My question above seems to have gotten lost in the shuffle.

To come at this from a slightly different perspective, for Amir or any other Microsoft insider, any thoughts on why it is that Universal seems to generally get so much better results from VC-1 than Warner does?

I don't understand why people continue to think that Warner is softening their transfers by using poor compression methods. Just because of a couple of movies that were shot with soft lens filters? (ie. Batman and Vendetta). Even Superman looked exactly as I remember it on dgital projection.

What about all of the sharp transfers like Training Day, Grand Prix, Four Brothers, ATL, The Searchers and The Ant Bully? Just off the top of my head.

Nils, do you want to pick one or two questions to ask and then I can address? This is not a thread for 10-part arguments/rants about issues :). I have good answers to everything you asked btw :). Moving your question to the other threads won't do any good as I don't participate in any other thread than this one on HD optical topic.

Alternatively, you could wait until I have time to write up to a question asked a while ago, which is what science in AVC would make it have softer video than VC-1 (this is the signature btw. I have said it a dozen times as has Ben so I don't know why you still ask what the signature is).

Sure, here are a couple of my main questions, the others you can ignore or come back to when you get the time...

Since you claim there is a definite signature on AVC encodings (softness), are you saying that both MPEG2 and AVC cannot reach visual transparency using the bitrates avialable to it on Blu-ray?

When HD DVD's reviewed are rated less than 5/5, is that always the fault of the master?

Thanks.

2. Support for 44.1Khz (and multiples) for audio HD DVD discs to support cases where the production is done natively in that sampling rate. The request had come from the music industry to use HD DVD as a high-quality music delivery system.

Hello Amir,

Has it anything to do with the DVD Audio format from the good old DVD?

I wish, since there aren't almost any releases any more.

Thank you.

Hello Amir:

You mentioned previously about how all three codec's were compared against each other and VC-1 won/approved, I'm not sure if it was through the DVD Forum or some other org.

Anyway with the recent posts about having new A/B blind tests between codec’s I was wondering if the same type of review that VC-1 initially won was a one time comparison or do they re-review all the codec’s as time goes on and the codec’s evolve with the technology?

Hello Amir:

You mentioned previously about how all three codec's were compared against each other and VC-1 won/approved, I'm not sure if it was through the DVD Forum or some other org.

Anyway with the recent posts about having new A/B blind tests between codec’s I was wondering if the same type of review that VC-1 initially won was a one time comparison or do they re-review all the codec’s as time goes on and the codec’s evolve with the technology?

Hello Tom. Good question :).

The original DVD Forum tests consisted of two shoot outs. We won the first one. People thought we cheated :) (used pre-/post-processing to get better results which we had not). So they made the rules much more strict, but we won again. After this, VC-1 was made mandatory in the standard so there was no additional need for such tests.

However, the studios are not obligated at all to believe the above data even though most of them participated in it. As such, one by one, they conducted their own shoot outs, with their own material, data rates, and test conditions. This is how we won the support of Warner, Paramount and Universal. As there is no exclusivity arrangement here, the studios continue to evaluate new revisions of encoders from different companies and as such, we are in constant pressure to do better than our competition. This is why we know so much about the status of competing encoders and have gained in our confidence to stay ahead of them.

Your basic point that codecs, especially the new ones, keep getting better is very much true. But it applies equally to VC-1 as it has also improved (and rather rapidly) since the original DVD Forum tests.

As another meta point, folks should ponder why the codecs which did not win the shoot out, still became mandatory in the standard! I can assure you that if we had not won, there would not be a trace of our technology in either format. Anyone who tells you that business/other motivations don't get in the way of best man getting picked for the job, hasn't spent a minute in the circles where such decisions are made...

Hello Amir,

Has it anything to do with the DVD Audio format from the good old DVD?

I wish, since there aren't almost any releases any more.

Thank you.
In a way yes. By adding these sampling rates, you get the same suite as DVD-A because TrueHD is the same as MLP used in DVD-A compression. But in a way, it is better in that DVD-A had an expensive (to license) navigation system. In the case of HD DVD, one simply makes an audio-only playlist (my preference) and you have a nice CD-type of experience which doesn't cost a penny more in licensing costs. And of course, you can also go crazy with HDi menus and such and blow away anything DVD-A could (not my preference).

Having said this, my personal opinion is that the time has come and gone for any kind of mass market around high-resolution optical formats. I much rather download the same files over the internet when I want it, than to get stuck with a few specialty retailers stocking a couple of discs here and there. And most of the world could care less about any improvements over CD it seems.

Amir,

I would like to hear your comments on what you are seeing in these MPEG-2 vs. VC-1 comparison uncompressed PNGs:

http://www.avsforum.com/avs-vb/showthread.php?t=813105

Amir,

I would like to hear your comments on what you are seeing in these MPEG-2 vs. VC-1 comparison uncompressed PNGs:

http://www.avsforum.com/avs-vb/showthread.php?t=813105
I am on a slow cellular modem in Hiroshima so can't look at the pictures easily. I have asked the team to look at it and when I hear back, and finish taking pictures at Miyajima (hey, work without play doesn't make sense in my book! :), I will comment.

For now, I am surprised that you would say this in that thread, and then expect me to respond to you here:

The whole "my movie must be in VC-1" hype that seems to have died down on here seems to be just that, hype. Portrayed and spread very slickly by Microsoft insiders on here I might add to a very gullible audience.

If you don't believe my answers, not sure why you are asking me more questions here. We are not gladiators who you poke to fight for entertainment. You should at least give us some food and water before doing that :D.

I feel it's time to post something I've been PM-ing to those who have had their posts deleted:

"This thread is not a court of law. You're not a lawyer trying to catch someone in a lie or misstatement. Or badger them until they come around to your way of thinking. Take the answers the insiders give you and you do with them what you will. Questions. Answers. That's all that goes in that thread. I understand that the answers may come off as slanted. No surprise, there. Most of our insiders are admittedly not independent. You might not agree with the answers you get but this thread is not the place for challenging those answers. There is a whole forum where you can discuss it in an intelligent, non-bashing, non-confrontational way."

Ask your question. Get your answer. Move on.

Thanks

I am on a slow cellular modem in Hiroshima so can't look at the pictures easily. I have asked the team to look at it and when I hear back, and finish taking pictures at Miyajima (hey, work without play doesn't make sense in my book! :), I will comment.


Dear Amir,

while you are at it please ask your team to take a further look at the recent "The Departed" VC-1 transfer.

Despite rather good marks "in the press" this Warner VC-1 transfer contains some of the most questionable sequences (purely from a technical point of view, the movie itself fully deserves it's praise) I have ever come across during my career in this business.

I'm preparing a thread dedicated to show the defects of this transfer. Right now I'm still figuring out a way to take meaningful "screenshots" (digital screen captures).

Before that please allow me to ask you for your professional opinion about this transfer.

Please ask your team for example to take a look at chapter 33 (dialog in hallway). There is some SERIOUS noise reduction going on. This scene is just an example of the worst case of noise reduction I have ever seen applied on any disc based HD transfer so far. The Departed has quite a lot of natural film grain. eg. in Chapter 33 there is no grain at all. It is not just that excessive NR is applied but the used filter is subpar - it smooths and smears the picture significantly, ALL detail and definition is lost, typical noise reduction artifacts are all over the place AND the VC-1 video stream AVERAGES 6.5 MBit/s during this scene (which is not surprising as there is no detail left to encode). Wait, it gets even better. After each cut in this scene the first 2 frames contain excessive amount of (natural) film grain. This film grain "magically disappears" from the third frame of each cut onwards - till the next cut. Cheap NR at it's worst.
As mentioned the VC-1 bitrate during this scene averages ~ 6.5 MBit/s. The very next scene of chapter 33 is not filtered - film grain is still present - and the VC-1 bitrate suddenly jumps to ~20Mbit/s. I don't presume VC-1 encoding itself is at fault regarding this inconsistent and excessive filtering (chapter 33 is just one example of many) but I would like to ask you why studios using VC-1 choose to filter their transfers prior to encoding in the first place. I just finished watching Chicago from Buena Vista - this AVC encoding with regular 40MBit/s peaks conserves the inherent, excessive film grain in a wonderful and natural way (remark for the moderators: suitable for this thread I worked as a projectionist for several years - so I would consider myself a "insider" regarding this matter).

Why do studios supporting your codec choose to apply excessive filtering/NR for many of their releases (Warner's Batman Begins, Charlie and the Chocolate Factory being other prime examples)? Does MS/the VC-1 codec team encourage this? If not, are you discussing this matter with the studios?

Looking forward to your comments! Thank you very much for your time.

Amir-

Any info on what's up with TL-51? I thought Toshiba would have submitted it this week for approval... Are they still planning on submittied it and having it approved by the end of the year?

Thanks,
Chris

Hello Tom. Good question :).

The original DVD Forum tests consisted of two shoot outs. We won the first one. People thought we cheated :) (used pre-/post-processing to get better results which we had not). So they made the rules much more strict, but we won again. After this, VC-1 was made mandatory in the standard so there was no additional need for such tests.

However, the studios are not obligated at all to believe the above data even though most of them participated in it. As such, one by one, they conducted their own shoot outs, with their own material, data rates, and test conditions. This is how we won the support of Warner, Paramount and Universal. As there is no exclusivity arrangement here, the studios continue to evaluate new revisions of encoders from different companies and as such, we are in constant pressure to do better than our competition. This is why we know so much about the status of competing encoders and have gained in our confidence to stay ahead of them.

Your basic point that codecs, especially the new ones, keep getting better is very much true. But it applies equally to VC-1 as it has also improved (and rather rapidly) since the original DVD Forum tests.

As another meta point, folks should ponder why the codecs which did not win the shoot out, still became mandatory in the standard! I can assure you that if we had not won, there would not be a trace of our technology in either format. Anyone who tells you that business/other motivations don't get in the way of best man getting picked for the job, hasn't spent a minute in the circles where such decisions are made...

