[mythtv-users] component video capture cards

Cory Papenfuss papenfuss at juneau.me.vt.edu
Tue May 3 15:49:55 UTC 2005


On Tue, 3 May 2005 match at ece.utah.edu wrote:

> On 2 May 2005 at 23:14, Reza Naima wrote:
>
>> Is this still true?  There are _no_ video capture devices for the above
>> formats?
>
> As far as I'm aware, this is still true. You can get them in the $2500 range, but no
> "consumer" cards.
>
> I have 3 students working on one as a senior project. I dunno if they'll be
> successful. They'll research it all summer then begin the design & build it in the fall.
> It's not trivial. If anyone has any suggestions for the design, pass them along.
>
> Marvin Match
> U of U
> Electrical and Computer Engineering Dept.
>
 	Aside from the bandwidth implications, it shouldn't be that 
difficult, I wouldn't think.  The fastest bandwidth of standard HDTV is 
1080i, which is 1920x1080x30 = 62.2Mpixels/s.  At 8bits/Y, 4bits/Pb, 
4bits/Pr, that's 2 bytes/pixel, or 124MBps... roughly 1 gigabit/s for the 
RAW video.  Pretty ugly to deal with that way.

 	Are they planning to do some rudimentary (or non-rudimentary like 
MPEG[24]) compression on it in an FPGA or something?  I would think that 
NUV or RTJPEG could cut it down by 10:1 pretty easily if it could be coded 
up tight enough to deal with.

 	Other issues that will be a problem are jitter on the sync 
signals.  You'll probably need to get ahold of the standards for the 
signal... I believe EIA 770.2 for SDTV and EDTV, and EIA 770.3 for 
tri-level sync of HDTV.

-Cory


*************************************************************************
* Cory Papenfuss                                                        *
* Electrical Engineering candidate Ph.D. graduate student               *
* Virginia Polytechnic Institute and State University                   *
*************************************************************************



More information about the mythtv-users mailing list