[vdr] OT: Best DVI Video Card for Xine
pasi.juppo at iki.fi
Sat Feb 26 23:14:53 CET 2005
jori.hamalainen at teliasonera.com wrote:
>> I fear that 10 years is way too short time for anything to die on
>> TV standards.
> Well not to die, but become obsolete. As you know DVB is much more
> than TV standard, but it is very hard to use terms when there is no
> term. DVB can broadcast HDTV and AC3/DTS, so DVB will stay, but I
> hope that for new displays people are acquiring, they want same
> quality from sources. And I think that in 10 years majority of sold
> TV sets could be plasma/tft/some other compared today's TV's with big
> tubes and "720x576" grid for electron gun.
Most likely majority of sold "TV sets" will provide much better
resolution than old ones but I simply cannot see it happen that TV
companies start to broadcast better quality material without getting
paid for it - at least extra. Hopefully I'm wrong here..
> I think TV business is currently changing. Youth are not so keen to
> live TV, a new generation of people are growing with TiVo, Replay TV,
> VDR, Myth TV etc. So they are used to watch TV when appropriate, not
> when episode airs. And cap is very short to download content from
> computer network. Both are sort of time shifting.
Yes, that's very true. TV business is definately facing a lot of
problems in the near future when trying to get people back to watching TV.
> So perhaps one day we log on to studio's site, and download episode
> of your favorite TV-series (already dubbed to German by studio :-)
> and subtitles for all languages. And all in HD format protected by
> DRM, and you can watch episode when suitable to you. Why you have to
> wait in Europe 1 year for that episode to be broadcasted. It is
> reality today, you can have it from Internet after episode has aired
> in US. And naturally in HD format.
Propably possible someday. I'm not very keen about this due to DRM..
>> There should be no need for upsampling of legacy stuff nowadays -
>> it's actually bad for transmission purposes (wastes bandwidth).
>> HDTV receivers are required to handled rather great amounts of data
>> so when receiving SDTV material all scalings, if needed, would be
>> done at customer end.
> But when HDTV channel broadcasts 15Mbps HDTV that bandwidth must be
> allocated on transponder. Of course on HDTV PID SDTV could be sent,
> but you cannot add any temporary channel etc. to released bandwidth
> from HDTV. Of course more bandwidth could be allocated to SDTV
> channel(s) on same transponder if no HDTV broadcast is available. But
> I think it would be inconvenient for people to switch channels when
> HDTV broadcast ends, just to see same channel on SDTV side.
Maybe the whole channel idea needs to be re-thought. Instead of channels
you can watch different programs. Total amount of bandwidth is split
between programs. This would allow e.g. HDTV level movie to be sent
while at the same time there are News, TV shop etc. that don't require
that much bandwidth. For the end user "channels" are still used to keep
things common. Oh well, now I'm getting into dream land :)
> But I am against of up/downsampling. IMHO is much better to adapt
> display frequency to content, not vice versa (content to display).
> This way we could see 1080p24 material as good quality as possible,
> and for example not 1080i50 with 4% PAL-speedup, which changes also
> audio pitch. Or video judders for down converting 60i to 50i.
Makes me wonder why this is not used. It would be cheaper for broadcasters.
> I have find it difficult to find proper settings for high quality
> playback. Best have been (on windows side) using 72,75 and 90 Hz
> frame rates (adapted for content: cinema, pal or ntsc) and with
> directshow reclock-plugin. With this I've been able to eliminate
> judder from videos, or sound jumps. But immediately when source is
> 24fps and monitor is 75Hz, picture or sound jumps. You can choose
> which one.. :-)
Shouldn't be that difficult to write small SW that checks the framerate
of the video and adjusts monitor freq according to it.
More information about the vdr