HEVC .265 vs .264 vs uncompressed DVB-S2

[comfortably_numb]

So as I've been analyzing FTA channels on the various KU satellites, I notice some differences in encoding, compression scheme and bandwidth.

The NBC feeds on 103 seem to be uncompressed, varying between 9mbps and 21mbps.

Velocity on 121 is H.264, but only about 3.65 mbps.

PBS HD East on 125 is uncompressed at 13.5 mbps.

Why does the Velocity H.264 stream look so good but the bitrate is so low? Why don't other networks use .264 and .265 if the picture looks as good as an uncompressed DVB-S2 stream?

Is H.265 considered "4K"?

So much to digest. I'm confused.

 

[KE4EST]

Is H.265 considered "4K"?

No. It is just another a more efficient form of compression as compared to H.264. The reason you see 4K pop up when reading about H.265, is because 4K raw files are so much larger than HD.
So most if not all companies are using this for 4K, since they can reduce the file size over AVC(H.264). Mainly for streaming purposes.
If one wanted to though they could use HEVC(H.265) for SD or HD and they do.

 

[Cham]

There was some SD H265 channels on 30w a few months ago. Think they are gone now though.

 

[Jim S.]

My MiniHD RE just scanned in a "radio" channel on 30W this morning named "HEVC test". Some day I'll have to record it and take the disc to the computer to see what's being tested.

BTW I see that there's a new set of the Cubavision family of channels now that it flags as HD although they're not (they're probably PAL instead of NTSC.) More to the point, they have some EPG info. (I forgot to check if it's full info or just titles.)

 

[Brct203]

My MiniHD RE just scanned in a "radio" channel on 30W this morning named "HEVC test". Some day I'll have to record it and take the disc to the computer to see what's being tested.

BTW I see that there's a new set of the Cubavision family of channels now that it flags as HD although they're not (they're probably PAL instead of NTSC.) More to the point, they have some EPG info. (I forgot to check if it's full info or just titles.)

on the Amiko (and some others), MPEG4 channels initially scan in as HD, but once you watch them, if the are SD, the HD flag gets turned off. In the case of Cubavision, some of the channels are in simulcast in MPEG2 and MPEG4, so the MPEG4 ones scan in as HD but are really SD. I think Cuba is NTSC (they must have adopted it before 1959 )

As for HEVC, I read on Portal EDS that Top Latino is adding a channel in h.265 (HD 1080). That could be what you scanned in. I have not tried it yet. It's supposed to be on the same transponder as the regular Top Latino at 11791V2500

BTW, about Cubavision, some of the channels have recently added EPG data.

 

[Martyn]

All of these channels are compressed. That's what H.264 and H.265 are: compression systems, or "codecs" (coder, decoder).

What you're seeing on NBC and PBS is less compression. An uncompressed HD signal inside NBC is probably in the hundreds of megabits per second, if not 1Gbps. But that's unwieldy for broadcast.

As the feeds are destined for TV stations and over-the-air broadcast, less compression is used to keep the signal quality as high as possible.

Velocity is probably fed to Dish Network at a higher bit rate and Dish compresses it down to 3Mbps so it takes up less space and they can fit more channels in.

This is also why an over the air signal from your antenna will usually be superior to the same signal via cable, Dish or DirecTV. Pay TV service providers compress signals more aggressively to get more channels in.

H.264 has been around for a few years and is widely used but not as efficient as the newer H.265, but many set top boxes and TVs don't have the H.265 codec inside so they cannot decode it.

That's why your STB interpreted the HEVC channel as radio - because it has the audio codec but not the H.265 video codec.

Picture quality is determined by how aggressively compression is applied. To complicate matters, many providers use variable bit rate coding if they have a set of channels being broadcast together. So if channel 1 is broadcasting, say, the news where there are many seconds of images of a news reader that don't change much and channel 2 is broadcasting an NFL game where the picture is always changing, the encoder can give a bit less bandwidth to channel 1 and a bit more to channel 2. It's dynamic and changes all the time but more efficient.

With the advent of 4K broadcasting, the streams require more bandwidth and H.265 does a better job of compressing this down so is being introduced with the new format.

That's why some STBs erroneously assume all H.265 is 4K. You sometimes see the same with STBs that assume all 16:9 or H.264 content is high-definition, as noted above, when sometimes it's not. That's just sloppy programming.

 

[comfortably_numb]

That's why your STB interpreted the HEVC channel as radio - because it has the audio codec but not the H.265 video codec.

This is correct. I have both the Amiko RE and the HD.265 and there are some channels the RE can’t decode.

 

[harshness]

This is also why an over the air signal from your antenna will usually be superior to the same signal via cable, Dish or DirecTV.

"Superior" is not a term that I would use with respect to modern OTA. Given that most stations have at least one subchannel, the quality is diminishing. Of course the pay services re-compress the OTA signal so it may take a hit there. Most of them do some clean-up and perhaps some punch-up to hide the effects.

Pay TV service providers compress signals more aggressively to get more channels in.

So are TV stations. My local PBS, KOPB, has two 1080i and one SD feed along with a four-channel audio feed with art cards.

To complicate matters, many providers use variable bit rate coding if they have a set of channels being broadcast together.

This isn't complicated at all. It is very simple and usually gives a good result if you're going to jam a lot of content into a small space. The only thing that is complicated is figuring out the bitrate and that's just a statistic that doesn't mean as much as most people think.

With the advent of 4K broadcasting, the streams require more bandwidth and H.265 does a better job of compressing this down so is being introduced with the new format.

If only they could do h.265 efficiently in real time... h.265 is pretty intensive and the result when it can't spend time considering how to best compress the content isn't nearly as good. I figure that this is why the Winter Games were only available on a delay.