Amir,

Sony Pictures participated in both of these events and my recollection of the final results is different. First off, the tests were based primarily on sub 10Mbps encoding, the total amount of encoded content was on the order of minutes and there were no direct view full resolution panels present. If you are claiming that these events proved anything accept that a reasonable picture could be produced as judged on 5 year old display technology, then I beg to differ.

We received this message which I think has some good advice:

Please ask the insiders to stop this tit for tat with each other. Half of these comments seems little more than a propaganda war between MS and Sony and they are doing a disservice to the forum membership.

It would be helpful to all for these insiders to STOP addressing comments to one another and to confine their posts to answering questions posted by forum members.

Just as this forum is not a court of law where members get to badger the insiders, this is also not a space where insiders should be allowed to argue against one another in a similar fashion.

What is good for members is also good for the insiders.

Thank you.

and then we receive this:

"You did have this as the one place for conversation between insiders

At least their tit for tat gave some information and was restrained or affected their credibility if it got out of hand.

If they cannot challenge the information of other insiders no one can and will lead to even more propaganda.

I think the last change in policy may have some unintended impact.

I like talkstr8t and paidgeek challenging amirm and benwaggoner and vice versa. If no one AVS member or even insider can , it might lead to pure propaganda.

Sorry I disagree with you here, because of unintended consequences"


So Insiders: carry on : this was not intended as a change in policy: just repeating some comments we thought were interesting :o

In 5 years you will probably be be storing your movies on a server in Utah and retrieving them via Verizon FIOS at 100mb or better download speed over Fiber Optic lines. Hmmm, I wonder what encoding format they will use? :)

All sort of kidding aside. The cost will come down.

This will be driven by a combination of price pressure from retailers and pricing / positioning decisions made by Major Studios.


The question that comes to my mind is: what will happen after 5 years, is the cost for BD movies going up for the consumers?

Richard

Welcome to AVS

Thanks for posting here

My pleasure, I trimmed my post because I felt it was off topic..trying to keep everyone happy here..:)

Rich

Richard

Welcome to AVS

Thanks for posting here

May I ask what Encode format we are comparing VC1 to?

Amir,

Sony Pictures participated in both of these events and my recollection of the final results is different. First off, the tests were based primarily on sub 10Mbps encoding, the total amount of encoded content was on the order of minutes and there were no direct view full resolution panels present. If you are claiming that these events proved anything accept that a reasonable picture could be produced as judged on 5 year old display technology, then I beg to differ.

What formal and informal lines of communications are there between BD & Universal Studios and HD DVD & Fox/Disney/Sony? Do the sides meet up regularly (monthly, quarterly, semi-annual, etc.) to discussion the state of affairs and to check up on progress?Speaking for the former (BD/Uni), I'm aware of no formal discussions. One would assume that individual BD company execs use whatever relationships they have with Universal execs to encourage Universal to support Blu-ray.

I can't speak from experience, but I would imagine the same scenario applies for HD DVD and Fox/Sony/Disney. Of course, there are far more companies with far more execs in the Blu-ray camp to lobby a single studio, Universal, than there are HD DVD companies/execs to lobby Fox and Sony and Disney...

To use an audio analogy, if I you knew that a speaker was "bright" (i.e. had exaggerated highs), would you need proof on every song that it sounds bright? Surely you form your opinion appropriately after listening to your favorite tracks on it, and observing the same results over and over again. Such is the case with AVC and softening grain and detail as compared to VC-1.Amir, this strikes me as a gross oversimplification. A speaker is a physical construction with a bit of electronic shaping available. You might be able to tweak the sound slightly via changing the crossovers, amplification, etc., but the physics set some pretty hard limits.

A compression spec, on the other hand, can produce wretched results or extraordinary results depending on who does the implementation, efficiency of algorithms used, computational horsepower applied, etc. Joe Kane, an industry expert whose opinion I suspect you value very highly (based on his services frequently being hired by Microsoft) has said that AVC quality will most certainly surpass that of VC-1 as processor improvements allow more complex algorithms to become feasible. This opinion surely doesn't square with your contention that VC-1 is essentially always going to do a better compression job than AVC.

No persistent storage is required for these players.Paidgeek, I believe you are incorrect. There are two types of non-volatile memory ("local storage") defined in the Blu-ray spec - the Application Data Area (known as "Persistent Storage" in MHP) for preferences, bookmarks, and such, and the Binding Unit Data Area for storage of A/V clips. All Blu-ray players must support the Application Data Area, albeit not a huge amount. All BD-Video 1.1 and BD-Live players must also support the Binding Unit Data Area (256MB for BD-Video and 1GB for BD-Live).

- Talk

Sony Pictures participated in both of these events and my recollection of the final results is different. First off, the tests were based primarily on sub 10Mbps encoding, the total amount of encoded content was on the order of minutes and there were no direct view full resolution panels present.In addition, as I recall the event didn't include AVC high profile as one of the tested codecs, only main profile.

- Talk

Unfortunately, there is no way to make such generalizations. Every movie will be subjected to so many production steps and mastering that it is hard to know what shape they wind up in by the time the land in a VC-1 encode pipeline. GDMX (Warner's post house) has stellar expertise in VC-1 encoding, going back two years and with wide open support channel to us. So little fault can be put in that part of the process as opposed to what Uni uses.

Amir, perhaps then you might consider my original question:

"Originally Posted by patrick99
Amir, in light of your extensive comments on Batman Begins and on VC-1 as compared to AVC, I wonder if you have seen the BD of The Prestige? If so, do you think that the differences that appear to exist in the PQ are attributable to differences in the masters? Since the director and the DP are the same on both, and since both have similar colorations, one would have expected both to look very similar. Thanks"

I think that Universal has demonstrated repeatedly that VC-1 is capable of producing outstanding results. But I assume you would agree that using VC-1 doesn't guarantee outstanding results.

It seemed to me that the comparison of Batman Begins versus The Prestige might give some insight into the question of whether the softness in BB is attributable to softness in the original material or is instead an effect introduced during the transfer process.

So is it possible that you might address my original question?

Thanks

Amir,

Sony Pictures participated in both of these events and my recollection of the final results is different. First off, the tests were based primarily on sub 10Mbps encoding, the total amount of encoded content was on the order of minutes and there were no direct view full resolution panels present. If you are claiming that these events proved anything accept that a reasonable picture could be produced as judged on 5 year old display technology, then I beg to differ.

paidgeek -

Interesting. Are you willing to suggest rules for a new shootout? Would Sony participate?

- Tom

We are currently performing an Encode Test using several different types of content and several different labs. I am open to anyone who wants to contribute time, services, or content to the cause.

We would be using the following content:

1 - 100% Animation – Animation with a great deal of fine detail and precise motion. This is challenging content and we have used it in the past for standard DVD Compression tests. None of the encoders could produce results that we were happy with, but we did find one that was better than any of the others.

2 – 50/50 Hybrid – A mix of 1080i footage with CGI Overlays with a few very tricky elements that are encoding “challenges.”

3 – 65mm Film – Clips from a variety of IMAX films ranging from excellent to poor condition with and without pre-encode processing.

In general, we feel that we know the strengths and weaknesses of each Codec well enough to choose content designed to show their strengths and reveal their weaknesses.

It is easy for any encoder to look good if the film is squeaky clean and heavily processed to remove all dirt, noise, grain, artifacts, gate weave, flicker, and other artifacts prior to encoding. I do not consider encoding “pristine” prints ad true test of encoding capability.

For the record, I offered this to Microsoft.

Posting screenshots of paused high def titles currently seems to be the most effective way for members to make visual comparisons of the same scenes and then discuss the advantages and disadvantages of the specific titles and the differences between them.

As with most public comparison tests more effort is usually put in by some members to debate how the test was done in the first place rather then the results of the test. Example in how the different devices are calibrated, what type of camera was used, how many megapixels the camera has, size & resolution of the posted pictures, whether the player displays differently while paused etc…

I’d like to ask reps from both formats what you would consider to be fair as far as taking these screenshots for posting and making public comparisons by us members?

This way we can have a bit more uniformity and credibility with taking and posting these pictures for comparisons and the results wont be unfair towards either format.

Thank you in advance

We are currently performing an Encode Test using several different types of content and several different labs. I am open to anyone who wants to contribute time, services, or content to the cause.

We would be using the following content:

1 - 100% Animation – Animation with a great deal of fine detail and precise motion. This is challenging content and we have used it in the past for standard DVD Compression tests. None of the encoders could produce results that we were happy with, but we did find one that was better than any of the others.

2 – 50/50 Hybrid – A mix of 1080i footage with CGI Overlays with a few very tricky elements that are encoding “challenges.”

3 – 65mm Film – Clips from a variety of IMAX films ranging from excellent to poor condition with and without pre-encode processing.

In general, we feel that we know the strengths and weaknesses of each Codec well enough to choose content designed to show their strengths and reveal their weaknesses.

It is easy for any encoder to look good if the film is squeaky clean and heavily processed to remove all dirt, noise, grain, artifacts, gate weave, flicker, and other artifacts prior to encoding. I do not consider encoding “pristine” prints ad true test of encoding capability.

For the record, I offered this to Microsoft.

Richard -

Electronic cameras are increasingly able to offer 1080p/24 (or faster) material that might on good days approach 'squeaky clean'. Is that also a criterion?

Dunno how to phrase this as a question (what was that old quiz show?) but it seems the very existance of profitable electronic media distribution will exert a force to create movies that look good after being electronically coded. So wouldn't you also want to test some of these?

- Tom

Paidgeek, I believe you are incorrect. There are two types of non-volatile memory ("local storage") defined in the Blu-ray spec - the Application Data Area (known as "Persistent Storage" in MHP) for preferences, bookmarks, and such, and the Binding Unit Data Area for storage of A/V clips. All Blu-ray players must support the Application Data Area, albeit not a huge amount. All BD-Video 1.1 and BD-Live players must also support the Binding Unit Data Area (256MB for BD-Video and 1GB for BD-Live).

- Talk

Talk,

Strictly speaking, you are correct, but since the question was with regard to version 1.0 players and only 64KB is required, I rounded to zero....

