大象传媒

I'm getting really confused about 大象传媒 blogs now. TV things like 大象传媒 HD and the Red Button are posted on the internet blog. Computer things, such as the new Doctor Who computer game, get posted on the TV blog.
Does this make sense to anyone?

Andy, I see your point, it probably is more appropriate to have posts about technology here than the TV blog. Perhaps its more the naming that is confusing, maybe calling it the 大象传媒 Tech blog would be clearer (though then it might get confused with dot.life).
Personally, I'd love to see a more Web-focused blog. I think its a pity that so many of the smaller blogs are being merged to form larger blogs with less of an identity.

Look forward to the rest of your posts.

Andy said
TV Panels can't fully resolve the 1920 signal so he said it would be pointless to increase 大象传媒 transmissions to this resolution.

This is just not true. I have a Sony KDL-46x2000 which has a display resolution of 1920x1080 and I run it in pixel mode. This means that each pixel transmitted is mapped to one pixel on the screen if it is transmitted as 1920. Close up I can see each individule pixel. I must admit I have not counted them I just take Sony's word.

With a viewing distance of 3 times the screen height a person with 20/20 vision can resolve 1920 horizontal pixels. I find it quite easy to see the difference between a 1920 and a 1440 picture. Most other broadcasters use 1920.

The "disapointment" issue was a bit of an own goal wasn't it. Of cource the 大象传媒 have always been aware that the picture quality has deteriated. So we come to the "using the spectrum efficiently argument". The flaw in this argument is that spectral efficiency is not the primary requirement of HD. The essence of HD is high quality pictures and sound and this should be the primary requirement. Spectral efficiency is secondary. If you want spectral efficiency stick to SD or even go back to Baird's 32 line standard.

The ITU has a definition for for HD. It says that at 3 times the screen height the picture should look the same as real life. Well that is the goal. As we know the even the best pannels cannot reproduce the full spectrum of colours visible to the human eye but this definition should be our goal.

What the 大象传媒 realy mean by spectral efficiency is low cost. Unlike colour the 大象传媒 has not introduced an HD licence fee to fiance the increase cost of HD. So the 大象传媒 is trying to deliver an HD service within the existing budget. Something has to go and that something is called quality.

Welcome Back Andy

Very interesting,look forward to the other posts.

Is todays software update/fix the first for the new (10 months old)encoders?

It seems a long time to solve a problem that you were aware of from day 1.

Please keep us updated on any further updates and fixes.

Just a short comment for now until we red all three parts.

Firstly thanks to all concerned for their efforts and input - I hope all posters will address the issues raised in the reports as they are made available.

So my one comment is also about the 1920 issue.

My 42 Toshiba Z3030 is a full 1920x1080 panel and it also offers 1 to Pixel Mapping.

Cheers, daveac

Hi all, thanks for the comments so far, and I look forward to reading the rest following our other 2 Blogs. Sorry for the long wait too, there was a lot to discuss after our visit, all done by email in our respective spare time.

I'm away now for a couple of weeks so you'll probably not see much comment from me during that time, if any. But I'll certainly read all of yours on return and I'm sure that, in the meanwhile, my fellow visitors will be monitoring and responding to your comments in my absence.

Thanks too, Andy for being there to respond to comments, and for taking the time during a very busy week to do so.

Hi Andy, can we have this report/s posted on the main bbc blog as previously requested as well.

A grating frequency is a concept associated with the sampling of an analogue signal. In digital television the picture is built from pixels and no analogue sampling is required througout the chain. However in a CRT the digital signal is converted to an analogue signal to modulate the electron intensity. So a grating frequency is relavent to a CRT display but not to a digital display.

SDI and HDMI send the video as digital stream. Conversion should not affect the resolution. Please can you tell us how this was demonstrated.

I point out that the colour difference streams are transmitted with only a quarter the number of pixels. Eg 960 x 540.

I would add that a digital filter could mimic some of the effects of sampling. Subsampling at 1440 pixels is realy a form of digital filter. So to is the video codec which can lead to a loss in resolution which is so clearly demonstrated by the 大象传媒. If the 大象传媒 were to increase to 1920 pixels they would also have to increase the bitrate for the same level of artifacts. The bottom line is that 1920 pixels could be clearly resolved if the bitrate was adequate.

