Dieter wrote:
Are there any plans for advanced hardware de-interlacing?
There are no plans to NOT do it. If we can fit it in with all the
other stuff, we'll do it.
Interlace problems are incredibly annoying. Good interlace/de-interl=
ace
is not optional.
For video capture, which OGD1 doesn't do. OGA specifies graphics
rendering and video output. I'm talking with some other people about
a separate video-capture device.
I meant output. For example 1080i HDTV video (from digital stream/file
source) usually needs to be deinterlaced. Obviously this can be done also=
in
software so while hq hardware acceleration would be very nice, it is not
really a must. Deinterlacing 1080i with quality requires quite a lot of
processing capasity, and it seems like a problem that would parallelize
quite nicely on hardware.
That doesn't make a lot of sense to me. 1080i is an interlaced mode.
If it's to be deinterlaced, that would be done by the monitor/tv.
Huh? My TVs all expect to receive interlaced video. My computer monitors
all expect to receive progressive video. Some recent digital TVs can handle
either, but that is a limited subset of the displays that users will want
to use with OGC.
Let's say I have two files on disk. One is from ABC, and thus will be
mpeg2ts 720p. The other is from PBS, and will be mpeg2ts 1080i. If the
display device is a computer monitor, the ABC 720p file is fine, but the
PBS 1080i file needs to be converted from interleaved to progressive.
If the display device is a conventional analog TV, the PBS 1080i is fine,
but the ABC 720p needs to be converted from progressive to interlaced.
So, something between the disk and the display has to be able to
interlace and deinterlace.
Not sure about this. IIUC, 1080i was chosen for transmission because
the bandwidth for 1080p simply was not available. 1080i is shown on
768p monitors so it must be converted to non-interlaced before it can be
scan converted. Does the monitor convert it? Is a 1080 CRT interlaced?
I think we need some HDTV research.
--
JRT
_______________________________________________
Open-graphics mailing list
[email protected]
http://lists.duskglow.com/mailman/listinfo/open-graphics
List service provided by Duskglow Consulting, LLC (www.duskglow.com)
