All true, but don't forget, it's a also a recording interface (for which HDMI is not capable). Ron Kilroy Hughes wrote: >1394 output from cable settop boxes never made any sense anyway. >It was limited to compressed bandwidth for SD video, and it's still >limited to compressed bandwidth for HD video. A compressed display >connection is a non-starter in many ways. > >One reason is that most of the "value add" of cable (and satellite) >services, like EPGs, PVR, VOD, VOIP, email, etc. are not provided in the >MPEG transport streams relayed from broadcast networks. They are >uncompressed video/graphics rendered in the settop box. =20 > >You could always add a realtime MPEG-2 MP@HL encoder and 1394 on the >output of each settop box to take advantage of the piddling percentage >of "digital" displays with ATSC tuner/decoders, but then you're limited >to some arbitrary subset of image formats that might decode properly >rather than an MPEG compliant decoder, and it would cost more than the >STB and make ugly pictures worse. But, other than being a bad idea both >technically and from a business perspective, and obsolete before it's >deployed ... it's a great idea. > >The best solution is for settop boxes to output uncompressed, and these >days that means HDMI because of content protection. Displays with HDMI >input will remain useful long after various broadcast modulation and >encoding schemes are dead and gone (and same for audio). > >The current challenge with uncompressed connections is to do a good job >with the bulk of movies and dramatic TV shot at 24 frames per second and >turn that back into 1080P in today's displays, which are progressive in >nature (DLP, plasma, LCD, LCOS, SED, etc.). Raster scan displays are >becoming an historical footnote for consumer HD displays, but live on as >1080i30 signal format delivering an unknown mix of 24P and 30i to the >progressive display system. When displays throw away the repeat fields >and leave the 24P images alone, 1080P24 (with whatever display update >rate) looks great. Unfortunately, the images processors often >"deinterlace" resample, filter, and even scale up and crop (to simulate >the bezel of a cathode ray tube ... how stupid is that?). The result >looks like twitter filtered 1080i, or worse. =20 > >Kilroy Hughes > > > ---------------------------------------------------------------------- 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.