RBFilms - I suggest that in addition to those you already mentioned you try film content that's been cleaned up and edited using 1080i@60Hz equipment, as was done early on before 24psf/50i became the standard. Those will typically have difficult cadence breaks causing challenges for the equipment. For the animation test try both CG and cel based animation, as they're difficult to handle in different ways.

paidgeek -

Interesting. Are you willing to suggest rules for a new shootout? Would Sony participate?

- Tom

Tom,

There is not much upside for Sony to participate in a codec shootout at this point. The obvious problem is the the possibility of endless second guessing as to whether such a shootout has been conducted properly. If an impartial shootout can be arranged, SPE can contribute some samples for encoding that are known to be difficult for all of the encoders in current use.

Another issue that should be considered is the characteristics that a codec imparts on the results. For instance, in a shootout, one may compare the results of each codec side by side and a preference may be shown for the sample which looks "clean". Herein comes the question of whether codecs should be chosen based on how they potentially alter the look of the master versus how close an approximation they achieve of the master. For my part, any shootout, should allow side by side comparisons of the master to the encoded samples using identical, full resolution, properly calibrated displays.

Regardless of my own feeling about color, focus or film grain on any particular title, we have spent enough time with film makers to know they don't want their work altered according to the "taste" of a particular market, compressionist or studio. The agreement is, we are provided an approved master and our job is to reproduce it as accurately as possible on whatever the distribution format is: Television, DVD, BD and so on.

Sorry for the longish answer to your simple question...

Posting screenshots of paused high def titles currently seems to be the most effective way for members to make visual comparisons of the same scenes and then discuss the advantages and disadvantages of the specific titles and the differences between them.

As with most public comparison tests more effort is usually put in by some members to debate how the test was done in the first place rather then the results of the test. Example in how the different devices are calibrated, what type of camera was used, how many megapixels the camera has, size & resolution of the posted pictures, whether the player displays differently while paused etc…

I’d like to ask reps from both formats what you would consider to be fair as far as taking these screenshots for posting and making public comparisons by us members?

This way we can have a bit more uniformity and credibility with taking and posting these pictures for comparisons and the results wont be unfair towards either format.

Thank you in advance

The results of screenshots can be very misleading because, as I have mentioned in my previous posts, all encoded movies have flaws. Consider this, every encoder benefits from customization (a human looking at the encoded stream and adjusting bit rate, or other parameters). Since customization is useful, it stands to reason that all encoders produce artifacts (as they must given between a 50:1 and 100:1 reduction in data from the master). If you accept the above, then the question becomes, what can the codecs do with the least opportunity for human error (in requiring a great deal of customization), with reasonable bit rates and on difficult titles (those that have high entropy etc.). All this is further complicated by the fact that film grain is not an artifact and in the absence of a master to reference, the encoders success at capturing it can't really be judged.

To summarize, screenshots will tell you less about the codecs and more about whether or not the compressionist was paying close attention or had the right equipment (e.g. displays) to find faults during encoding.


PS.
Even though screenshots can be misleading, if you have the opportunity to view the same movie, side by side from two codecs, it can be useful and revealing. Full rez direct view LCD's are getting inexpensive right? just borrow the one from the bedroom and put it beside the matching unit in the family room. Kidding aside, this type of direct comparison is most telling...

Question about 2.35 subtitles and menus to all insiders -

Many pages ago there was discussion on how to accommodate users with constant height systems by adding some code (HDi and BD-Java??)that would automatically allow the menus and subtitles to fit in the 2.35 frame vs the letterbox or black bar portion of the 1.78 frame.

Have any studios expressed interest in doing this or should we just give up hope on this front? If there was interest are there any US titles that are released or soon to be released that will have this feature?

Thanks for your time,
Jeff

I don't know about menus, but for subtitles I imagine that kind of positioning trick would easily be possible if the studios used scripted subtitles instead of bitmaps.

Since we're on the subject, one thing I would like to see is the possibility of user-added subtitles in some way, or at least support for downloading official subtitles for titles that were created/made after the disc was released. That way you could buy the disc for your local region, and then get extra subtitles for your preferred language. I know it's not exactly a priority (most english native speakers probably don't even think about it), but it's one of those things that could provide extra value to customers so they don't have to import discs and players from other regions to get their preferred language.

Question for Amir or anyone else involved with MS:


I read the Toshiba HD-DVD player firmware upgrade details.
http://www.highdefdigest.com/news/show/Toshiba/Firmware_Upgrades/Hardware/Toshiba_HD_DVD_Firmware_Upgrade_V1.3_Now_Available/496
Some of the issues addressed appear to be similar to issues found on the HD-DVD add-on player for the Xbox 360.

Are these same "fixes" shared for the Xbox 360 and its upcoming update, as the player is made by Toshiba I believe?

This update is already available for the Toshiba players, although the Xbox 360 update appears to be coming with the spring system/Dashboard update.


Thanks.


-Brian

Are these same "fixes" shared for the Xbox 360 and its upcoming update, as the player is made by Toshiba I believe?
Toshiba (really, TSST -- Toshiba/Samsung optical manufacturing arm) only provides the drive to us. All the software comes from us and not from Toshiba.

As we find bugs, if they are in the same HDi implementation we provide to Toshiba, we also give a drop to them. It is their option to provide that update to their customers. There is little the comes back in reverse though as Toshiba does not provide any code to us for Xbox 360.

Thanks, Amir!

We are currently performing an Encode Test using several different types of content and several different labs. I am open to anyone who wants to contribute time, services, or content to the cause.

… None of the encoders could produce results that we were happy with, but we did find one that was better than any of the others.

...

In general, we feel that we know the strengths and weaknesses of each Codec well enough to choose content designed to show their strengths and reveal their weaknesses.

For the record, I offered this to Microsoft.
Hmmm. I just checked with my team and none of the technical people seem to know anything about your project, let alone having the clips and such, or participated in your testing. They also have not had any contact with you one way or the other. The only thing I personally know about is you offering to send us the current retail version of your title in MPEG-2.

Maybe you have been working with Ben and given the fact that he has been sick, the content hasn't made it to the codec team. Please follow up again as we are happy to help out here in any way we can. Until then, please kindly note that until you work with us directly and we can bring you up to speed on ins and outs of VC-1 and our tools, I don't agree that you know the weaknesses and strength of VC-1 :). Can you also clarify if you are doing your own in-house encodes and not using an outside shop as we think you currently are?

On MI3 VC-1 vs MPEG-2….

Just heard back from the team that that the preliminary assessment is that the segment in question does indeed show some artifacts. It is our opinion that this could have been remedied easily as the segment is pretty easy to encode so one could easily fine tune it/bump the data rate.

What is puzzling though, is Benes’ claim of he got those screen snapshots. He says he used Media Player Classic with a filter for VC-1. This doesn’t compute. Media player (any version) lacks code to properly de-multiplex (separate the a/v streams) in HD DVD. HD DVD shipped some 8 years after the last update of Media Player Classic! While we use VC-1 in WMV encoded files for PC use, those files use our file format ("ASF"), which is wildly different than what is used in HD DVD. So the player has no way of playing HD DVD content, even if the copy protection is stripped away. We have provided specialized tools to play such files without access to an HD DVD player but I don't see how Benes' would have access to such a tool, without working for one of the companies with access to our professional encoder toolset.

In addition, our VC-1 decoder that he says he used, lacks post processing as that was too CPU intensive for the time period we release that code. So you are bound to see more artifacts in some cases as compared to real HD DVD hardware (sadly, I understand some software HD DVD players lack the same post processing).

Anyway, maybe adventurous hackers out there have reformatted those streams to get it to play in Media Player (but heaven forbid, have not actually re-encoded it). Or that there is a de-muxer I don’t know about but Benes does but didn’t list. Either way, it would have been good to have such data so that we know how to investigate. Otherwise, I am left wondering how really got those shots…

As an aside, he should also rename the title of his thread. The MPEG-2 encode is probably running at 2X the data rate of VC-1. Remember, the HD DVD version actually has two video streams with a Picture-in-Picture on top of the move itself – BD version in MPEG-2 lacks that. So if that scene is the pinnacle of what is wrong with the VC-1 encode running at half the data rate of MPEG-2, while at the same time providing a secondary video stream, I still stand proud even if this were an issue with VC-1 :).

Maybe you have been working with Ben and given the fact that he has been sick, the content hasn't made it to the codec team. Please follow up again as we are happy to help out here in any way we can. Until then, please kindly note that until you work with us directly and we can bring you up to speed on ins and outs of VC-1 and our tools, I don't agree that you know the weaknesses and strength of VC-1 :). Can you also clarify if you are doing your own in-house encodes and not using an outside shop as we think you currently are?It sounds like only Microsoft can bring out VC-1's capability and posthouses that have been using Microsoft provided encoder can't do that. Is it what you meant? If so, "shootout" with Microsoft's hand tuned stream won't really simulate the quality of actual production, unless all VC-1 titles are going to be encoded by Microsoft?

Maybe compression shootout by one posthouse which has capability of all encoders and has been doing Hollywood titles with those systems may reflect more "real" productivity/quality of actual products?

Or that there is a de-muxer I don’t know about but Benes does but didn’t list. Either way, it would have been good to have such data so that we know how to investigate.If you check the doom9 forum, you'll find EVO demultiplexer is there.

The MPEG-2 encode is probably running at 2X the data rate of VC-1. Remember, the HD DVD version actually has two video streams with a Picture-in-Picture on top of the move itself ? BD version in MPEG-2 lacks that. So if that scene is the pinnacle of what is wrong with the VC-1 encode running at half the data rate of MPEG-2, while at the same time providing a secondary video stream, I still stand proud even if this were an issue with VC-1 :).I think bene can strip out both elementary stream of VC-1 and MPEG2 as well as secondary PiP VC-1 stream (above demultiplexer can do this), and would let us know the "real" bitrate of both, if he wanted.

It sounds like only Microsoft can bring out VC-1's capability and posthouses that have been using Microsoft provided encoder can't do that.
Not at all. Go to GDMX, DLUX, Cinram, etc. and you can get top notch encoding done in VC-1. That is because their people have attended our training and know how to use the system. But more importantly, they have experience with it. One can’t say they know VC-1’s capabilities, not having done any encodings of their own.

