[mythtv-users] HD input in the future (DVI and HDMI)

Sean Cier scier at PostHorizon.com
Mon Jan 17 16:53:59 EST 2005


Brad Templeton wrote:
> 
> Then the push will come from the studios to get rid of component video
> and DVI.  (They are already fully underway with DVI, almost all new
> TV sets have HDMI instead, which is backwards compatible with DVI.)
> Uncertain how that battle will go.

Just wanted to clarify here; I presume you were just simplifying for 
clarity's sake, but wanted to point out that HDMI does not mean DVI + copy 
protection: DVI+HDCP means DVI + copy protection, while HDMI basically means 
DVI+HDCP+Audio+BidirectionalCommunications (to negotiate formats, send IR 
signals, etc) with a new connector.  So nothing about HDMI is really about 
copy protection -- it just happens that the version of DVI that HDMI is 
based on already included copy protection (HDCP) -- and AFAIK, nearly all 
currently-for-sale displays which have DVI input already have DVI+HDCP (as 
do many or possibly most video sources with DVI output, i.e. HD tuners and 
some DVD players).

If anybody doubts this, a bit of searching reveals $30 dongles that take 
HDMI as input and send DVI output, simply by stripping out the 
'unneccessary' bits of HDMI; no decryption involved, since the encrypted 
video stream is the same between the two.

Also note that there's been a lot of suggestion that HDCP is quite readily 
crackable, quite possibly already effectively cracked; there's just not yet 
a good motivation to distribute the cracks ala libdvdcss, since nobody could 
yet do anything useful with the unencrypted DVI stream anyhow.  It's a 
slightly fuzzy area since the industry(*) could fight back against *some* 
kinds of cracks by using key revocation -- but key revocation is an entirely 
untested technology, and is likely utterly, laughably impractical in the 
real world.

(*)I'll leave determining *which* industry as an exercise to the reader.

So even if full-resolution component video and 'vanilla' DVI become a thing 
of the past, using DVI+HDCP or HDMI -- currently the state-of-the-art 
connector, with no suggestions of anything higher-tech on the horizon -- and 
MPEG2 encoders which, as Brad suggested, Moore's law will inevitably bring 
us -- will mean that HD input to our Myth boxes will be a real possibility 
long-term, with likely the only 'hack' being obtaining a 'libhdcp' the same 
way you obtain 'libdvdcss' now (and, worst-case, the use of an external HD 
tuner box used the same way cable tuners are used by many now).  Of course, 
it would be nice if even this turned out not to be necessary for our 
fair-use manipulation of the datastream, but if it does, it's not the end of 
the world.

Of course, for OTA or unencrypted cable, using a GnuRadio card
<http://comsec.com/wiki?UniversalSoftwareRadioPeripheral> and ATSC (or QAM) 
software codec could result in a more elegant solution even in a 
post-broadcast-flag era; but that doesn't necessarily work for encrypted 
cable signals, and its legal status for OTA remains to be determined (though 
it would have to be a pretty flaky legal decision to make that hardware 
illegal).

Even better would be if the EFF and others succeeded in getting the 
ridiculous broadcast flag struck down in court (well, more specifically, the 
FCC's self-granted oversight over the entire data chain and hence the entire 
technology industry).  And/or maybe a cable tuner PCI card that took a 
CableCard and delivered clear MPEG2 bitstreams or, hell, even uncompressed 
frames... hey, a guy can dream, can't he?

-spc

-- 
  /-                 Sean Cier <scier at PostHorizon.com>                  -\
( Everyone should believe in something; I believe I'll have another pint )
  \-                 http://www.PostHorizon.com/scier                   -/


More information about the mythtv-users mailing list