As I said before I can clearly see the difference between a 1920 and a 1440 picture.

Just before I go, I'd like to make a clarification regarding Andy's introductory comments where he says: "This is Paul's blog, the content is his and the rest of the group's" In the final rounds of editing we'd missed the point that, when read in conjuction with the way I introduced the Blog, it could be deduced that I wrote most of it. That would be an incorrect deduction.

The Blog was written by all of the visitors (pictured above + the photographer), with every one of us making contributions, some of them quite considerable, e.g. the entire account of the day's activites was written by citizenloz. The help and assistance of others not able to make the day should be acknowledged too and they may wish to step forward here and take due credit.

I am grateful to them all, as I hope is made clear in the rest of the forthcoming Blogs.

@12 Trevor, I agree. STV HD has been testing, with 1920x1080, and (when not upscaled) is definitely a clearer picture than ITV HD, which is at 1440, with the same source material. It is also running at slightly higher bandwidth. I guess some of it comes down to how well the TV comes with 'stretching' from 1440 to 1920. mine doesn't have any of the newer extra processing.
@9, Paul, have a nice break!

Re #13.

Thanks Paul for clarifying that.

The main issue in all this has been ignored once which is the quality of the encoder that is being used. Just like a car, the performance of encoders varies dramatically amongst different manufacturers.

It is interesting that both the British and the Germans are complaining about mediocre picture quality on public-broadcasters' HD services. Complainants on both sides of the North Sea are pointing to incorrect solutions. Us Brits are saying that the bit-rate should be increased and the Germans want 1080i. Both are missing the actual reason which has let do picture quality problems around the world - the encoders are poor. Virtually all of them are optimising for metrics like PSNR instead of actually bothering to produce a picture that is pleasing to the human eye. It is possible for PSNR to increase and visual quality to decrease.

For example I have seen a case in which only 4 out of a possible 16 intra-prediction modes were used - consider this H.264 intra-prediction done very lazily.

It is also worth pointing out that the quality of the distribution encoders are poor too. There was severe blocking during some of the tougher scenes during this year's and last year's Eurovision. Some of the playout server encoders I have seen also are poor too.

Having said that, throughout Europe there are many broadcast engineers who share my concerns. It is the fault of some people in the industry for merely accepting such issues and drinking too much "manufacturer kool-aid". (3-5mbit 1080i comes to mind as a good example of this)

Frankly, it is becoming silly to read all the various "explanations" for picture quality deficiencies that are posted on this blog and in the comments every week. Some of them are valid but in relative terms are mere drops in the ocean.

It doesn't matter much anyway because in a few years we will be at the level of American HDTV, which shouldn't even be called HDTV. A few channels on the continent are already not far from that level.

"鈥latform neutrality has to be considered when reviewing transmission parameters"

I'm really concerned and disappointed by this statement by the 大象传媒 as this amounts to nothing less than an admission that Freesat is being dumbed down in quality of transmission so as to match Freeview HD which is much more bandwidth restricted.

This to me is totally wrong.

Its not biased to have one platform of better quality than another when that is caused by a restriction on the available tranmsission space on the other that simply cannot be avoided. This situation is entirely different to that where both platforms are equal but one is given favourable treatment.

It is however, morally wrong to deliberately dumb down the other broadcast channel and spoil people's viewing pleasure simply to match the restricted one.

The 大象传媒 owes a duty to deliver the best possible quality to its viewing audience and licence payers and is failing to deliver that where it reduces quality deliberately for these reasons.

To put it simply, why should the viewing audience of one medium suffer a deliberate low quality picture simply because it cannot be avoided on another channel?

Also, I briefly make the point I've been making for a long time on all of this.

Freesat and Freeview need to be seperated out and allowed to play to their own strengths.

There's room for both platforms to co-exist simply in different roles. Freeview as the easy to get cheap option and Freesat as the ultra high quality, max content enthusiasts / frequent viewer option for those who put picture quality over and above cheap reception. That way everyone is happy and the public are given a clear choice over which medium to choose according to their viewing habits, pockets and installation preferences.