VC-1 is an advanced codec with a lot of features. And our tools are designed with professionals in mind who go through training, rather than casual users (the level of documentation simply is not there for this kind of use). It is not like we charge for our training anyway. We are more than happy to help out, including such specialized and small companies such as RB’s. Given that, there should be no excuse not to take us up on learning how to use our tools and technology.

BTW, I hear Panasonic now has a preference for doing the encoding for studios, rather than selling their AVC encoder to others. If so, that shows they are uncomfortable supporting users of their encoder. Or they think they can do a better job than others. So if are more confident in AVC being used by others, what’s up with this position by Panasonic?

Is it what you meant? If so, "shootout" with Microsoft's hand tuned stream won't really simulate the quality of actual production, unless all VC-1 titles are going to be encoded by Microsoft?
Per above no. But as people who invented VC-1, and not getting paid a fixed budget to do a project (or paid at all by the studios :)), it is true that we can probably extract that last bit of quality out of an encode. But that is not synonymous with others not being able to do an excellent job. To date, every VC-1 encode is done by a post house, not by us. And that includes the many reference quality titles.

Maybe compression shootout by one posthouse which has capability of all encoders and has been doing Hollywood titles with those systems may reflect more "real" productivity/quality of actual products?
Unless that one post house does the encoding for the entire industry, with the same operator who does the shoot out, and they spend equal time on the shoot out than they do on real title, with the test title being representative of all titles they encode, and him/her not having too much to drink on the day encoding, the test, maybe you have a point :). Otherwise, read Paid’s post for excellent reasons on why public shoot outs are not of much value.

If you check the doom9 forum, you'll find EVO demultiplexer is there.
Thanks. I don't hang out there anymore. But sure would have been nice for him to give the details as otherwise, we don't know how to replicate his results. Neither does anyone else.

I think bene can strip out both elementary stream of VC-1 and MPEG2 as well as secondary PiP VC-1 stream (above demultiplexer can do this), and would let us know the "real" bitrate of both, if he wanted.
Measuring peak rate is difficult without specialized tools (hence the reason PS3 has trouble with it). Just because you can play it, doesn’t mean you can measure the instantaneous peaks. I suspect you have the data already but just not saying it. Are you claiming that MI-3 uses HD DVD data rate for its MPEG-2 encode then?

Not at all. Go to GDMX, DLUX, Cinram, etc. and you can get top notch encoding done in VC-1. That is because their people have attended our training and know how to use the system. But more importantly, they have experience with it. One can’t say they know VC-1’s capabilities, not having done any encodings of their own. So, if someone wants to know the strength and weakness of some codecs, asking those posthouses to do encoding supposes to work for getting quality in real world, right? It sounds little bit contradicting what you said "Until then, please kindly note that until you work with us directly and we can bring you up to speed on ins and outs of VC-1 and our tools, I don't agree that you know the weaknesses and strength of VC-1 :)", anyway...

Measuring peak rate is difficult without specialized tools (hence the reason PS3 has trouble with it). Just because you can play it, doesn’t mean you can measure the instantaneous peaks.It's not difficult job to measure the amount of bits certain picture is using, if someone knows bitstream syntax and has elementary data. And, anyone can download SMPTE 421M specification.

I suspect you have the data already but just not saying it. Are you claiming that MI-3 uses HD DVD data rate for its MPEG-2 encode then?I don't have those numbers, and no interest to do that. BTW, have you seen this bene's thread (http://www.avsforum.com/avs-vb/showthread.php?t=760714)? If you are interested in ABR of BD titles, it's quite telling you something.

So, if someone wants to know the strength and weakness of some codecs, asking those posthouses to do encoding supposes to work for getting quality in real world, right? It sounds little bit contradicting what you said "Until then, please kindly note that until you work with us directly and we can bring you up to speed on ins and outs of VC-1 and our tools, I don't agree that you know the weaknesses and strength of VC-1 :)", anyway...
No it is not contractory. RB said he knew the difference between the codecs, not that his post house did. For him to have such personal experience, he would need to use to learn to use the tools and have done some work with it.

It's not difficult job to measure the amount of bits certain picture is using, if someone knows bitstream syntax and has elementary data. And, anyone can download SMPTE 421M specification.
Have you done it? If so, then you would know it is not so easy. Even our tool approximates the rate. The only way to have 100% accurate data is to look at the encoding report. Otherwise, things like the buffer model and such, get in the way.

Rio, any comment on the Panasonic encoder business model? I am sure someone in your company has the answer ;).

Amir "Media Player Classic" is a 3rd party media player that's coded from the ground up, and is not a Microsoft product. While it mimics the look of the old media player included in windows, that's merely a matter of cosmetics:

http://sourceforge.net/projects/guliverkli/

It has full support for customized filter graphs, and also some internal codec support. If someone says they used MPC to play back a HD-DVD my guess is they're using additional dshow filters to achieve that as the current version doesn't come with it out of the box so to speak.
Ah! We use the term "classic" to also talk about our original OCX control :p. Thanks for clarifying that. My comments regarding our video decoder stands though.

In addition, our VC-1 decoder that he says he used, lacks post processing as that was too CPU intensive for the time period we release that code. So you are bound to see more artifacts in some cases as compared to real HD DVD hardware (sadly, I understand some software HD DVD players lack the same post processing).Amir, Can you clarify that all of the Toshiba standalone players that use the Broadcom chip do in fact have the "post processing" for VC-1? It was "I imagined" elsewhere that it is missing also in those players.

Thanks!

Amir, Can you clarify that all of the Toshiba standalone players that use the Broadcom chip do in fact have the "post processing" for VC-1? It was "I imagined" elsewhere that it is missing also in those players.

Thanks!
Yes it does have it. They are not constrained by the PC power of the yesterday.

Amir, perhaps then you might consider my original question:

"Originally Posted by patrick99
Amir, in light of your extensive comments on Batman Begins and on VC-1 as compared to AVC, I wonder if you have seen the BD of The Prestige? If so, do you think that the differences that appear to exist in the PQ are attributable to differences in the masters? Since the director and the DP are the same on both, and since both have similar colorations, one would have expected both to look very similar. Thanks"
I have not yet watched it. I was sick and now in Japan and way behind watching movies :). But even if I did, I wouldn't be able to say anything useful here. I already mentioned that even within one movie, Batman Begins, there are variations in apparent sharpness. So what can you expect, comparing different movies? Really, there is nothing to be learned from these kinds of comparisons.

I think that Universal has demonstrated repeatedly that VC-1 is capable of producing outstanding results. But I assume you would agree that using VC-1 doesn't guarantee outstanding results.
Actually, Universal does not do any encoding themselves. They use a post house that others also utilize. But I do agree with your point on VC-1. Feed us VHS, and we can only make it look like VHS :).

It seemed to me that the comparison of Batman Begins versus The Prestige might give some insight into the question of whether the softness in BB is attributable to softness in the original material or is instead an effect introduced during the transfer process.
It really doesn't. Per above, there is zero information to be gained here. Thousands of people work on a movie. Budgets vary. Equipment varies. Experience changes. Intent evolves. I know you all looking for some way of getting relative comparisons done but "this dog don't hunt. :)" If you don't trust me, as Paid to see what he thinks. I am confident he would agree.

So is it possible that you might address my original question?

Thanks
Hopefully we can put this to rest and move on to other things...

Is there an inherent bandwidth issue with HD DVD. I ask this because I've yet to see a movie with True HD and a high VC-1 encode.

Hi,

On another subject, 24p... Following Toshiba postponing 24p on XE1/XA2 to july; Can insiders told us what is the status of 24p regarding each format ?

Is it well defined in red/white/yellow books ? up to the player implementation ?

I've read subtitles (still being pictures, unfortunately) can be a problem on HD-DVD, then how BD implement it ? (it does works afaik).

--Patrice

There are two types of non-volatile memory ("local storage") defined in the Blu-ray spec - the Application Data Area (known as "Persistent Storage" in MHP) for preferences, bookmarks, and such, and the Binding Unit Data Area for storage of A/V clips. All Blu-ray players must support the Application Data Area, albeit not a huge amount. All BD-Video 1.1 and BD-Live players must also support the Binding Unit Data Area (256MB for BD-Video and 1GB for BD-Live).So is this why SMP8634 Blu-ray player development kit (http://www.sigmadesigns.com/public/Products/SMP8630/pdf_files/bluray8634_br.pdf) has two flash modules with one being 1 GB and the other being 64 MB? Also does Blu-ray require that the flash modules for local storage and persistent storage be kept separate or was this done just for ease of design?

So is this why Blu-ray player development kit has two flash modules with one being 1 GB and the other being 64 MB?Ease of design. The 64MB flash is used for persistent storage and also to store the encrypted player firmware. The 1GB flash module is for local storage and can be easily changed for more or less depending on the manufacturer's specific design.

Is there an inherent bandwidth issue with HD DVD. I ask this because I've yet to see a movie with True HD and a high VC-1 encode.
"Phantom of the Opera", which was a launch title?

There haven't been any True HD titles with insufficent bitrate I'm aware of.

Keith, just curious but have any of the manufacturer's specific designs included both Dolby TrueHD and DTS-HD MA decoding?

Question about 2.35 subtitles and menus to all insiders -

Many pages ago there was discussion on how to accommodate users with constant height systems by adding some code (HDi and BD-Java??)that would automatically allow the menus and subtitles to fit in the 2.35 frame vs the letterbox or black bar portion of the 1.78 frame.

Have any studios expressed interest in doing this or should we just give up hope on this front? If there was interest are there any US titles that are released or soon to be released that will have this feature?

Thanks for your time,
Jeff

We are looking into this... don't give up hope.

Is there an inherent bandwidth issue with HD DVD. I ask this because I've yet to see a movie with True HD and a high VC-1 encode.

Phantom of the Opera, Batman Begins, Troy, Harry Potter and the Goblet of Fire all had TrueHD and a great PQ, IMO. HPGoF and Batman Begins had that AND IME.

