[opendtv] Re: Samsung Rep admits HDMI won't support 1080p until v.1.3
- From: Tom Barry <trbarry@xxxxxxxxxxx>
- To: opendtv@xxxxxxxxxxxxx
- Date: Tue, 08 Nov 2005 23:30:08 -0500
Craig Birkmaier wrote:
> This is one of the reasons that we pushed for lower spatial / higher
> temporal rate formats for DTV broadcasting. You can distribute
> spectacular pictures by encoding about 1/2 million pixels - e.g. 1024
> x 576@ 60P. You can still acquire the source at 1280 x 720@60P, then
> simply re-sample prior to compression for emission.
>
Craig -
Curiously I've had a chance to test this assertion in the last couple
days. Over the weekend I bought a $900 XGA front projector at CostCo to
replace my dying RPTV. This is a DLP with a 4:3 1024x768 native
resolution. Since it automatically switches to 16:9 with an HD signal
that means it's really using a 1024x576 (wide-PAL) resolution for movies.
Now everyone on this list has probably heard my laments that very few
broadcast shows seem to have an effective resolution much more than this
anyway. And since I equally will use this projector as an easy-eye
computer monitor it seemed an economical choice.
It has been so far.
I'm currently projecting on a white wall about 9-10' from my eyes,
watching an image about 3' high (when 16:9) so I'm about 3 screen
heights away. And I don't notice any loss of detail compared to my
previous 1080i RPTV. However I'm just on the edge of being able to see
the pixel structure at that distance. For instance in computer mode as
I type this I can see the slight jagginess of the computer fonts. But
for HD video images that is much harder.
However when I first set it up I had the projector set back a bit
further because it was temporarily convenient. That made the image
about 10-20% larger, probably just over the size/distance threshold
where I could pretend it was pixel free. For my vision and seating
distance I would personally need a bit more display resolution if I
wanted to keep a 4' high 16:9 image.
I'm not sure whether that would also mean a higher transmission
resolution for current HD broadcast. But I am certain sure I'd need it
if I was really watching full resolution 1080p, say with possible new
HD/DVD's of some flavor.
I also notice with the very large screen size that I do NOT want a full
field of vision for web browsing. It makes me more aware of why reading
a magazine with separate narrower columns seems easier, not zooming my
eyes back and forth across 80 some inches of wall to read each line.
Luckily Windows can make windows where this is needed.
All in all I'm pretty happy here, and do still tend to believe that 576p
can indeed match most current HD as a delivery format, even if I know we
could theoretically do better.
- Tom (who still expects to buy a 1080p display when it's priced right)
> At 10:09 AM -0800 11/7/05, dan.grimes@xxxxxxxx wrote:
>
>>"What we need is HD source that has not been trashed by excessive
>>compression in the distribution chain." --Craig Birkmaier
>>
>>Amen!
>>
>>But don't you think that it would be good to have 1080@30p production? I
>>realize that this requires more bits, but interlacing is also a big
>>problem. Or do you think that 720@30p is good enough? While I hope for
>>1080@30p production, I would rather have only 720p production if it removed
>>most of the artifacts.
>
>
> 30P production is certainly possible today, but it is rarely used.
> Like 24P it presents problems with rapid motion, and one must be very
> careful with camera moves and zooms. Generally speaking, if a
> producer is looking for that "look" they will choose 24P.
>
> That being said, a great deal of streaming media content that is
> available via the web is limited to 30P, even if the original source
> was 24P. The need to conserve bits for web streaming may lead to
> greater use of 30P, but for now, it seems that 24P and 60P are the
> most common HD acquisition formats for the U.S. - whoops, I almost
> forgot 1080@30i.
>
>
>>I want to mention (though beating a dead horse) that I was so frustrated
>>with the compression artifacts watching NASCAR yesterday that I went to my
>>old NTSC TV. But the artifacts were so bad, even the low resolution
>>picture looked bad! The distribution chain needs fixin'!
>
>
> Yup. ANY 60P format with high motion is still difficult to compress
> into 18 Mbps for an ATSC channel, and if that channel is being shared
> for a multicast or other services (like USDTV channels) then it is
> likely that all of the programs will be compromised. H.264 can help a
> great deal, but it is not an available tool for OTA broadcasts.
>
> This is one of the reasons that we pushed for lower spatial / higher
> temporal rate formats for DTV broadcasting. You can distribute
> spectacular pictures by encoding about 1/2 million pixels - e.g. 1024
> x 576@ 60P. You can still acquire the source at 1280 x 720@60P, then
> simply re-sample prior to compression for emission.
>
> Unfortunately, most broadcasters are still hopelessly hooked on
> "FORMATS;" they have a very difficult time with the concept that they
> are just delivering bits that can represent pictures of any size or
> frame rate. In this, the web video folks are light years ahead.
>
> The folks with plenty of bandwidth abuse what they've got and deliver
> crap, while the folks trying to build a business with a distribution
> network that is very constrained, are the ones working hard to
> improve delivered video quality.
>
> Go figure.
>
> Regards
> Craig
>
>
> ----------------------------------------------------------------------
> You can UNSUBSCRIBE from the OpenDTV list in two ways:
>
> - Using the UNSUBSCRIBE command in your user configuration settings at
> FreeLists.org
>
> - By sending a message to: opendtv-request@xxxxxxxxxxxxx with the word
> unsubscribe in the subject line.
>
>
----------------------------------------------------------------------
You can UNSUBSCRIBE from the OpenDTV list in two ways:
- Using the UNSUBSCRIBE command in your user configuration settings at
FreeLists.org
- By sending a message to: opendtv-request@xxxxxxxxxxxxx with the word
unsubscribe in the subject line.
Other related posts:
