[linux-dvb] Re: Generating progressive video

[Michael Plagge <plagge@onlinehome.de>]

On Tue, 2004-02-17 at 20:53, Michael Hunold wrote:
> Hello Michael,
> 
> On 02/17/04 15:33, Michael Plagge wrote:
> > On Tue, 2004-02-17 at 14:35, Michael Hunold wrote:
> >>On 02/16/04 23:38, Michael Plagge wrote:
> >>
> >>Anyway, what solution (ie. which gfx adapter) do you have in mind to put 
> >>the result to your beamer via DVI?
> 
> > Especially for this i have bought a matrox G450 with dvi-out since this
> > is the card which seems to be supported best by DirectFB and other
> > software using framebuffers (mplayer, xine, etc.). Therefore i want to
> > use the matrox gfx driver.
> 
> Have you checked if the DVI output is actually supported?
> 

Yes, it works perfectly. I was able to use mplayer and fbtv and tvtime
with DirectFB and/or framebuffer. I also can use a X-Server with this
card. 

> 
> If you omit the full-featured card and use software mpeg-decoding or 
> something like the cle266, the you can have a fully digial chain.
> 
> > Do you think it is reasonable to use a DXR3 card for hardware decoding
> > (as you can see i like the idea of hardware-decoding). The main reason
> > for preferring hardware-decoding is to circumvent the syncing problems
> > you are describing in the IDEAS-File in cvs.
> 
> You cannot manipulate the video after you've send the mpeg data to the 
> dxr3 decoder. There is no way to get the decoded video frames back (like 
> via Video4Linux with the DVB cards). If you don't have a hardware mpeg 
> decoder that actually implements your ideas of interlacing (which I 
> think doesn't exist), then you have to use software mpeg decoding.
> 

I see :-(

> > As you can see from my comments, i just start playing around with these
> > problems and maybe i am a little bit naive, but while reading you
> > comments there was an idea coming into my mind. Do you think it could be
> > possible, if i do the hole thing (software mpeg decoding and
> > deinterlacing) in software, to use the movement vectors from mpeg2 for a
> > motion adaptive deinterlacing algorithm (i hope this idea is not too
> > silly) 
> 
> Adaptive interlacing is a CPU burner already. It's probably a good 
> academic research topic if the motion vectors can be used for 
> deinterlacing, but I doubt that this will ever work for live tv.
> 


To subsume this discussion, i have to do everything in software (with
non CLE266 hardware) to get a 'high quality' solution. I will think  a
little bit more about my last point (using movement vectors). Could be a
good starting point for working with mpeg2. Do you think it would be
interesting to add deinterlacing support to your softmpeg library, and
is it possible to support your work on the library.

best wishes 
michael



-- 
Info:
To unsubscribe send a mail to ecartis@linuxtv.org with "unsubscribe linux-dvb" as subject.