SD resolutions dropping on 110W

I don't want go in lengthly discussions - for me all the arguments very subjective and we grinds it already many times;
the bar of acceptance of good quality is set already: for SD - DVD, for HD - HD DVD or Blu-Ray disks.
I'm not referring to original source - films, studio recordings what usually not available for mere mortal.
 
DVD video is almost 9Mb/Sec...You can't push that over satellite with current technology, at least not reliably. It's a stretch with FiOS, and you quickly run out of capacity. Dimensionally, 720x480p is doable, but not at that bitrate.

I miss EDTV; 850-ish by 480 panels, down-scaling HD content to fit, native-ish to DVD, and all you have to do is de-interlace SD content to make it right.
 
I have a VIP622 and, for the last week or two, have not been able to tune SD from satellite 110 on one of my two tuners. Fox Movie Channel, for example, is almost completely pixelated. For some reason I can tune it in okay on the alternate tuner. Makes it kind of frustrating to record things.
 
DVD video is almost 9Mb/Sec...You can't push that over satellite with current technology, at least not reliably. It's a stretch with FiOS, and you quickly run out of capacity. Dimensionally, 720x480p is doable, but not at that bitrate.

I miss EDTV; 850-ish by 480 panels, down-scaling HD content to fit, native-ish to DVD, and all you have to do is de-interlace SD content to make it right.

Tell that to Dish as they advertising their service as "DVD quality" and as"Blu-ray" for HD.
 
It all seems to defy some of the most basic principles of proper video encoding.

The phrase "proper video encoding" on the Internet is used 99.9% of the time to refer to further compression of anime and pirate DVD, TV, Blu-Ray and HD-TV material, often in a two-pass method.

Satellite and Cable use real time encoding . This is done quite differently from what is called "encoding" on the Internet.

Dish Network was one of the first to to use different encoding that is targeted for the content of the channel. (For example, when they first added Fox Soccer Channel, the ball would move in a jagged way, and a few months later, that effect was corrected.)

If you have any professional real time encoding experience, then we would be interested in hearing about it.

Another major fallacy of PQ threads is the idea that any particular event or channel that you are viewing, automatically represents an overall policy or methodology or PQ level. Given what I have experienced in corporations, it is highly possible that the engineers know that Encore Drama needs some work, but that instead they have been ordered to work on those new HD Locals that are added in large quantities every week since Ciel-2 came online.
 
The phrase "proper video encoding" on the Internet is used 99.9% of the time to refer to further compression of anime and pirate DVD, TV, Blu-Ray and HD-TV material, often in a two-pass method.

Satellite and Cable use real time encoding . This is done quite differently from what is called "encoding" on the Internet.

When I referred to "proper video encoding," it was quite clear that I was referring to the use of lower bitrates for less filtered (less compressible) video and the use of higher bitrates for more filtered video (more compressible, if filtered intelligently). Dish seems to have the bitrate allocations backwards for the two general types of filtering they apply to SD channels on WA. I can't see how you made the leap from that topic to online video piracy.

Based on your comments, it seems that your definition of real-time encoding excludes the possibility of any multi-pass encoding. You almost seem to believe that multi-pass encoding in general can only be used for illegal activities, which is just silly since it is probably used to prepare everything from movie trailers online to streaming Netflix content to video downloads via iTunes. For a real time (or near real time) system, the video duration over which an analysis pass could be done would have to be limited to at most a few seconds in order to push all of the video through on a short delay close to real time, but multi-pass encoding could definitely be done. Their current system appears to only offer a channel in a mux more bits after increased activity has been detected rather than when the activity begins, and short-duration multi-pass encoding would help to reduce artifacts in those cases. In general, anything that can help further to dynamically prevent compression artifacts and not overspend bits on low-detail/low-activity scenes reduces the need to apply aggressive filtering to the input/source material in order to meet an average bit rate target (over an extended period of time). This can result in overall improved PQ, which would be very nice.

It seems that you chose to overlook the matter of how ugly the video becomes when Dish overuses edge enhancement. The resulting video has noise added, edges distorted, and details blurred together or shifted. When video has just been softened, as Dish does for many channels on 110W, it is reasonable for end users to adjust the sharpening on their TV to partially compensate for this effect in they want to. However, when excessive edge enhancement is applied, as it is for many channels on 119W, the end user can't reasonably fix the broken picture. The filtering has messed up the picture to a point that no TV adjustment, no improvement in encoding, and no increase in resolution or bitrate can improve the PQ.

