Hi, > When running at say 1080p and the framing is side-by-side or top-bottom, are > the contents of all active display layers surfaces expected to be > automatically downscaled by half?
Yes. The hardware would take care of that. > If so, does all of the active display layers being downscaled include those > configured with DLOP_LR_MONO and DLOP_STEREO, as well as ones without either > flag? The one without (if any) would just need to be duplicated w/o applying any horizontal shift (0 parallax). > What about if the application wants to handle things more efficiently and > configure the layer at the appropriate half size, thus only rendering to the > actual half sized surface (of course just when the display mode is > side-by-side or top-bottom)? Is this possible? But then what about legacy applications that expect to just still render @ 1920x1080 (so the primary surface is expected to be @ HD). 32 MB of VRAM is the least amount required for stereoscopy :) Yes, that's too much. -Ilyes On Fri, Nov 4, 2011 at 2:05 AM, Strelchun, Timothy <timothy.strelc...@intel.com> wrote: > Hi guys, > > After reviewing the API and impl. changes for adding > stereoscopic support, some clarifications on intended usage and handling > behavior would be helpful. > > When running at say 1080p and the framing is side-by-side or top-bottom, are > the contents of all active display layers surfaces expected to be > automatically downscaled by half? > > If so, does all of the active display layers being downscaled include those > configured with DLOP_LR_MONO and DLOP_STEREO, as well as ones without either > flag? > > What about if the application wants to handle things more efficiently and > configure the layer at the appropriate half size, thus only rendering to the > actual half sized surface (of course just when the display mode is > side-by-side or top-bottom)? Is this possible? > > Thanks, > Timothy > > -- > > Timothy Strelchun > Software Engineering > NTG Service Provider Division (formally DHG) > Intel Corporation > > _______________________________________________ > directfb-dev mailing list > directfb-dev@directfb.org > http://mail.directfb.org/cgi-bin/mailman/listinfo/directfb-dev > > _______________________________________________ directfb-dev mailing list directfb-dev@directfb.org http://mail.directfb.org/cgi-bin/mailman/listinfo/directfb-dev
