DirectFB - Home of the pragmatist Roadmap


[directfb-dev] Re: directfb on xscale
Mailing List archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[directfb-dev] Re: directfb on xscale



Quoting Clemens Kirchgatterer:
> Sven Neumann <neo@convergence.de> wrote:
> 
> > Clemens Kirchgatterer <c.kirchgatterer@amatic.com> writes:
> > 
> > > my next aproach was a basic DFB test programm, that loads 2 png
> > > files
> > > (each 256 colors without alpha, i hope) and tried to blit one as the
> > > screen background and the other over that background. this is
> > > awfully
> > > slow. i can follow the blitt easily with naked eye. (1 fps, i think)
> > 
> > Do you render directly from the image provider to the destination
> > surface or do you use an offscreen surface the same pixel format than
> > the destination? There are quite a few things you might be doing wrong
> > (or rather ineffeciently), so you will have to give us some more
> > details. Perhaps show us some code snippets?
> 
> i just made some progression.
> 
> when i set the pixelformat "explicitly" to DSPF_LUT8 on the screen and
> the source surface, framerate suddenly raises to 42!FPS. :-)
> why isn't the pixelformat set corrctly for best performance? is this a
> bug in DFB? or a limitation of the xscale fb?

IDirectFBImageProvider::GetSurfaceDescription() normally uses the pixel format
of the primary layer. If the image has an alpha channel it uses DSPF_ARGB.

-- 
Best regards,
  Denis Oliver Kropp

.------------------------------------------.
| DirectFB - Hardware accelerated graphics |
| http://www.directfb.org/                 |
"------------------------------------------"

                            Convergence GmbH


-- 
Info:  To unsubscribe send a mail to ecartis@directfb.org with 
"unsubscribe directfb-dev" as subject.



Home | Main Index | Thread Index


directfb.org / Development / Old Archives