I don't have and don't need the specific "professional" experience you seem to think I need in order for me to make educated comments and express opinions on the subject being discussed here. Nothing I've seen or heard so far has demonstrated that Dish is actually giving SD PQ much consideration, and evidence has been given in this and other threads to demonstrate that they definitely aren't giving SD PQ much consideration. I understand that Dish's resources are finite and that only so much can be done given bandwidth constraints and limitations of MPEG-2 video compression, but there are definitely some things that could be done to improve on the current situation that don't involve huge expenses such as launching more satellites, swapping out all receivers, or even replacing encoders. If they would drop the edge enhancement on the channels they're applying it to then that would shut me up quite a bit, but that doesn't seem likely to happen any time soon.
 
Tell that to Dish as they advertising their service as "DVD quality" and as"Blu-ray" for HD.
Yes, Dish made two mistakes. First, it's DVD/BR resolution, not quality. Second, it's not even that resolution anymore, as mentioned in the first post of this thread.

I will say this; my Dish SD is almost as good as OTA SD, Dish HD upscale (Food, TNT, Hist) picture quality is on-par with my $99 upconverting DVD player, and my Dish HD (Smith, HDNet, LiLs) picture quality is better. OTA HD and Bluray still beat Dish, but for most people, for most viewing, it's adequate. I think that's the bar that Dish is trying to meet -- Adequate.
 
I was surprised that last night's Daily Show on the new HD version of Comedy Central looked almost the same (i.e. only very slightly better) than on the SD channel. People forget a lot of shows just have poor production quality to begin with, and all the bandwidth in the world won't help the picture quality in these cases.
 
DVD video is almost 9Mb/Sec...You can't push that over satellite with current technology, at least not reliably. It's a stretch with FiOS, and you quickly run out of capacity.

That's utter nonsense. Back when HDNet was one of a handful of HD stations on Dish, I would regularly see bitrates in excess of 15Mbps. There's nothing to stop them from committing an entire transponder to a single HD channel and sending down something in excess of 30Mbps, except the receiver's ability to decode such a high bitrate signal.
 
Well, I count 54 full-time, non-premium, high-def channels on the
Dish website
. They state "Over 65" in TurboHD Gold, but I count 54 including PlatinumHD, but that's just counting bullets and pictures. Again, that doesn't count HBO, Cinemax, Showtime, Starz, RSNs, or satellite-based PPV. Let's say there are 250-300 SD channels that they also MUST keep online (or aren't their responsibility). Let's not even mention locals.

Even if you clean-slate the channels-per-orbital, which is what Dish should be doing with Eastern Arc, the bandwidth just isn't there. I miss the days of Voom as an operator; they could have had a chance if they could have kept their own birds.
 
Last edited:
Actually, Voom had their own bird before they became just a provider for Dish HD content (Rainbow 1, bought later by Echostar).

Regarding DVD-video at 9MB/sec - that's using MPEG (or is it MPEG-2) compression. That should be considerably less if compressed in MPEG-4, in theory at least the resolution and bitrate could be maintained to present that quality to the end-user.

More fodder for Smith: I think Dish SD is DVD-quality with one caveat. The DVD I'm talking about is something I converted from a VHS tape to the yellow, red and white inputs on my standalone DVD recorder. :D
 
Buckchow said "I don't have and don't need the specific "professional" experience you seem to think I need in order for me to make educated comments and express opinions on the subject being discussed here."

My response: that's true, but if you actually had professional encoding experience you might be able to explain why Dish Network is doing what you apparently think is exactly backwards with their processing. There may be engineers who frequent this forum who know exactly why and how Dish is functioning, but who are prohibited by commercial and job reasons from explaining. I doubt that Dish is degrading quality for any reason other than maximizing bandwidth use, although there is always the possibility that there are better patented encoding processes not used for cost reasons. Notwithstanding that, I'm glad you started this thread since I'm always interested in technology advances that improve picture quality.

Off Topic: CowboyDren said "I miss the days of Voom as an operator; they could have had a chance if they could have bought their own birds."

