[opendtv] Re: Interlace Artifacts
- From: "Stephen W. Long" <longsw@xxxxxxxxxxxxxx>
- To: opendtv@xxxxxxxxxxxxx
- Date: Mon, 10 Jan 2005 22:45:11 -0500
At 09:29 PM 1/10/2005 -0500, you wrote:
>Amen, brother. I wondered a long time ago (on this list) why interlace
wasn't thrown out with the subcarrier. I know we're far from along, but I
just don't get it.
As has been quoted by people smarter than me, follow the money. Find out
who owns the patents for interlace coding/decoding inside MPEG-2 and you
will have part of the story. Find out who owns the patents for the
1035/1080 interlace imaging systems and you will have part of the story.
Find out who owns the patents for segmented frame and related interlace
based tape formats, and you will have part of the story. Add up all of the
pieces, and you can learn why we still have interlace.
What is truly amazing is the linkage between interlace proponents and the
8VSB proponents. Many of the same companies and same people involved.
See, if the answer has to be 1080i, then you have to have a coding system
that cares less about receivability and only cares about having enough
(turns out, not quite enough) bits to transmit 1080i. The only reason 8VSB
was proposed was to protect the patent pool of one company and also provide
bits for 1080i. Remember, much of what is in place now is leftover piece
parts from the failed analog interlace system proposed before a digital
system was mandated. Thank heavens we at least have a digital system (even
with all of its 8VSB flaws).
Here is the saddest part of the story. We could have had a DTV system in
the US that is robust and allows mobile applications. I am not talking
about COFDM - I am talking about 2VSB. 720p60 using MPEG-2 will just fit
inside the 14Mbps payload of a 2VSB transmission system (hell, we run it as
low as 8Mbps on constrained links, gets the job done). So, 1080i supports
8VSB and vice versa. Are they linked - you bet - follow the money, follow
the patent pools.
Almost 4 years ago several of us spoke about drafting a NEW digital
standard for the USA, taking the best of breed from systems around the
world and building a DTV system for the USA that would truly last for 50
years. Robust digital modulation (SFN COFDM or better), H.264 video, SMPTE
KLV metadata and data essence transmission, IP data casting, and Dolby
Digital sound (but would fix the ability to carry separate language tracks
without carrying full 5.1 mixes), a PSIP that actually works, better ADA
compliance, etc. Only problem was a war got in the way and consumed all
available bandwidth for my team, so we did not get around to it.
Someone else will do what I just outlined, building a commercial system
that delivers high value to customers. Maybe in the bands reserved for
selling at auction. It is so interesting to watch VOOM. They almost got
the idea - using improved coding to get more channels on a satellite. They
just backed the wrong horse (MPEG-4 Part 2). Had they waited for Part 10,
they could have been somebody. But that's another story.
At 09:29 PM 1/10/2005 -0500, you wrote:
>Amen, brother. I wondered a long time ago (on this list) why interlace
>wasn't thrown out with the subcarrier. I know we're far from along, but I
>just don't get it.
>
>The "i" formats give license to perpetuate a bad format for who knows how
>long. I wish 1080/60p had been included from the get-go, with all i's
>excluded. If 720/60p were the highest resolution format transmitted today,
>that would seem fine to me.
>
>I always figured plenty of people would WATCH interlaced signals, and
>manufacturers should be free to make whatever display they want to. But I
>like the way you put it - "stop collecting the images using interlace". I
>think it's tragic that we're still originating interlaced material. Phooey
>on that.
>
>
>At 06:13 PM 1/10/2005, you wrote:
>>There is a sub-plot here that is not being discussed. Why were/are people
>>such as myself and many others (Tom for instance) pushing for an all
>>progressive universe? Because we knew "way back when" that the world would
>>eventually move to native P scan devices (such as we are seeing with LCD,
>>DLP, Plasma) and the interlace artifacts would never be properly corrected
>>with affordable conversion devices - it is cheaper and better to just fix
>>the problem at the beginning - stop collecting the images using interlace.
>>It is interesting to see the asymmetric relationship in P versus interlace
>>source origination. Any P source looks ok on an interlace display device
>>(my first HDTV set was native interlace, all I could afford, so even I USE
>>TO have a hated interlace CRT in my home). Interlace on Interlace looks
>>ok. Interlace on P display looks terrible. Coupled with frame judder,
>>some fast motion imagery using interlacing scanning is just awful - have
>>you ever seen a 1080i basketball game over compressed in an ATSC channel -
>>its awful. Contrast that with a 720p60 football game on ABC (and now Fox).
>> On the same bandwidth channel, the pictures are STUNNING. When 720p60 is
>>displayed on a native 720p60 device (my home projector), my mouth drops
>>open every time I see the pictures in my own home.
>>
>>It is amazing that the old guard still sings their tired old songs.
>>Interlace in NOT the future (it never was). P scan forever!
>>
>>At 09:59 AM 1/10/2005 -0500, Craig Birkmaier wrote:
>> >At 12:16 AM -0800 1/10/05, Bill Hogan wrote:
>> >>When Tom McMahon says something you can believe he knows and can tell the
>> >>difference between artifacts from displays and the signals feeding those
>> >>displays. Yes, Tom's observations can be taken at face value.
>> >>
>> >>Regards, Bill Hogan
>> >
>> >I'll second that, and note that NONE of these display technologies
>> >have artifacts that can easily be confused with interlace artifacts.
>> >Contouring, the lack of detail in dark and bright regions, color
>> >fringing (single chip DLP) , and colorimetry issues as DISPLAY
>> >artifacts.
>> >
>> >The biggest problem continues to be that which Tom alluded to:
>> >
>> >It is very difficult to do a good job de-interlacing in the receiver
>> >as opposed to using a high(er) quality professional system prior to
>> >encoding for emission. It get's even harder if the receiver is
>> >forced to work with a noisy analog signal (aka cable) or a trashed
>> >MP@ML encoding that presents the de-interlace chip with excessive
>> >quantization noise (AKA DBS).
>> >
>> >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.
>
>
>
>
>----------------------------------------------------------------------
>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:
