> > > > for you.. I believe it should be possible to get 50 fps progressive
> > > > output
> > > > from 50i material using vdpau deinterlacing.
> > yes, my PCI Geforce 8400 card on GPU G98 can do 10...@50 but with the
> > simplest bob deinterlacing only. I prefer to use
> > more advanced deinterlacing algorithm - temporal_spatial or temporal. But
> > with them I have jerky video
> Hm, do you think you were only able to use simple deinterlacing because
> of limitations in CPU speed, bus bandwidth (only PCI - I've been
> thinking about one of the 8400 GSs myself so I can play with it in my
> PCI-only VDR box..), or power within the G98 GPU itself?
I don't know exactly
there's report that on G98 temporal works well. I believe in it
So, I suppose that PCI bus is limited
In order for either VDP_VIDEO_MIXER_FEATURE_DEINTERLACE_TEMPORAL or
to operate correctly, the application must supply at least 2 past and 1 future
fields to each VdpMixerRender call. If
those fields are not provided, the VdpMixer will fall back to bob
> > > .. But you might get better results if you outputted 1:1 interlaced
> > > material
> > > using 1080i50 mode and let the LCD tv do the deinterlacing..
> > Yes, exactly what I want to try, But I couldn't run properly 10...@50 -
> > even in xorg.log I have the report about
> > validated 10...@50 mode - my LCD TV Philips 9703PFL reported about 1080p
> > source from hdmi
> :( Native output and letting the TV do the grunt work would be my
> prefernece, too - in what way doesn't it work? No output at all or very
> juddery output ?
I could reach more less good quality output for 10...@50 , but I decided to
come back to 10...@50 because that mode was
better than 10...@50 (I repeat mt TV set always informed me about 1080p mode ,
never - about of 1080i)
vdr mailing list