 

[N5XZS]

Little off the track, is HEVC proprietary or not?

The reason I ask can be use for ham radio TV "ATV" for modified HEVC version to send in at 480i up to least 720p 50 or 60Hz.

Someday to send in narrow band digital fast scan TV on VHF Hi 2 and 1.25 Meter band and maybe VHF low band 6 meter band using at least 480i format!

Think of that for sending video via E-skips.

 

[harshness]

Little off the track, is HEVC proprietary or not?

HEVC incorporates IP from many organizations totaling over 3,200 patents. It is heavily encumbered.

AV1 or VP10 may be a better choice.

 

[Martyn]

h.265 efficiently in real time... h.265 is pretty intensive and the result when it can't spend time considering how to best compress the content isn't nearly as good. I figure that this is why the Winter Games were only available on a delay.

I think that was mostly to do with backhaul. After all, Korean viewers were getting live 4K pictures from the games and NHK has managed live 8K broadcasting. Not to say it isn't intensive, but it can be done at both ends of a fiber/satellite link.

I think NBC was pushing the 4K content over its fiber from Pyongchang when OBS wasn't originating any live content, thus the delay.

 

[harshness]

I think that was mostly to do with backhaul. After all, Korean viewers were getting live 4K pictures from the games and NHK has managed live 8K broadcasting.

While many assume that the 4K and 8K content was "broadcast" in South Korea, I'm betting that for most viewers it was more likely delivered via the Internet (netcast?).

ATSC 3.0 support for 8K is at least a year away according to statements made at the recent NAB conference.

Is it really "broadcast" if only a tiny fraction of the population can see it?

 

[N5XZS]

Where did you get the report on ATSC 3.0 may support 8K?

New super HECV codecs or better modulation scheme instead of COFDEM?

Want to see NAB report, since I think 8K is still couple of years away even trying to fit into 6 Mhz channel bandwidth.

 

[Martyn]

While many assume that the 4K and 8K content was "broadcast" in South Korea, I'm betting that for most viewers it was more likely delivered via the Internet (netcast?).

ATSC 3.0 support for 8K is at least a year away according to statements made at the recent NAB conference.

Is it really "broadcast" if only a tiny fraction of the population can see it?

??

Yes, if it was broadcast over the air, it was broadcast over the air. As with any new technology, the number of receivers was probably low at first, but that doesn't mean it only counts when most people are watching it that way.

8K is not currently a South Korean or ATSC3.0 project but something being done by NHK in Japan. NHK already has a test 8K channel broadcasting via satellite and regular programming is due to begin later this year. The Tokyo Olympics will be offered in 8K, at least to viewers in Japan. STRL has demonstrated, and I have seen, terrestrial 8K transmission in Tokyo.

For reference:

KBS program schedule: ?????
NHK schedule: https://www.nhk.or.jp/shv/pdf/shvprogram.pdf#page=1

 

[harshness]

Yes, if it was broadcast over the air, it was broadcast over the air.

But was it something that could be seen live by anyone outside of the television broadcasting community or was it more of a proof-of-concept as all of the broadcasts in the US have been? Experimental stuff is all well and good but if it never makes it to whomever want it's living room, it's mostly theory and promises.

At the time of the Winter Games, the Next-gen standard had only been ratified for a little over a month but since the marketplace is so tiny, I suppose they can move fairly rapidly.

KBS program schedule: ?????

I can't read Korean but from what I can figure out, it appears that this is mostly scenic demo loops. There also appear to be links to view the content online. They appear to be up to episode 51.

We've heard grand claims of several things being broadcast here in the US but we also know that virtually nobody outside of the broadcasters and their technology partners got to see it.

At this point, I'd imagine that the hardware people and the two or three large station groups are footing much of the bill for this experiment but what happens when the scope of the experiment expands?

As readers here, we were afforded some pretty great insights into ESPN's UHD efforts and we know how those turned out. Planning to do something doesn't necessarily make it happen the way we expected.

Then there's the repack and its associated simulcasting...

 

[harshness]

Where did you get the report on ATSC 3.0 may support 8K?

Because the ATSC isn't very conscientious about putting date stamps on their blogs, the information was quite stale and I misread it. Back in 2015, they were saying that 8K wouldn't be on the docket for 2016. They have been jawing about 8K since 2014 or earlier and that's the magic of not having to actually produce deliverables to go along with the hype.

New super HECV codecs or better modulation scheme instead of COFDEM?

That's really their problem to fulfill but I'd imagine that once HEVC hardware develops, they'll get at least most of the way there. Remember when they said that two HD channels couldn't be done with h.264? Of course the quality isn't what it used to be...

 

[comfortably_numb]

Does a signal have to be DVB-S2 in order to be H.264?

 

[harshness]

Does a signal have to be DVB-S2 in order to be H.264?

If you're talking DVB-S versus DVB-S2, yes. DVB-S specifies MPEG-2.

Other modulation schemes may or may not be limited to a specific compression scheme.

 

[comfortably_numb]

If you're talking DVB-S versus DVB-S2, yes

Correct. I've been using TSReader to determine bitrate of video streams, HD flavor, etc. TSReader displays whether a stream is H.264, but not whether a channel is DVB-S or DVB-S2. So if a stream must be DVB-S2 in order to be H.264, then I can deduce that a file, and therefore a channel with H.264 compression is also DVB-S2.

 

[harshness]

So if a stream must be DVB-S2 in order to be H.264, then I can deduce that a file with H.264 compression is also DVB-S2.

It is probably better to say that if the stream is H.264 it can't be DVB-S. H.264 can be modulated all sorts of other ways other than DVB-S2.