As you probably know, the owners of Voom network spent over $600 million on the network before they gave up; they just couldn't get enough subscribers. As I understand it, they sold Rainbow 1 to Dish as part of the process of moving to Dish Network. It's too bad they didn't just go ahead and invest $100 million per year in new programming when they should have.
 
How many transponders do you think Dish has?

Dish has a ton of transponders as mentioned above, but they do not do any good. Only the ones on 61.5/72.7/77 and 110/119/129 matter.

Dish has 29 on 110, 21 on 119, and 32 on 129. Of which 19 on 110, 15 on 119 and 16 on 129 are available for national channels. The rest are on spot beam (and one leased for distants).

On EA they 30 on 61.5, 2 on temporary auth on 61.5 but 4 TPs are broke on E3 so effectively 28. 72.7 they have 16, 77 could have up to 32 but unknown until we see the configuration on the new satellite what will be USA and what will be Mexico. 8 of the 61.5 are being used for spots right now, they can turn more to spot later when they either finish moving international off or get another satellite over to cover the broken 4 TPs.

105, 121, 118.7, 148 they have all the TPs but are not used for regular programming.
 
I doubt that Dish is degrading quality for any reason other than maximizing bandwidth use, although there is always the possibility that there are better patented encoding processes not used for cost reasons. Notwithstanding that, I'm glad you started this thread since I'm always interested in technology advances that improve picture quality.

The technology for much higher quality SD MPEG-2 video at the current bitrates definitely exists, so that's not the issue. I like your thinking that patents may somehow be involved with limiting the technology being used. If they decided to try to come up with their own filtering and/or encoding methods in order to avoid patented technology, or they decided to only use technology based on expired patents, that could explain a lot.

I'm trying to get a couple of video samples together in another attempt to squash the "they don't have enough bandwidth for better quality" claims. May take a few more days, but the idea of the comparison will be to show what a Dish channel with a higher bitrate than a Bell TV channel looks like, then to show what the same sample of a Bell TV channel looks like when transcoded down to ~1.5Mbps (around where Dish tends to bottom out) without filtering or lowering the resolution. The tricky parts are finding something that is currently airing on both Dish and Bell and arranging for someone to get a sample from Bell. Should be fun. :)

Getting back to the original topic of this thread, the resolution has dropped on more channels on 110W including Encore Westerns. Encore Westerns is of specific interest because it is one of the relatively few channels on 110W that has edge enhancement (EE) applied to it (most channels with EE are on 119W), so the quality is quite noticeably worse than other channels having their resolution further reduced. The fact that some of the source material for this channel is of questionable quality will only be exaggerated by these further reductions in picture detail.

A new observation is that the resolution reductions to 480x480 appear to be accompanied by an increase in the duration of the picture sequences. The number of pictures per sequence tends to hover around 30 on channels still at 544x480, but it's closer to 50 on the 480x480 channels. This means, among other things, that more are bits available to B and P pictures (good), more artifacts are present when scene changes aren't detected (bad), and there is a slightly longer average delay before video appears when tuning the channel in (bad for channel surfers, probably of no relevance to everybody else.
 
buckchow, as I recall DVD books require 1 to 10 for I-frame - correct ?

Sorry, I've never really looked at the DVD requirements. If it is 1-10 as you said, that wouldn't be unreasonable for a DVD. In general, more bits are available for more I-frames. Also, minimizing seek delays is probably a higher priority for DVDs than with digital cable/satellite.

Last time I checked multiple channels, which was almost a year ago, DIRECTV was using about 30 pictures per GOP and Bell was using exactly 30 for every GOP on some channels and hovering around 45-50 on some others. As of yesterday, one of Bell's channels was alternating between 24 and 30 (ensuring one new I-frame per second, or one picture per frame as the channel alternated between "film" and NTSC content).

I don't think there's any limit in the MPEG-2 spec for the number of allowed pictures per sequence since the temporal_reference value used to specify each picture's index in the sequence can go up to 1,023 and wraps around (mod 1024) to 0. Of course if a sequence becomes too long, the pictures contain complex motion (or scene changes), and the number of bits allocated to P and B frames is too low, then the PQ will fall apart quite unpleasantly. Definitely need good detection of scene changes to prevent that from happening.
 

Which barrel connector for HD?

can I do TV1 and TV2 with 2 different UHF remotes?

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

Who Read This Thread (Total Members: 2)