sound alike in both level and quality.
> > > > >
> > > > > So it must be a problem with the compilation (or better, the
> compiler).
> > > > > Indeed relooking at the code, I couldn't find a bug causing that
> strange
> > > > > behaviour.
> > > > > Please, could you tell me the target architecture (x86?) and the
> gcc's
> > > > > version?
> > > >
> > > > Does the API always require floats? Maybe emulation or library for it
> is
> > > > broken.
> > > >
> > > > --
> > > > Best regards,
> > > >    Denis Oliver Kropp
> > > >
> > > > .------------------------------------------.
> > > > | DirectFB - Hardware accelerated graphics |
> > > > | http://www.directfb.org/                 |
> > > > '------------------------------------------'
> > > >
> > > > _______________________________________________
> > > > directfb-users mailing list
> > > > [email protected]
> > > >
> http://mail.directfb.org/cgi-bin/mailman/listinfo/directfb-users
> > > >
> > >
> > >
> > > --
> > > Registered Linux User: #293401
> > >
> >
> >
> > --
> > Registered Linux User: #293401
> >
>
>
> _______________________________________________
> directfb-users mailing list
> [email protected]
> http://mail.directfb.org/cgi-bin/mailman/listinfo/directfb-users
>
>


-- 
Registered Linux User: #293401

_______________________________________________
directfb-users mailing list
[email protected]
http://mail.directfb.org/cgi-bin/mailman/listinfo/directfb-users

Reply via email to