Or am I misunderstanding the question? Did you mean a high video bitrate or were you referring to PQ?



edit: And I want to add that Troy had IME as well.

The question of how much it cost to replicate BD and HD DVD is still a subject of speculation on this board. And while this was understandable prior to the launch of BD, I can't see any reason why some of the insiders haven't gotten the real cost and price by now. Surely several of you know people who are telling you what people are being charged, and estimates for the actual cost to master and produce SL and DL HD DVD and BD disc.

By now, someone should know where we are at with yields on BD50. Given that Sony is about to produce 500K BD50's for Europe, one might think they have that process somewhat under control now.

So the general question I have is: To what extent has the BD solved it's replication and mastering cost issues?

More specifically, if someone can, What do we know about the price and cost of BD SL and DL compared to HD DVD SL and SL? Do we knowing anything about yields?


Thanks in advance.

I have not yet watched it. I was sick and now in Japan and way behind watching movies :). But even if I did, I wouldn't be able to say anything useful here. I already mentioned that even within one movie, Batman Begins, there are variations in apparent sharpness. So what can you expect, comparing different movies? Really, there is nothing to be learned from these kinds of comparisons.


Actually, Universal does not do any encoding themselves. They use a post house that others also utilize. But I do agree with your point on VC-1. Feed us VHS, and we can only make it look like VHS :).


It really doesn't. Per above, there is zero information to be gained here. Thousands of people work on a movie. Budgets vary. Equipment varies. Experience changes. Intent evolves. I know you all looking for some way of getting relative comparisons done but "this dog don't hunt. :)" If you don't trust me, as Paid to see what he thinks. I am confident he would agree.


Hopefully we can put this to rest and move on to other things...

Amir, thank you for taking the time to respond to my question.

At your suggestion, I will ask if Paidgeek has any comments on my question regarding whether the difference in results in terms of PQ as between Batman Begins and The Prestige are likely to be attributable solely to differences in the source material. Paidgeek, any comments on this?

Amir, I am not sure what is the "this" that you think should be put to rest? It seems to me that the general question of whether differences in PQ are attributable to differences in the source material or instead are attributable to differences introduced in the transfer process is not a question that deserves to be put to rest so easily.

Amir, thank you for taking the time to respond to my question.

At your suggestion, I will ask if Paidgeek has any comments on my question regarding whether the difference in results in terms of PQ as between Batman Begins and The Prestige are likely to be attributable solely to differences in the source material. Paidgeek, any comments on this?




Film making has changed quite a lot in the last 10 years by incorporating electronic cameras and production tools into the process. For instance, if an actor or director is not happy with the way their character looks in a particular scene, they might use electronic methods in post to say, remove a few wrinkles... This happens more than you might think as the capability to make adjustments once you are in a DI color correction suite may be too tempting to pass up.

It is really not possible for a consumer of BD or HD-DVD to determine if a picture looks soft or lacks grain as a result of compression. The compressionist can make direct comparisons with the master, but naturally consumers don't have that available to them. There are artifacts that can be attributed to compression such as blocking, but even then you have to be careful because on certain types of content, the director may elect to show a scene shot in a consumer HD format as a way of expressing the immediacy of a handheld camcorder. Normally a situation like this is too obvious to misinterpret.

As I mentioned in a previous post, comparing the same version of a movie on two different formats that used different codecs is fair game as these should be sourced from the same master and any observed differences should be attributed to the results of compression.

Film making has changed quite a lot in the last 10 years by incorporating electronic cameras and production tools into the process. For instance, if an actor or director is not happy with the way their character looks in a particular scene, they might use electronic methods in post to say, remove a few wrinkles... This happens more than you might think as the capability to make adjustments once you are in a DI color correction suite may be too tempting to pass up.

It is really not possible for a consumer of BD or HD-DVD to determine if a picture looks soft or lacks grain as a result of compression. The compressionist can make direct comparisons with the master, but naturally consumers don't have that available to them. There are artifacts that can be attributed to compression such as blocking, but even then you have to be careful because on certain types of content, the director may elect to show a scene shot in a consumer HD format as a way of expressing the immediacy of a handheld camcorder. Normally a situation like this is too obvious to misinterpret.

As I mentioned in a previous post, comparing the same version of a movie on two different formats that used different codecs is fair game as these should be sourced from the same master and any observed differences should be attributed to the results of compression.

Thanks for the response, Paidgeek. So if it is correct that Warner is planning to put out The Prestige in a European version using VC-1, it would be fair to compare that to the BVHE AVC version?

I know my costs...and I have a good idea of what others pay. However it would not be appropriate to post specific costs in a public forum as that is confidential information.

What is your purpose for wanting this information by the way? I can safely say it costs about $1.00 more per disc than a DVD Disc to produce the HD Formats....in addition to much higher mastering and authoring costs.

Why do you wnat to know?

Rich



The question of how much it cost to replicate BD and HD DVD is still a subject of speculation on this board. And while this was understandable prior to the launch of BD, I can't see any reason why some of the insiders haven't gotten the real cost and price by now. Surely several of you know people who are telling you what people are being charged, and estimates for the actual cost to master and produce SL and DL HD DVD and BD disc.

By now, someone should know where we are at with yields on BD50. Given that Sony is about to produce 500K BD50's for Europe, one might think they have that process somewhat under control now.

So the general question I have is: To what extent has the BD solved it's replication and mastering cost issues?

More specifically, if someone can, What do we know about the price and cost of BD SL and DL compared to HD DVD SL and SL? Do we knowing anything about yields?


Thanks in advance.

Yes, there is a bandwidth spec and limitation for both formats.

HD-DVD has less bandwidth for Audio/Video encodes than BD. However, as encoders get more efficient over time, there may be more HD-DVD discs with DTS Master Audio Lossless and Dolby True HD, which is really just MLP Audio.

It is a fine line between audio/video and where to compromise when it comes to producing HD Titles. Most people will more readily notice the impact on Video before they will the Audio when it comes to choosing which one takes the "squeeze" of lossy compression so to speak.

Rich

Is there an inherent bandwidth issue with HD DVD. I ask this because I've yet to see a movie with True HD and a high VC-1 encode.

Film making has changed quite a lot in the last 10 years by incorporating electronic cameras and production tools into the process. For instance, if an actor or director is not happy with the way their character looks in a particular scene, they might use electronic methods in post to say, remove a few wrinkles... This happens more than you might think as the capability to make adjustments once you are in a DI color correction suite may be too tempting to pass up.

It is really not possible for a consumer of BD or HD-DVD to determine if a picture looks soft or lacks grain as a result of compression. The compressionist can make direct comparisons with the master, but naturally consumers don't have that available to them. There are artifacts that can be attributed to compression such as blocking, but even then you have to be careful because on certain types of content, the director may elect to show a scene shot in a consumer HD format as a way of expressing the immediacy of a handheld camcorder. Normally a situation like this is too obvious to misinterpret.

As I mentioned in a previous post, comparing the same version of a movie on two different formats that used different codecs is fair game as these should be sourced from the same master and any observed differences should be attributed to the results of compression.
Thank you for the informative post, it's appreciated.

I've one question: What happens if there's a film with a very soft master (let me mention HOFD as a prime example). Does the soft master always have to be a consequence of the original 35mm film recording? Or is it possible, that the DI manipulation steps are responsible for the softness? Or in other words: Is it totally impossible for such a movie to get some more details/sharpness out of the master without resorting to digital algorithms like sharpening/detail enhancement? E.g. could the DI be redone? Or could the original film be scanned again with better equipment? Or is all hope lost? I'm wondering because very old movies such as Casablanca and Robin Hood look quite fine after restauration. So I'm wondering whether such a restauration would be possible for newer movies with soft masters (like HODF) without negative side effects, too?

Yes, there is a bandwidth spec and limitation for both formats.

HD-DVD has less bandwidth for Audio/Video encodes than BD. However, as encoders get more efficient over time, there may be more HD-DVD discs with DTS Master Audio Lossless and Dolby True HD, which is really just MLP Audio.

It is a fine line between audio/video and where to compromise when it comes to producing HD Titles. Most people will more readily notice the impact on Video before they will the Audio when it comes to choosing which one takes the "squeeze" of lossy compression so to speak.

Rich

But how much does HD-DVD need lossless (especially when it's only 16/48) if 24/48 1.5mbps Dolby Digital Plus is virtually transparent? This seems to be a question that a lot of people don't want to address, as
1. BD supporters don't want to bring it up because BD can't do 1.5mbps Dolby Digital Plus.
2. Certain people want to keep up the illusion that DTS Master Audio and True HD are something spectacularly better, even at 16/48.
3. Certain people want to keep up the illusion that PCM is better than everything else, even when it is only 16/48.

Why do you wnat to know?

Rich

Because we analyze everything to the Nth degree around here :)

One of the advantage of HD DVD was lower cost of replication. Obviously, that cost is not always passed on to smaller studios, because your price may be dictated by supply and demand rather than cost to replicate - until you get high volume. But there was some evidence that a BD disc cost about 30% more to replicate than an HD DVD.

In another thread you said that the glass master cost the same for each format - about $3k. You are the first person I know that has said that. So apprearently Bd has brought down the cost of creating the glass master. There were some advances we had talked about before, and maybe those are coming onto the market now.

But I haven't heard anything about replication cost around here in a long time. That could be because they have made advancements in yields, and there is no longer a big difference. Or maybe people just got tired of talking about it. Still, i'm particularly interested in the experience of small studios in replication.

Without getting into specifics, can you give me a percentage of how much more BD cost to replicate? Is it 0%? 30%? These are the kind of things we like to talk about around here.

And thank you in advance. I really appreciate the insiders on this board giving us this kind of insight.

Hello,
I have two questions for paidgeek, you may or may not know the answers
Firstly at what point does a film become a catalgue title in Sony's eyes is it when it finishes it's run in the cinema or does a certain amount of time pass.