I think I wrote "distribution encoder" in the comment (can't see it yet) above when I actually meant "contribution encoder".

Hi Andy,

Can you expand a bit on your comments about not requiring noise reduction and "as we have found a better way to handle noisy pictures". Sounds intriguing.

NR is not something I remember us talking about in much detail during the visit so it would be interesting to find out what you are doing to handle it with this latest change.

Thanks

"鈥latform neutrality has to be considered when reviewing transmission parameters"

Why?

re #10, Mo, 鈥淭V Panels can't fully resolve the 1920 signal so he said it would be pointless to increase 大象传媒 transmissions to this resolution.鈥� I'd just like to point out that the first half of this sentence is Andy's suggested formulation, though the second is ours. I only say this to clarify whence any response might come.

The incorrect capital P is all mine. :/

Part 2 of our blog is now available:

/blogs/bbcinternet/2010/06/picture_quality_on_bbc_hd_a_vi_1.html

鈥ost of bandwidth is not a significant factor in reducing bit-rates.

If it is no issue, then why has bitrate been stuck at a mere 9.7Mbps for 10 months, when there is so much free space on that transponder?

鈥latform neutrality has to be considered when reviewing transmission parameters.

WHY? Why then does the 大象传媒 HD satellite feed to Europe transmit at much higher bitrate & resolution?

鈥ence there is an obligation to always use the available spectrum efficiently.

Then why have we had half a transponder transmitting nullbytes for the last 10 months?

鈥ncreasing the bit-rates for 大象传媒 HD now, using the spare satellite capacity, would only lead to disappointment when they are reduced again to make way for further channels.

ROTFL, what a feeble rationale, & an own-goal too

Andy claimed that:
鈥V Panels can't fully resolve the 1920 signal so he said it would be pointless to increase 大象传媒 transmissions to this resolution.

Oh really, in that case why do manufacturers persist with their 1920*1080 FULL HD panels, with their 'full pixel' setting? Surely they should be making 1440*1080 panels to reduce manufacture costs?
And why are all those other HD channels transmitting 1920 - could it be that they all are wrong to do so?

Andy Quested showed a series of clips from other HD channels which he believed highlighted similar problems to those the group has raised with regard to 大象传媒 HD. Examples included lack of definition, poor colour, or poor source material. These were viewed on a large 65" plasma display.

Interesting to see that shows from competitor channels that were selected for poor PQ purposes were demonstrated on such an enormous panel. Was the viewing at the correct distance for this size of panel? I take it that the visitors were also shown 大象传媒 HD's Ashes 2 Ashes & Survivors on this same screen?

Andy Q
During the visit we quite clearly demonstrated that the uncompressed 30MHz grating fed via an HD SDI to HDMI convertor was not fully resolved.

Sorry, what is this 30MHz grating with its SDI to HDMI converter? In reality we have a viewer (not a grating!) watching via a Full HD panel 1920, connected to STB via HDMI lead, watching full-pixel mode at the recommended viewing distance - where does the SDI to HDMI conversion come in?

Andy Q, thank you for the detailed reply - can't say as I agree with it all though ;-)

Andy #25

When you say

"On a waveform monitor the gratings on the test card are all the same level."

is thatbefore or after encoding?

Surely the issue is not whether a Full HD consumer panel can display the full 1920x1080 broadcast or not, but whether it can display better resolution than when fed with a 1440x1080 broadcast.

Whilst Andy did demonstrate to us that there were indeed inadequacies in the handling of the 1920 broadcast - in comparison with a very expensive HD CRT - we didn't see a comparison of a 1920 vs 1440 broadcast.

So like so many things, it seems Andy is able to prove his point (that 1920 displays cannot display the full resolution) but not actually addressing the real point that viewers are making - namely that 1920 is better than 1440.

Of course a 1920 panel can display 1920 pixel resolution, that is its native res - the res at which it performs optimally. 1 broadcast 1920 pixel = 1 display 1920 pixel.

Contrast with 1440 broadcast res. - how many panels are manufactured at this res? None that I know of, so ALL panels have to interpolate the 大象传媒 HD signal.

