Does Dish downrez HD channels?

Answer is in my sig. As someone else stated, I have an R5000 and can capture the transport streams off one of my 211s.




From Lyngsat:

8PSK modulation
21.5 megasymbols per second
2/3 Viterbi forward error correction
188/204 Reed-Solomon error correction

3 * 21.5 * 2/3 * 188/204 = 39.6Mbits/sec


I think DISH is or migrating to "DVB-S2, 8PSK, MPEG-4/HD, NAGRAVISION 3" and DVB-S2 uses BCH outer code instead of Reed-Solomon error correction giving closer to 43 Mbits/sec at 21.5M SR and over 50Mbp/s with 22.5M SR which the 27Mhz wide transponder could handle just fine.
EchoStar 12 @ 61.5° West - frequencies - freq - channels - packages :: TrackSat.com

I still say that the content provided in MPEG4 (hbo, show, max, others?) is just passed through and the content still sourced in mpeg2, has so much headroom that transcoding would only add some loss in quality. I remember VOOM and the mosquito noise was very visable, I dont notice that any more. Think about it, you'll take a 10% hit in quality transcoding and if you down res anothor 10% that's a 20% loss, can anbody say DTV looks 20% better than DISH.


File size does not really relate to bit-rate and quality, Mpeg's, the "I Frames" can be up to 4 times as large in size as the "P Frames" or "B Frames", file size is directly proproptional to the number of scene changes that generate a new "I Frame" and the number of objects and moving objects, in addition by using VBR (Constant Quality) which is in a 'single pass encorder', can give up o 3 x reduction or more in file size then the CBR equivalent at the same bit-rate with no loss of quality. I can imagine, depending how long they buffer the video stream, that the short term average bit-rates would be bouncing all over place with highs and lows which would kind of explain maybe that statistical multiplexing is used to figure out how many null packets to throw in to keep the the through-put constant, I read somewhere that the transponders through put can have 20%-40% null packets, but they would still need to have the bandwidth available for the data peaks to avoid a data bottle neck.


Why 1440 x 1080, HDCAM tapes have been the typical transfer and distribution media for movies for HD television. HDCAM holds the progressive film image telecined to 8 bit 1440x1080i. It uses 3:1:1 color space and about 3x intraframe DCT compression to a 144 Mb/s data rate vs. 90Mb/s used for SD DigiBeta. Under playback, the 1440x1080i recorded signal is stretched out to 1920x1080i/29.97 or inverse telecined and scaled for 1280x720p/59.94 playback. Later HDCAM models can also play progressive 23.976 PsF tapes out to 1080p/23.976, 1080i/29.97 or 720p/59.94.

Read this one to understand what they are up-linking to the satellite for typical TV network produced material. 1440x1080 is the highest resolution any of this is recorded in the field or studio.
HDCAM, XDCAM & DVCPRO HD: Some Questions Answered : HD High-End

The huge difference is these broadcast formats avoid or reduce intraframe compression so that editing is less lossy. Typical bit rates are far higher.

Production formats:
HDCAM 135 Mb/s 1440x1080i or 1280x720p 3:1:1 coding
DVCPro-HD 100 Mb/s 960x1080i or 1280x720p 4:2:2 coding

News/Reality formats:
XDCAM-HD 35Mb/s 1440x1080i or 1280x720p 4:2:0 coding
HDV 25Mb/s 1440x1080i or 1280x720p 4:2:0 coding

After editing, these are distributed from either HDCAM or DVCProHD master formats to playout servers.

DBS and broadcast bitrate transmission to the home is much reduced to between 8 to 19 Mb/s for MPeg2 HD or about 5-15 Mb/s for MPeg4.

There is absolutely no point to upscaling these to 1920x1080i before downlink. It would only decrease quality. The cable/sat tuner makes the conversion to 1920x1080i or 1280x720p in the box for output over HDMI or analog component.