And secondly, Is there anyway, or any section of Sony Pictures that we in other regions can contact to find out if a Region A disc is region ABC or just A.
I ask because there are a few titles coming up that I would love to get but no Blu-ray release is scheduled yet, or due to international disctribution won't show up. Below are a few of the titles from Sony that I would like to get,

Stranger Than Fiction
Layer Cake
The Holiday (2006)
Rocky Balboa
Identity
Volver
Secret Window

Thanks for the response, Paidgeek. So if it is correct that Warner is planning to put out The Prestige in a European version using VC-1, it would be fair to compare that to the BVHE AVC version?

Yes, that could provide some interesting insight on how the the two codecs (and compressionists) perform on these titles.

Thank you for the informative post, it's appreciated.

I've one question: What happens if there's a film with a very soft master (let me mention HOFD as a prime example). Does the soft master always have to be a consequence of the original 35mm film recording? Or is it possible, that the DI manipulation steps are responsible for the softness? Or in other words: Is it totally impossible for such a movie to get some more details/sharpness out of the master without resorting to digital algorithms like sharpening/detail enhancement? E.g. could the DI be redone? Or could the original film be scanned again with better equipment? Or is all hope lost? I'm wondering because very old movies such as Casablanca and Robin Hood look quite fine after restauration. So I'm wondering whether such a restauration would be possible for newer movies with soft masters (like HODF) without negative side effects, too?

I don't see the DI process causing the master to be softer. It is possible to make to make the image softer by over use of grain reduction tools, but this is rarely a problem. At the risk of disappointing you, I don't think you will ever see a version of HOFD that is dramatically sharper than the one you have. These days, the film transfer is only one step in actually finishing the film. Special effects, wire removal, color correction and so on are all done in the digital domain. It is possible that the tools used to perform these functions will get better over time, but I think it will be a rare case to get an editor and other creative people to redo a title in order to try to squeeze out another 5% improvement, that is unless it is a classic...

You raise a good point about some of the older titles. Films were shot, edited and transferred using more basic techniques in the past. I think the saying that sometimes less is more holds true here because I have seen masters for films 10 - 40 years old that can stand up to or even exceed the quality of new films being released for the first time.

Hello,
I have two questions for paidgeek, you may or may not know the answers
Firstly at what point does a film become a catalgue title in Sony's eyes is it when it finishes it's run in the cinema or does a certain amount of time pass.

And secondly, Is there anyway, or any section of Sony Pictures that we in other regions can contact to find out if a Region A disc is region ABC or just A.
I ask because there are a few titles coming up that I would love to get but no Blu-ray release is scheduled yet, or due to international disctribution won't show up. Below are a few of the titles from Sony that I would like to get,

Stranger Than Fiction
Layer Cake
The Holiday (2006)
Rocky Balboa
Identity
Volver
Secret Window

As a rule, if a title is still in theatrical release anywhere in the world, it is not considered catalog. This can take up to a year.

The labelling of the discs to correctly identify them as region ABC should be sorted out by now, so you should be able to get this information straight off the packaging.

I know my costs...and I have a good idea of what others pay. However it would not be appropriate to post specific costs in a public forum as that is confidential information.

What is your purpose for wanting this information by the way? I can safely say it costs about $1.00 more per disc than a DVD Disc to produce the HD Formats....in addition to much higher mastering and authoring costs.


Would you say there was a difference in cost for you to produce Chronos on Blu-ray versus HD DVD? Common perception on these forums says it's more expensive for BD but that Sony is subsidizing...

Do you have any thoughts to ofter on that? Being a smaller studio, I'd find your perspective on the whole thing to be very illuminating for any studio in a similar position.

Yes, that could provide some interesting insight on how the the two codecs (and compressionists) perform on these titles.
I think you are saying this but to be clear, even that doesn't do the job (compare "two codecs"). We are not talking about codecs that auto-encode the movies and as such, the only difference would be the technology. This is a manual process and with different operators, equipment, budget for encoding, and skill of the operator, you get different results. Worse yet, one scene may look better in one, and another in the alternative version.

And of course, the HD DVD version will use a different video data rate and may accommodate other a/v tracks, making the comparison more difficult yet, if you want to figure out if the BD version could have benefited from use of VC-1 (which is I think the topic people have angst over). So at objective level, this is not a codec shoot out (although subjectively, may not matter much).

Anyway, I may be mixing up my titles but from what I recall hearing, The Prestige looks very good in VC-1/HD DVD. If so, this whole discussion will be academic for the world at large...

As an aside, he should also rename the title of his thread. The MPEG-2 encode is probably running at 2X the data rate of VC-1. Remember, the HD DVD version actually has two video streams with a Picture-in-Picture on top of the move itself – BD version in MPEG-2 lacks that. So if that scene is the pinnacle of what is wrong with the VC-1 encode running at half the data rate of MPEG-2, while at the same time providing a secondary video stream, I still stand proud even if this were an issue with VC-1

does this suggest that the bandwidth limitations of HD DVD force inherent compromises when 1080p video, PIP secondary streaming, and high-quality audio reproduction are all desired by a disc producer?

So at long last, here are some of the key reasons that VC-1 differentiates itself from AVC in preserving more of the natural texture/resolution and grain of the original HD content. Everything I state here is factual and can be confirmed by looking at the specs and history of these technologies. And apologies in advance for any typos, grammar errors. I am still pretty jetlagged in Japan.

As the discussion of codec technology at mathematical level quickly runs into “Star Trek language” (my term for fancy stuff that make no sense to someone not schooled in them), I am going to provide some high level context and historical perspective for these design choices first. Hopefully that makes it easier to digest the meat of the post, which follows.

MPEG-4 AVC’s origins come from work done by ITU (international standards setting organization) to create an alternative to then leading compression technologies on the internet: namely, Real network’s video codec and Microsoft’s. At the time, MPEG-4 ASP (or Part 2 as it is called) had failed to gain any traction due to not being competitive from performance point of view with these other codec. And MPEG-2 lacked the efficiency to be useful at internet rates.

The ITU initiative, called H.264 (coming after an earlier standard called H.263) was led by Dr. Gary Sullivan who happens to work on my team. Gary is world renowned in the compression/standards circuits and was honored for his work on H.264/AVC with IEEE fellowship award. He has the temperament and skills to drive these things like few have.

Unlike previous standardization efforts, the ITU put computational complexity at very low priority, instead focusing on best compression efficiency as a top goal almost regardless of implementation cost. This allowed them to take on algorithms which were very compute intensive, but generated compression efficiency gains nevertheless (the so called “CABAC” is an excellent example where it might take 20% of a chip just to perform this one function).

H.264 gained the interest and contributions from some of the top compression experts worldwide. As such, the codec started to show significant gains over then best open standard, MPEG-4 Part 2. Not to be upstaged by ITU, MPEG group proposed that H.264 become a joint project between the two standards group and hence the moniker “JVT” that you sometime see (“J” standards Joint). Later the standard was renamed, confusingly so, to MEPG-4 AVC (or Part 10) giving more power to the standard (see below).

The joint initiative got a lot of interest around the world. The aging patent portfolio around MPEG-2 motivated some of the largest companies to contribute to AVC, in the hopes of replenishing their revenue stream one day with patents in that format, when MPEG-2 royalty stream starts to run dry. Alas, that proved difficult as a form of gold rush occurred here, with some 150+ organizations contributing to the final standard (almost 10 times more than did with MPEG-2!). I remember Gary telling me how shocked he was as to the number of attendees in some meetings.

We did our share too and evidenced by Microsoft being one of the members of the MPEG-4 AVC patent pool (we were not part of MPEG-2 but do hold a similar position in MPEG-4 Part 2). And like the other patent holders, we stand to make a few pennies out of AVC one day. Given our involvement both in driving the standard and contributions to the algorithm and business terms/licensing, I hope people realize that my discussions around AVC is not the typical “bashing of your competitor” as we take some ownership and pride in development of AVC (although obviously not nearly as much as VC-1).

Given that 99% of video on the internet at the time of AVC development was CIF resolution (quarter screen/SD) and lower, strong emphasis was put to do well there. Test clips where used to evaluate effective performance various proposals with almost all at CIF resolution. This masked any issues that might have existed in higher resolution material with these algorithms.

JVT work finished around 2003, generating a truly advanced codec, rivaling anything we or Real Networks had. At SD resolution and lower, AVC is a formidable competitor to VC-1 and whether one wins or not in any comparison, is a very subjective thing. More so in its favor, AVC has the power of MPEG brand (with government mandate in some countries), and lack of direct association with Microsoft, which in many situations puts it ahead of VC-1 before the game even starts. This combination has resulted in many strong design wins for AVC in a number of applications from broadcast to satellite, IPTV, and of course, HD DVD/BD.

Computationally, AVC was about 3X slower than MPEG-2 to decode (and many times slower in encode). So making this a lower priority goal, did indeed take its toll on difficulty of implementation.

Of course, we were not sitting still during the development of AVC. We were hard at work, designing our next generation video codec, while being fully aware of work being done in ITU/MPEG.

Like AVC, we wanted to create a “standard.” And by that, I mean that we wanted to license it to others in the industry to implement in their hardware which meant that once we did that, we could not change it again without breaking compatibility. This meant that the algorithm had to last and stay competitive for a long time (5+ years). The final implementation, known as Windows Media Video 9 (or WMV-9) pushed way beyond our previous revisions of our video work, producing significant compression efficiency. Using MPEG’s own test clips, we were able to show 3X gain in objective PSNR measurements as compared to MPEG-2 and 2X compared to MPEG-4 Part 2 (at internet data rates).

The WMV-9 codec later became known as VC-1, when we opened its specifications, and submitted it to SMPTE organization for standardization. This was a requirement of it being adopted by DVD Forum.

Unlike AVC, we did not want to put computational efficiency at the bottom of the goal list. We wanted to run in millions of portable devices and there, battery power becomes directly proportional to how many MIPS you use. So even if we didn’t care how much the hardware cost, we still wanted to be more efficient.

During the development of VC-1, a sequence of events led to changes to our design which literally put us on the map when it comes to HD encoding. While our mainstream business was internet video with SD resolution and lower, we started to do some prototype encodings at HD resolution to see how well the codec performed there.