25 Andy Quested wrote:
At some point in the future we will go to 1920 horizontal if we believe it is necessary.

But why don't you believe it is necessary now, give the almost universal use of 1920 by other UK HD channels, 大象传媒 HD abroad, the notes about the need for 1920 that have been made by the 大象传媒 themselves to the trust, preferences by the 大象传媒 for source material to be in 1920, etc, etc, etc, - all of which have been pointed out in the blogs many times typically without reply.

What people find so infuriating is that the 大象传媒 just can't simply accept this, do it, and move on, when even one man and a dog HD channels can.

Or at least offer a valid reason - instead of obfuscation such as "panels can't display it" - or at least a time frame rather than a vague "when necessary".

Andy no disrespect meant by this as I know your a busy man, but I couldn't help but notice that you skipped around my point in post 17 above.

#29 Andy Quested

"Coming to the ITU test, this was very interesting to set up and to watch happen. There were concerns on both sides as to the implications of such an open test but I hope you all appreciate how impartial the test is and that the results are not only reliable but repeatable."

Reliable and repeatable maybe, but not conclusive as our third blog tomorrow will show.

Fascinating reading this folks. A big thankyou to everyone involved. Only one comment to make so far, HD_fan428 - love the suit and sunglasses! :D

#36, hdblogger, glad you like it so far. But it's jtemplar in the cool outfit--I stayed behind the camera! :)

HD_fan428,

Sorry my mistake. Still, looks like an audition for a Crimewatch identity parade.

Seriously though, I'm still of the opinion that it's down to the source material. I've seen some pretty outstanding stuff on 大象传媒 HD as with other channels and I've also seen some really poor stuff on other channels. I'll wait until the next installment though before I'm convinced either way. It's a shame Paul is away for two weeks. That guy has more holidays than Judith Chalmers.

Jonatham Ross, last Night
Andy, lets get to the point of all this, why then does 大象传媒 HD look
so bad then, i still can see sub standand HD Broadcasting all the
time on 大象传媒 HD, and the ture fact is that ITV HD looks so much better
most of the time, But there broadcast is at the higher Resolution,
which I CAN SEE LOOKS BETTER and sharper and cleaner looking,
The very sad fact to me is that ITV HD Picture Quality is still
very badly let down by a Bit Rate that is far to low, Yes and the
same old fact , a lack of bandwith on the Freeview HD Platform.

Andy
I still remember the Winter Olympics, so much better on Eurosport HD,
has you no it was Broadcast at the best quality, ie 1920 x 1080i with
a higher bit rate of 16 to 18 mbps.
it was so easy to see the difference, my point is to you Andy it was
in a class of its own, not second class HD i keep seeing on 大象传媒 HD...

@Andy, #35 that wasn't actually my point.

I was seeking comment on the point that Freesat is being deliberately dumbed down so as to match Freeview quality and the fact that, that isn't being "platform neutral" at all.

I've copied the post down here to avoid you having to search the blog:

"鈥latform neutrality has to be considered when reviewing transmission parameters"

I'm really concerned and disappointed by this statement by the 大象传媒 as this amounts to nothing less than an admission that Freesat is being dumbed down in quality of transmission so as to match Freeview HD which is much more bandwidth restricted.

This to me is totally wrong.

Its not biased to have one platform of better quality than another when that is caused by a restriction on the available tranmsission space on the other that simply cannot be avoided. This situation is entirely different to that where both platforms are equal but one is given favourable treatment.

It is however, morally wrong to deliberately dumb down the other broadcast channel and spoil people's viewing pleasure simply to match the restricted one.

The 大象传媒 owes a duty to deliver the best possible quality to its viewing audience and licence payers and is failing to deliver that where it reduces quality deliberately for these reasons.

To put it simply, why should the viewing audience of one medium suffer a deliberate low quality picture simply because it cannot be avoided on another channel?

Also, I briefly make the point I've been making for a long time on all of this.

Freesat and Freeview need to be seperated out and allowed to play to their own strengths.