BluRay allows use of 1920x1080p/24 formats at up to 25 Mb/s for added quality. It would be possible to distribute movies over satellite at this bitrate (or ~12Mb/s for MPeg4) but half the amount of channels could be sent over a satellite transponder. Someday this quality may be offered at a premium charge.

ATSC DTV broadcast does upscale to 1920x1080i or 1280x720p for broadcast in anticipation of future enhancement. They are investing in fixed infrastructure intended for decades of life for transmitters and HDTV internal tuners. 1440x1080 was not approved as one of the supported 18 transmission formats. Still the production behind the transmitter is typically recorded at lower resolution. DirectTV and Dish can do what they want because both the downlink and tuner are under their control.
 
Ok... A's were Directv and B's were Dish... ;)

OK, I admit that I had picked B. When looking at them some parts of the scene looked better on one while others looked better on the other, so I ended up going with what my PS3 showed higher burst rates on.

Don, couple questions:

- Was the capture done on Dish Easter or Western Arc?
- What did you use for the capture device on this?

Thanks
 
I think DISH is or migrating to "DVB-S2, 8PSK, MPEG-4/HD, NAGRAVISION 3" and DVB-S2 uses BCH outer code instead of Reed-Solomon error correction giving closer to 43 Mbits/sec at 21.5M SR and over 50Mbp/s with 22.5M SR which the 27Mhz wide transponder could handle just fine.
EchoStar 12 @ 61.5° West - frequencies - freq - channels - packages :: TrackSat.com

As far as I know, the tuner in the Dish receivers does not support DVB-S2.

TSReader shows the Mux Bitrate as 41.2 megabits on a Dish Turbo 8PSK 21500 SR 2/3 FEC transponder.

They do have a few spotbeams around running at Turbo 8PSK 21500 5/6... those are some where around 51-52 megabits. http://www.satelliteguys.us/thelist/index.php?search=dnall&sub=true&sort=MOD&order=
 
[Wouldn't be easy to ask ppl in SFO area to measure such config ?]

Anyway in reality the parameters provide barely 50 Mbps, usually 49.4 Mbps.
 

Attachments

  • 119w7.JPG
    119w7.JPG
    113.6 KB · Views: 171
I have R5000 mods to both DN 211s and a 4DTV. I also have a number of USB FTA tuners for my C and Ku dish farm. I've been a DN subscriber for about 2.5 years and have noticed a degradation in the quality of my HD recordings since I started. As I have the original transport streams from the R5000 (null-stripped), I have compared the downrezzing and increased compression on the HD movie channels using the same movie and channel that I recorded in the past.

While I have to compliment Dish on their ability to squeeze HD down to very low bit rates, today's images are noticeably softer and challenging material looks horrendous now. I would also note that comparing a DN receiver outputting live through HDMI doesn't look as good as a R5000-captured stream played through a better decoder via HDMI.

I don't know which HD master feeds Dish uses for HBO and Starz, but my 4DTV gets the C-band versions off of 127W (G9). HBO's 4DTV feeds are noticeably better than Starz's and both are dramatically better than what DN uplinks. I no longer subscribe to HBO and Starz on DN, but I did some R5000 comparisons of the same movie at the same time through both paths and that convinced me to drop these movie channels on DN.

I also have observed a few of the movie network DVB/DVB-S2 feeds pop-up in the clear on C-band, which gives me explicit comparisons of the known master feed to DN reprocessed version. I have used a R5000 to record DN at the same time I capture the C-band master feed on one of my USB tuners. I won't mention which networks these are in keeping with the forum rules, but the difference is at least as much as the HBO/Starz 4DTV feeds to DN, usually more.

For the most part the movie master feeds have been MPEG2 below 20 Mb/s null-stripped. I do have some H.264 master feeds at a somewhat lower rate. All are 1920x1080. As of late, the MPEG2 master feeds usually fall between 3-6X higher bit rates than Dish's H.264 versions (both null-stripped). I also have a number of live events that came between 25-40 Mb/s with MPEG2 off FTA. Those are truly incredible.