As we expected, the “coding gain” (compression efficiency) was quite significant at HD resolutions just the same. Many of the techniques that improve the “quality per bit” are resolution independent, letting us produce similar quality to MPEG-2 but doing so at much lower data rate. Excited, we showed a sample of our HD VC-1 encode to one of the (technical) studio executives. That led to us participating in DVD Forum HD DVD codec shootout which per my earlier note, resulted in us defeating the other technologies and becoming a mandatory standard in HD DVD and later, BD.

The reason we did so well was the result of another unintended development earlier. Since there were no real applications at the time for HD on the internet, we used our new capabilities simply to showcase what the codec could do. You know, kind of how Honda does Formula 1 racing and uses that as bragging rights to sell you an Accord :). But someone thought we really wanted to sell them a race car, and next thing we know there was a company who was packaging a PC, a dark chip DLP projector, and chasing art houses and smaller cinemas to switch to digital using this low cost system! The package worked quite well, solving the major cost issue of higher end systems being developed in Hollywood and art houses started to deploy it rapidly (they could run digital advertisements on it to make money).

So we started to encode a bunch of independent movies and going through the evaluations with the creative community. Unfortunately, we quickly realized that some of the lessons we had learned on the internet were backfiring on us. In subjective test after test for internet applications, we had seen that people would sacrifice resolution to get a softer but artifact-free video (well, as artifact free as one can make internet video :)). When it came to digital cinema though, one couldn’t soften the picture when the thing is being blown up on such a big screen. And people didn’t want to be told to pick between softness or artifacts. They wanted to have their cake and eat it too with transparent resolution to the source and no artifacts.

Our early tests showed up that indeed, we were sacrificing some grain/textures with our algorithm as designed. Worried that if we optimized the codec differently for HD, we would lose the battle on the internet, we made the algorithm adaptive with respect to resolution/data rate. Namely, if you feed VC-1 content above SD resolution, it will behave differently, even though its core algorithm is the same across the board.

The above changes proved very effective and nicely improved our HD performance. We shortly became one of the standard submission formats for Sundance film festival (the only electronic format allowed). I also got to meet Robert Redford and learned that this job has some nice perks :). But I had to pay for it dearly, sitting at midnight (the only time we could time on the Sundance theaters during the festival) trying to prove to them that we had not lost the “film look.” After all, these guys don’t call themselves “film makers” for nothing!

So they demanded to have a shootout against film projection before giving us the green light to use the VC-1 encodes. And of course, they were always finishing their movies a couple of days before the show, given us little time to encode them and hence the midnight viewings.

Fortunately, in every case, the producer/director agreed that our VC-1 encode outperformed the equiv. film projection, while sitting in the front row of the theater and watching a 30 foot screen. And we went on to watch some really great, and some really not-so-great independent films on our technology. :)

During one of the Sundance festivals, we also met with Lionsgate folks who were thinking of releasing T2. Impressed with what we had done with VC-1 at Sundance, they agreed to do a dual disc release creating what later became “WMV-HD” format (VC-1 content but using our audio and file format on red laser DVDs). That started a trend and soon, we had some 50+ titles using the same technology, pushing us nicely to keep improving our codec for HD movie encoding. You really get good as this stuff, when you try to do 1080p in 10 mbit/sec maximum rate for both audio/video and video files the same size as regular SD DVDs in MPEG-2! WMV-HD also got my start at AVS Forum, answering questions about WMV-HD. So if you are unhappy about my postings here, you have Lionsgate to blame :D.

During this journey, Joe Kane, frustrated with the poor quality of D-VHS MPEG-2 on his HD DVE test patterns, got interested in releasing a WMV-HD disc. So we worked with him and learned a few things to fine tune there. His final VC-1 encode was below 12 mbit/sec (constant bit rate no less), yet outperformed his D-VHS equivalent showing that even on pathological test sequences (i.e. stuff that is hard but never occurs in real life), we had a significant efficiency and quality gain over MPEG-2. So Joe became a believer and continues to advocate VC-1 in his demonstrations to this date (going as far as using HD DVD/VC-1 in the Samsung booth at IFA last year instead of BD/MEPG-2!).

Computationally, we also met our goals in that we achieved great compression efficiency, yet our decoder would only take 2X more MIPS than MPEG-2 (compared to 3X on AVC). Encoding was a lot slower than MPEG-2 at higher quality settings, but way better than AVC. We also use less memory for both encode and decode than AVC.

OK, enough with history. Let’s get into the details.

Both AVC and VC-1, like MPEG-2 before it and many other video codecs, are transform based. The screen is divided into blocks which are then independently compressed (think JPEG) and their motion is tracked on screen (so that we don’t have to retransmit the whole block when some part of the image moves but other parts don’t). We call this “motion estimation.”

There is a lot more sophistication however in both AVC and VC-1 as compared to MPEG-2’s much simpler algorithm (which was heavily constrained by what hardware could do a decade back). We have such features as adaptive (as opposed to fixed) block size, quarter pixel motion estimation (as opposed to half pixel), and more efficient “entropy coding” (lossless component of all video codecs to squeeze the final set of bits as tightly as possible). And much more.

While both AVC and VC-1 are “advanced codecs,” they differ significantly in how they gain their efficiency over MPEG-2:

1. Transform block size differences. MPEG-2 uses fixed 8x8 blocks. AVC changed this to 4x4 blocks (with the transform designed by Microsoft btw :)). VC-1 on the other hand, supports 8x8, 4x4, 4x8 and 8x4. This gives VC-1 much more flexibility to pick the optimal block size based on picture content. For example, the outline of a person may be better coded using vertical blocks that suck in less of the background grain (letting the block be optimized for the subject or background gain and not both at the same time). And larger blocks can preserve texture better.

After losing the picture quality test to VC-1 in DVD Forum (and in some cases, finishing even behind MPEG-2), reality set in with AVC folks just as it had for us sometime back. Surprisingly, they chose to modify the standard and add adaptive block size of VC-1 to AVC, post standardization. I say surprisingly because this addition broke compatibility with the just finished standard which is a rare situation.

The paper proposal from Sand Video started it all and is an interesting read in this respect: http://ftp3.itu.ch/av-arch/jvt-site/2003_05_Geneva/JVT-H029r1.doc:

“The faithful reproduction of fine detail, including film grain, is required in high definition (HD) broadcasts, HD-DVD, and Digital Cinema. To meet this requirement, more high frequency information must survive quantization than is typical in lower bit rate applications. This contribution demonstrates that the frequency selectivity and reduction in boundary effects realized by using adaptive block transforms (ABT) helps meet the high standards of the HD community. We present average RD performance improvements of 9.75% on HD film sequences and significant perceptual gains in areas of fine detail and film grain.” (emphasis mine)

And then this:

“Test Sequences: Five HD-scans of major release Hollywood movies (designated Movies 1-5) (footnote 1).
(Footnote 1) Due to licensing arrangements with the Hollywood Studios, Sand Video cannot disclose the names of the movies nor show the results of the processing to non-DVD Forum members.” (emphasis mine)

Yes, you guessed it right. They are talking about the DVD Forum tests that I mentioned. They used the same test clips to show the improve quality using adaptive/larger block size which VC-1 used.

And who is “Sand Video”you ask? They were a start-up that was developing AVC encoders and decoders. The company was subsequently sold to Broadcom and their work became the foundation of the single chip AVC/MPEG-2/VC-1 decoders powering Blu-ray and HD DVD players. But at the time, they were a proponent of AVC and one of the companies participating in the tests, with an AVC encoder.

Since the AVC standard was already cast in concrete prior to this submission, a “High Profile” extension was created with the addition of 8x8 block size. The standard AVC spec was then called “Baseline” or “Main” profile. Despite being a late comer, HP extension was added to both HD DVD and BD specifications.

Note that per my previous note, the HP profile is incompatible with the Baseline Profile which many AVC systems use (i.e. crashes the decoder). So unless you see HP after MPEG-4 AVC, you don’t get adaptive block size. And even if you do see it, you have half the choices that VC-1 has. Per earlier mention, non-square blocks are quite handy in coding the picture, letting us to better avoid clustering non-similar pixels.

You may want to check out the pictures in the paper, showing the effectiveness of 8x8 blocks and its performance for preservation of grain.


2. Loop filter differences. “Loop” here refers to a feedback loop which is part of all modern video encoders. To figure out what to send next, the codec subtracts the decoded frame from the source (finding the “error signal”), and then compresses that difference and ships it to the decoder to keep refining the picture to bring it closer to the source (assuming the scene has not changed). Since we feed the output of the encoder back to the input, it creates a feedback “loop” in the classic engineering terminology.

AVC and VC-1 deviate from MPEG-2 in that they can insert a filter in the feedback loop. By taking into account the distortion created in the stream as a result of compression, and filtering it, one can gain significant compression efficiency. This is the key reason neither AVC nor VC-1 degrade as badly as MPEG-2 into a sea of blocks when starved for bits. Think of it as “soft clipping” for you all audiophiles :).

The existence of loop filter also means that increasing the bit rate may not necessarily improve perceived quality because the codec is able to some extent mask compression artifacts. The corollary of this is that the quality curve of these advanced codecs is more non-linear with a longer asymptote at higher rates, as compared to say, MPEG-2. They reach higher quality sooner, and increases in bitrate beyond some point do not gain you as much visually as it might in MPEG-2 (this is a good thing, not bad :)).

The strength of the loop filter in both codecs is dynamic and based on “heuristics” (educated guesses) by the encoder. They may also be tunable by the operator.

While both codecs sport loop filters, their filter characteristic varies substantially. The difference here is a direct result of our digital-cinema work where we found the loop filter to be very disruptive to film grain and texture. So we optimized the VC-1 filter by reducing its length to one pixel on each side of the block being encoded. Think of this as an eraser with a very sharp point, gently touching up those three pixels so that you can’t tell there is a line through them at the block boundary. Again, note that even this light touch is applied adaptively, i.e. only when needed and by the right amount.

Instead of just one, the AVC filter softens up to 3 pixels on either side of the block edge. Now think of a 4x4 block. If you go into the block two pixels from each side, you will be filtering the pixels twice (once from left, and then one more time from the right)! In contrast, the VC-1 filter only touches those pixels once since we only go one pixels to left or right. In the case of an 8x8 block, VC-1 never touches the center pixels whereas AVC filters them once.