There's room for both platforms to co-exist simply in different roles. Freeview as the easy to get cheap option and Freesat as the ultra high quality, max content enthusiasts / frequent viewer option for those who put picture quality over and above cheap reception. That way everyone is happy and the public are given a clear choice over which medium to choose according to their viewing habits, pockets and installation preferences.

In #35 Andy Q states 'I am pleased to say on Friday Sat HD actually reached the quality of Freeview HD!'

Was this said tongue-in-cheek, or are you really claiming this is worthy of celebration? That a platform of many years standing with much greater potential bitrate - DSat - has actually just managed to achieve equality with an inferior transmission platform - DTT - that's merely a few months old, bitrate-constrained & hardly out of the testing phase?

#43 Andy Quested

I was hoping after our visit that you would engage in all aspects of our visit report and not be selective in your responses, but that appears not to be the case.
The fact that you have ignored Alsone's post (#41) completely and only picked up on a secondary point in nur0's post (#42) just confirms that platform neutrality is a policy being pursued by the 大象传媒 for HD transmissions.

The third section of our visit report, containing conclusions and recommendations, has now been posted here:

/blogs/bbcinternet/2010/06/picture_quality_on_bbc_hd_a_vi_2.html

#46 Andy Quested,

But that contradicts the statement from Stephen Baily, acting Head of Distribution technology. Perhaps Alsone's choice of the term 'dumbing down' was wrong, but clearly a policy of not using any more resources on satellite than is available for DTT would leave capacity for a second HD channel on satellite, albeit at the same unacceptable picture quality as the current 大象传媒 HD channel

My point Andy is simple, the 大象传媒 ISN'T pursuing a policy of platform neutrality, what they're doing is a CLEAR BIAS TOWARDS FREEVIEW.

A policy of platform neutrality is quite simply to look at the best bit rate available for optimum transmission quality eg 16 mbs.

Then apply that that equally to all of the platforms eg. Freesat, 大象传媒 on Sky, 大象传媒 on cable, Freeview.

Where a platform cannot through bandwidth restrictions accomodate the optimimum bit rate chosen for the 大象传媒 HD medium, then bit rate on that platform should be reduced to the maximum level that can be accomodated along with a pledge to increase the bit rate as and when extra bandwidth ever becomes available.

That is platform neutrality. Where you have one standard and apply it across the board to all platforms without any bias, with only platforms that cannot accomodate it having a reduced bit rate.

What the 大象传媒 have done however, is entirely different and shows a CLEAR BIAS TOWARDS FREEVIEW.

They have taken a standard and set it, then looked at Freeview and said Freeview can't accodate this bit rate so all other platforms will have to suffer as a result as we can't have other platforms exibiting a better picture than Freeview, and then thrown the chosen standard out of the window and dumbed all of the other platforms down to Freeview's restricted bandwdith level irrespective of the orginal standard's optimum encoding or the poor picture quality for all that results.

That isn't neutrality. That is a clear bias towards Freeview and ensuring that other mediums don't exceed Freeview quality.

What I'm suggesting Andy is that the bit rate shouldn't be reduced just to accodate Freeview's bandwidth failings but rather if you are pursuing platform neutrality, what should be being done is a bit rate set which maintains high quality HD and then apply this to all available platforms. If that bit rate falls above Freeview's available space then Freeview should be transmitted at the reduced bit rate alone (but at the maximum level available for that platform). That is platform neutrality, not reducing every platform just so as to avoid others looking better than Freeview.

Astra commented a while ago that below 12mbs they get complaints about HD quality. Therefore this begs the question as to why the 大象传媒 haven't set 12 mbs or above as the 大象传媒 HD standard? There can only be one answer here from your answer above, because Freeview can't accomodate 12mbs.

However, the point I've been trying to make is reducing all service below the optimium level just to make Freeview look good isn't platform neutrality but rather bias towards Freeview.

What the 大象传媒 should be doing is applying the optimum bit rate, be that 12mbs or more, then applying it to all platforms and in the case of Freeview where it can't be accomdated, instead applying the highest possible level that can. That is neutrality.

Thanks for the blog one of the visit.