There is also the question of frame rates. I haven't paid a lot of attention to this until recently, because my HD scalar does a good job of converting 29.97 fps to 23.98 when the movie was so scanned. A week ago I had some master feeds open in Final Cut Pro and discovered they were being sent at 23.98 with flags for 29.97. I haven't had a chance to open DN's H.264 streams to see what is being done there. That could also play into the equation.
 
Lots of info

I have R5000 mods to both DN 211s and a 4DTV. I also have a number of USB FTA tuners for my C and Ku dish farm. I've been a DN subscriber for about 2.5 years and have noticed a degradation in the quality of my HD recordings since I started. As I have the original transport streams from the R5000 (null-stripped), I have compared the downrezzing and increased compression on the HD movie channels using the same movie and channel that I recorded in the past.

While I have to compliment Dish on their ability to squeeze HD down to very low bit rates, today's images are noticeably softer and challenging material looks horrendous now. I would also note that comparing a DN receiver outputting live through HDMI doesn't look as good as a R5000-captured stream played through a better decoder via HDMI.

I don't know which HD master feeds Dish uses for HBO and Starz, but my 4DTV gets the C-band versions off of 127W (G9). HBO's 4DTV feeds are noticeably better than Starz's and both are dramatically better than what DN uplinks. I no longer subscribe to HBO and Starz on DN, but I did some R5000 comparisons of the same movie at the same time through both paths and that convinced me to drop these movie channels on DN.

I also have observed a few of the movie network DVB/DVB-S2 feeds pop-up in the clear on C-band, which gives me explicit comparisons of the known master feed to DN reprocessed version. I have used a R5000 to record DN at the same time I capture the C-band master feed on one of my USB tuners. I won't mention which networks these are in keeping with the forum rules, but the difference is at least as much as the HBO/Starz 4DTV feeds to DN, usually more.

For the most part the movie master feeds have been MPEG2 below 20 Mb/s null-stripped. I do have some H.264 master feeds at a somewhat lower rate. All are 1920x1080. As of late, the MPEG2 master feeds usually fall between 3-6X higher bit rates than Dish's H.264 versions (both null-stripped). I also have a number of live events that came between 25-40 Mb/s with MPEG2 off FTA. Those are truly incredible.

There is also the question of frame rates. I haven't paid a lot of attention to this until recently, because my HD scalar does a good job of converting 29.97 fps to 23.98 when the movie was so scanned. A week ago I had some master feeds open in Final Cut Pro and discovered they were being sent at 23.98 with flags for 29.97. I haven't had a chance to open DN's H.264 streams to see what is being done there. That could also play into the equation.

What you haven't addressed and may be unable to is wht happens to the "raw Feeds" during the conversion process. I do believe that some of the quality lose that is being talked about takes place here. Let's face it the conversion from 2 to 4 is having to be done on the fly and not being done in the studio with much higher quality equipment than the sat co's can use. As far as bit rates for MPEG 2 having to be 3 to 6 times higher than MPEG 4 I would think that would be expected since 2 isn't as efficient as 4.
 
OK, I admit that I had picked B. When looking at them some parts of the scene looked better on one while others looked better on the other, so I ended up going with what my PS3 showed higher burst rates on.

Don, couple questions:

- Was the capture done on Dish Easter or Western Arc?
- What did you use for the capture device on this?

Thanks
129 Western, and Hauppauge pvr... :)
 
129 Western, and Hauppauge pvr... :)

Thanks for the response. So it looks like once the E* and D* STB's uncompress the video the Hauppauge PVR is then recompressing it, which 'might' impact the output? Here's my thinking, the PRV is doing on the fly compression of the component outputs from the STB's, if it has less 'data' to work with (a downrezzed source) it might be able to do a better job with the recompression then if the source has more data in it? Not saying it's so, just a possability?
 
As far as bit rates for MPEG 2 having to be 3 to 6 times higher than MPEG 4 I would think that would be expected since 2 isn't as efficient as 4.