Key points:

Have 大象传媒 admitted that the reduction in bit-rate has resulted in a reduction in PQ from what we got at the end of the Trial Period. [you reported them as saying they can not temporally increase the bit-rate as we will be disappointed when they have to reduce it]

I do not understand the statement that TV panels can not process 1920. I have a Full HD 1080 by 1920p panel that is used in unscaled mode!

#52 Bill-Taylor

"Have 大象传媒 admitted that the reduction in bit-rate has resulted in a reduction in PQ from what we got at the end of the Trial Period"

Not in so many words. What Andy said was that an increase in bitrate (presumably using the spare capacity on satellite) would only produce a small improvement in picture quality

What Stephen Baily (Acting Head of Distribution Technology) said was that increasing the bitrate now (on satellite), only to have to reduce it in the future (when the spare capacity was needed for another channel?) would lead to disappointment. This suggests there would be a noticeable improvement in picture quality, and that viewers would be disappointed when that picture quality wasn't maintained.

"I do not understand the statement that TV panels can not process 1920. I have a Full HD 1080 by 1920p panel that is used in unscaled mode!"

Join the club which no doubt also includes 1920 x 1080 TV manufacturers!

Further to the discussion on the reduced quality of freesat transmission.(not HD). Freesat signal strength on 大象传媒 channels seems to have reduced in the last week. We are in North West Wales, we now get "no satellite signal" on 大象传媒1 Wales and 大象传媒2 Wales but can get a reasonable picture on 大象传媒 West Midlands and 大象传媒 North West.
The same problem occurred late last year and then mysteriously went away again.
Our neighbours on freesat experienced the same issues.

Hi Andy
They are all SD channels, we do not have HD.
When we had the problem last year and again now the best reception
is on S4C. 大象传媒1 West Midlands and North West are OK but Wales and London
are very poor/ non-existent.
Cheers
John

The 大象传媒 HD bitrate and has recently dropped again. Last nights (10th June 2010) Opera Italia part 3 was broadcast at 7Mbps MBAFF 25fps instead of the usual 9.256Mbps PAFF 50fps.

TSReader showed the real-time bitrate hovering between 6.5Mbps and 7.5Mbps. TS Packet Editor showed an average bitrate of 7.156Mbps and a peak of 15Mbps. Just thought you'd like to know.

Just in case you're not aware upscaled SD on ITV 1 HD uses 9.8Mbps and looks fuzzy as you'd expect. The 7Mbps MBAFF 25fps still looks ok but not as good as 9Mbps PAFF 50fps. If the bitrate drops any further (to say 5Mbps) then 大象传媒 HD will start looking looking fuzzy like upscaled SD on ITV HD.

Hi Andy
re post 58
Our neighbour is 500 metres away over a hill.
This evening the picture on 大象传媒1 West Mids and ITV1 London are perfect, whereas 大象传媒1 Wales, 大象传媒2 Wales and ITV1 Wales get No Satellite Signal. Is this a Welsh problem?
John

Dear Andy Quested - yes, I suspected that you might be using VBR. A week or so ago you where broadcasting PAFF 50fps at 9.256Mbps. Picture quality was superb in this format. No movement between interlace fields gave a BluRay 1080p style picture - well almost.

When viewing a PAFF 50fps TS file with MediaInfo you'll see 2 frame rates. MediaInfo reports:
Frame rate: 50.000 fps
Original frame rate: 25.000 fps

If you strip the .264 video and .ac3 from the TS container and mux into a MKV container you have to use 50fps for the video and audio to be in sync. If you use 25fps the video is played in slow motion and out of sync with the audio.

The MBAFF 25fps is noticeably more blurred when viewed frame by frame because of the interlacing and the movement between fields. This is a bit of a step back into the past in my opinion. Admittedly when viewed in real-time (rather than frame by frame) no one is likely to see the difference unless you watch in slow motion.

As for the reason for VBR I assume this is to squeeze 3 or 4 HD streams into the one 36Mbps terrestrial multiplex that'll probably appear this Christmas. After the 6 terrestrial multiplexes were moved around last year to leave one multiplex reserved for HD I assume that you're now trying to squeeze 大象传媒 HD, ITV HD and Ch4 HD into this slot. Channel 5 have also said they'll be starting HD broadcasts on satellite sometime in the future so maybe you're trying to squeeze a forth HD channel into this one multiplex.