Of course Dish went to H.264 because they can cram more HD channels into a transponder with the same quality loss. But before they went pure H.264 they were chopping MPEG2 master feeds down by around 2-3X. Given the very rough rule of thumb that H.264 can equal MPEG2 at half the bit rate, I concluded H.264 was not brought in to improve the image quality, only to squeeze in more channels. In fact for the most part quality has gone down since the changeover. Even though they are not as sharp, some SD master feeds are now looking pretty good when compared to the corresponding Dish HD abomination.

The bottom line is quality is being sacrificed for Dish's perception that they are more competitive with more, poorer quality channels. Comparing master feeds to Dish downlinks makes this clear. Every time Dish adds more HD channels, I cringe.
 
Thanks for the response. So it looks like once the E* and D* STB's uncompress the video the Hauppauge PVR is then recompressing it, which 'might' impact the output? Here's my thinking, the PRV is doing on the fly compression of the component outputs from the STB's, if it has less 'data' to work with (a downrezzed source) it might be able to do a better job with the recompression then if the source has more data in it? Not saying it's so, just a possability?
Possible I guess... You can set the bitrate to constant, Variable or Variable peak, in the blind test I recorded at identical settings from both Stb's at Variable peak setting... I think I listed the settings earlier in this thread...
 
I agree completely! FIOS for someone who has never had a chance to see it in full force. Is like watching D* best HD channel, and Improving the PQ another step futher. Fios is the Blu Ray of the televison providers.
I wish I could have it.:( Fios is the only reason I love going to my IN-Laws.:D
I am seriously thinking of getting it. I am going to change my broadband from TWC to FiOS and I might change my TV from DirecTV to FiOS.
 
Possible I guess... You can set the bitrate to constant, Variable or Variable peak, in the blind test I recorded at identical settings from both Stb's at Variable peak setting... I think I listed the settings earlier in this thread...

I don't know if it would or wouldn't make a difference. Just wondering how good of a job that a $250 box could do with on the fly compression if the source was 1920x1080 vs. 1440x1080, less data to have to evaulate on the fly meaning it might be able to do a better job with it. What would really be nice if there was a way to get the output from the HDMI in a portable format to compare between the two.
 
I am seriously thinking of getting it. I am going to change my broadband from TWC to FiOS and I might change my TV from DirecTV to FiOS.
The DVRs Suck though. I would take HR over these anyday.
Motorola works fine, It just plain Jane and no recording space. Good for my In laws, but bad for me. It would be a tough switch for me to give up the Vip's for that Fios Motorola.
It would be tough for me to give up an HR for that motorola.:)
Equipment is most important to me , then price, programming, then HD. Fios has all but my Number 1 demand.
 
The DVRs Suck though. I would take HR over these anyday.
Motorola works fine, It just plain Jane and no recording space. Good for my In laws, but bad for me. It would be a tough switch for me to give up the Vip's for that Fios Motorola.
It would be tough for me to give up an HR for that motorola.:)
More HD, save money, bundle packages, and better quality PQ is the plus for FiOS
Better DVRs, the fact that I can take D* with me when I travel, and a better EI is a plus for DirecTV.

This is why I am still on the fence.
 
I don't know if it would or wouldn't make a difference. Just wondering how good of a job that a $250 box could do with on the fly compression if the source was 1920x1080 vs. 1440x1080, less data to have to evaulate on the fly meaning it might be able to do a better job with it. What would really be nice if there was a way to get the output from the HDMI in a portable format to compare between the two.
Seems to me that HP DVR wouldn't qualify for such tests. Definitely you should invest into R5000-HD, what have a deal with _native_ format without altering it.
 
I visited a local DISH dealer who had a 722 connected via HDMI to a Samsung 46" LCD HDTV. I sat relatively close and flipped through HD channels.

I will visit one more retailer with same setup but a 42" LCD LCD HDTV.

First impressions lead me to stick with what I have, but too bad we have to choose between quantity and quality!
 

Users Who Are Viewing This Thread (Total: 0, Members: 0, Guests: 0)

Who Read This Thread (Total Members: 1)