It would be better to get OFCOM to reserve at least 4 more frequencies to be used for future multiplexes and limit the number of HD channels per multiplex to 3x12Mbps HD broadcasts per 36Mbps slot. Then you'll have some spare capacity for the future. If all the spare frequencies are sold off it'll only cause problems in future.

Maybe you could find out if VBR is possible with PAFF 50fps but I realise that the aim of the exercise is to use the absolute minimum bandwidth so I understand why MBAFF 25fps is being used.

Sorry, I got the incorrect terminology with fields and frames.

PAFF at 50 fields per second with 2 fields per frame gives 25 frames per second. PAFF can be configured to have no movement (so I understand) between the 2 fields so the receiver can re-construct a 25 frames per second image or 1080p 25. The image quality can be made to be similar to BluRay 1080p. It's also possible to display the image or create files at 50 frames per second which is useful for fast moving stuff - sport, nature - that kind of thing but I assume that movement between fields must be allowed in this mode to give flow of movement. Most of the NRK Beta files is 50fps nowadays but it does increase CPU usage on playback and lower the quality factor which means using more bitrate.

MBAFF 25fps is like the traditional interlacing technique where you mix frames from different points in time giving the blurred image when viewed a frame at a time. When viewed in real time the human eye isn't quick enough to see the difference and software algorithms in the receiver can reduce the interlace effects but fast moving stuff is blurred.

That's how I understand it. As far as I can see 50 fields per second is the future for broadcasting in Europe. I've read that many American HD channels now use 60fps (I don't know if that's fields or frames) for broadcasts because of the improved picture quality for fast moving stuff.

My local transmitted is Waltham.

"Increasing the bit-rates for 大象传媒 HD now, using the spare satellite capacity, would only lead to disappointment when they are reduced again to make way for further channels. (NB: Take note of this statement - there is an inconsistent message here - Why would reducing the bit-rate lead to disappointment if, as we were told elsewhere, bit-rate has minimal effect on PQ?)"

To play Devil's advocate - couldn't he just be referring to disappointment amongst viewers monitoring the bit rate and being pleased it had increased then disappointed it had decreased rather than saying it would be due to a change in PQ? I'm not saying whether that would/wouldn't be the case, just that the statement can be read in more than one way.

#66 Alan Robertson

"couldn't he just be referring to disappointment amongst viewers monitoring the bit rate and being pleased it had increased then disappointed it had decreased rather than saying it would be due to a change in PQ"

An interesting perspective, but I was present when that statement was made and the context of the discussion around it was picture quality.

That statement has now been vindicated. With the introduction of VBR, the encoder has been utilising bitrates 70% higher than when it was fixed at 9.7 Mbps and viewers have been reporting improved picture quality coincident with that higher use of bitrates. Those picture quality improvements being sharper definition, improved clarity and less motion blur.

#68 Andy Q

Thank you for your comment on my post.

I am well aware that several changes were introduced over a short period of time, for which we all appreciate the efforts of the 大象传媒 HD team in introducing them and the improved picture quality resulting from them.

The fact that the encoder, in VBR mode, now uses bit rates as high as 16.5 Mbps for challenging scenes is proof in itself that an increase in bit rate was necessary and that the increased bit rate is indeed playing its part in improving HD picture quality, which is what we have been maintaining all along would be the case.

#70 Andy Q

Thanks Andy.

A suggestion. Can parts 1 & 2 of the blogs be closed for further comment with redirection to blog 3 for all future comments?

#72 Agreed

Thanks for all the comments, folks - didn't expect so many on a Sunday! All the talk about statmuxing, etc. takes me back to the days of ONdigital! (I had one of the first boxes and received it via a small set-top aerial thanks to being in a flat on a hill with a clear view of Black Hill) - even remember 大象传媒 Choice and 大象传媒 Knowledge (as they were called in those days) sending out some researchers to hold a focus group with me and my flatmates! Things have certainly moved on leaps and bounds since then :-)

#72 Andy Quested

The viewers group are happy for blogs 1 & 2 to be closed for comment, with a final post on each redirecting to blog 3.