So think of AVC loop filter as a giant eraser, three times bigger than that of VC-1, attempting to smooth the block boundaries.

If the scene gets too difficult to code and block boundaries become visible according to internal heuristics, AVC essentially filters every pixel on screen (and sometimes twice in case of 4x4 blocks), whereas VC-1 judiciously filters block boundaries only. Assuming same filter strength used in both codecs, VC-1 picture is bound to look sharper, with less “resolution pumping” (picture getting soft when it gets hard to encode and then sharper when not) than AVC.

The same video in MPEG-2 by the way, would show blocking artifacts as you see in live HD sports on TV. And since the efficiency of MPEG-2’s algorithm is lower in general as compared to AVC/VC-1 (e.g. in the entropy coding section of the codec), and its blocks always the larger 8x8, its blocking artifacts are more severe and easier to see. Some people may prefer this to the softer look of AVC though in some cases.

As you can imagine, if the content is easy to encode, the filtering does not kick in at full strength and the picture can look fine in AVC. But as the scene complexity increases, AVC becomes heavy handed, smearing detail. Give it easier (e.g. clean) material, or exceedingly high data rate, and it will do much better as its loop filter gets reduces in strength (hence my earlier comment that BD’s data rate “erases some of the AVC sins”).

Don’t even think of turning off the filter though, as some claim to get around this design limitation. Doing so, will take away the distortion mitigation system and you kind of wind up with MPEG-2 style artifacts (we know, we have tested it that way). Put another way, the AVC filter is more like a big on/off switch, without effective graduations, despite its adaptive nature. You are damned if you do, damned if you don’t with this loop filter for HD content.

Now you know why telling us this movie or that movie looks good in AVC, doesn’t impress us as much :). We know that some content will suffer less but in many general cases, you are going to run into this loop filter deficiency which works wonderfully for internet video, but not as well at the quality levels we are talking about. And that softening occurs in some scenes and requires knowing what segment is difficult and looking for its effect there.


3. Interpolation filter size. As I mentioned in the introduction, both AVC and VC-1 use quarter pixel resolution to track motion of blocks. This by definition means that you have to “interpolate” the intermediate values between pixels as you decide to move the same block in the decoder frame buffer within subpixel boundaries.

VC-1, roughly speaking, uses bicubic filtering with 4 taps. AVC on the other hand, uses a 6-tap filter. One would think that more of anything is better here but such is not the case when you are compressing HD video. By using more taps, you are using more of the pixels in the frame to find the interpolated one. This has the unintended effect of smearing adjacent pixels into each other with resulting loss of resolution (think of an object being averaged with black background with grain in it). And more taps in digital filters means more “ringing,” (some frequencies emphasizes more than others) causing edges to not be crisp. The shorter taps in the VC-1 filter is also one of the reasons we are much faster than AVC as we don’t need to touch as many pixels.

Are you still with me? If so, you are doing good and could do my job one day! :) I warned you this stuff gets complicated fast.

Hopefully the above gives you some insight into the differences between these two (very good) codecs and why there are some fundamental differences between them when it comes to HD encoding. While both algorithms can run circles around older technologies such as MPEG-2 (and contrary to popular myth, can be defeatured to behave just like MPEG-2), at the end of the day, one is a better fit for the application in mind.

VC-1 was optimized (and redesigned) before becoming final with real world HD applications and content. AVC was modified quickly, post standardization to perform better but not to the same level as VC-1. Smart people worked on both, but were given different design criteria so they came up with different solutions.

Is there a down side to the differences mentioned? Yes, a bit. VC-1 can become blocky sometimes because it attempts to both preserve detail while at the same time not resorting to excessive filtering of the source as AVC does. This impacts us the most at internet rates as compared to AVC (even there, we opt for sharper pictures). But we think for the formats in question, that can be dialed out by hand optimization and better automated analysis (the main focus of our R&D for the last year). And once there, you have a sharper picture for it. Not everyone agrees of course, especially at below SD resolution. But enough people do thankfully to have gotten us 200+ VC-1 titles in HD DVD/BD and great praise for many of the encodings…

does this suggest that the bandwidth limitations of HD DVD force inherent compromises when 1080p video, PIP secondary streaming, and high-quality audio reproduction are all desired by a disc producer?
In my opinion no. But to the extent the same people think BD should use VC-1 at higher rate, they can't turn around and claim that lower rate VC-1 should match or outperform MPEG-2/AVC at higher BD rate. Either accept that lower rate is just fine, or don't claim that we should beat the other guy at lower rate. The two stances together make no sense.

Amir,

Thank you for the very informative VC-1/AVC comparison. Can you recommend any books/articles on video compression technology that are not overly mathematical? I would like to learn more about this area.

Also, any dates yet on when the Mobile Experience will be in Atlanta?

The paper proposal from Sand Video started it all and is an interesting read in this respect: http://ftp3.itu.ch/av-arch/jvt-site/2003_05_Geneva/JVT-H029r1.doc:

“The faithful reproduction of fine detail, including film grain, is required in high definition (HD) broadcasts, HD-DVD, and Digital Cinema. To meet this requirement, more high frequency information must survive quantization than is typical in lower bit rate applications. This contribution demonstrates that the frequency selectivity and reduction in boundary effects realized by using adaptive block transforms (ABT) helps meet the high standards of the HD community. We present average RD performance improvements of 9.75% on HD film sequences and significant perceptual gains in areas of fine detail and film grain.” (emphasis mine)

...

And who is “Sand Video”you ask? They were a start-up that was developing AVC encoders and decoders. The company was subsequently sold to Broadcom and their work became the foundation of the single chip AVC/MPEG-2/VC-1 decoders powering Blu-ray and HD DVD players. But at the time, they were a proponent of AVC and one of the companies participating in the tests, with an AVC encoder.

Since the AVC standard was already cast in concrete prior to this submission, a “High Profile” extension was created with the addition of 8x8 block size. The standard AVC spec was then called “Baseline” or “Main” profile. Despite being a late comer, HP extension was added to both HD DVD and BD specifications.

Note that per my previous note, the HP profile is incompatible with the Baseline Profile which many AVC systems use (i.e. crashes the decoder). So unless you see HP after MPEG-4 AVC, you don’t get adaptive block size. And even if you do see it, you have half the choices that VC-1 has. Per earlier mention, non-square blocks are quite handy in coding the picture, letting us to better avoid clustering non-similar pixels.



Amir, did Tom McMahon not clairfy the ABT vs. 8x8 issue before?

http://archive2.avsforum.com/avs-vb/showthread.php?p=6594314&&#post6594314

"Finally, we didn't even make up an 8 bit 4:2:0 High Profile (which is the High Profile under discussion here) using the 8x8 transform until July of 2004. What we ended up over a year later was ***NOT*** ABT. It was only the addition of an 8x8 transform to the Standard (incidentally, variations on the 8x8 transform had been proposed by FastVDO and others well before ABT itself, in January of 2002 I believe)." - Tom

PS: I wish Tom well in his recovery.

Thanks for the explaination Amir.

Amir,

Thank you very, very much for the highly informative read. Like f300v10, I also would like some recommendations on books/publications to read more about the subject.

Thanks again :) .

Amir, did Tom McMahon not clairfy the ABT vs. 8x8 issue before?

http://archive2.avsforum.com/avs-vb/showthread.php?p=6594314&&#post6594314

"Finally, we didn't even make up an 8 bit 4:2:0 High Profile (which is the High Profile under discussion here) using the 8x8 transform until July of 2004. What we ended up over a year later was ***NOT*** ABT. It was only the addition of an 8x8 transform to the Standard (incidentally, variations on the 8x8 transform had been proposed by FastVDO and others well before ABT itself, in January of 2002 I believe)." - Tom
The key part of what Tom is saying is the same as what I mentioned. That is, 8x8 blocks were added in HP profile. He says this is not ABT but adaptive block size is just that, it lets the codec decide (adapt) which block size is better: 4x4 or 8x8. But if one assumes that Tom is saying the codec can only use one or the other, or not adapt, then that makes AVC weaker, not better :). Alas, my understanding is that AVC encoders in use do indeed sport an adaptive algorithm to pick the appropriate block sizes like VC-1.

The only thing I can think of is that Tom is saying the baseline profile, does not use ABT. That would be correct since the baseline only allows one block size: 4x4.

Tom does point out something I forgot to mention. That is, other block sizes were proposed for standardization in the baseline profile. But the AVC team thought they would not be advantageous to have. So they were dismissed -- only to be added later in HP, when evidence to the contrary became available.

PS: I wish Tom well in his recovery.
Me too :).

Amir,

Thank you very, very much for the highly informative read. Like f300v10, I also would like some recommendations on books/publications to read more about the subject.

Thanks again :) .
My pleasure.

Unfortunately, I don't know of any books that bring the topic level high up enough, for people to understand. Such books have to be accurate to the nth degree and as such, will get very detailed quickly. Certainly no book out there compares the technologies as I just did.

Not anymore. We changed our plans about a year ago, when we realized that there were a number of companies building HD DVD playback for PCs (and did so for XP, where we had no plans). So we decided to help them with compatibility suites and other assistance, as opposed to building our own player. The team did go on to build the Xbox 360 player however.

As to using the 360 drive, that is certainly possible now. The drive is WHQL certified and should work well with third-party players.
Hi Amirm, I have two more questions. First, according to this Engadget article (http://www.engadget.com/2006/12/30/never-mind-vista-heres-fiji-and-vienna/) from December 2006 HD DVD playback support would be added to "Fiji" or "Vienna". Would this be SP1 or SP2 for Vista? There are many who (http://www.avsforum.com/avs-vb/showthread.php?p=9932369&&#post9932369) really want to use (http://www.avsforum.com/avs-vb/showthread.php?p=9924214&&#post9924214) the Media Center interface w/ EVR to watch our HD DVD collection, which is one of the reasons why many of us opted to upgrade to Vista Home Premium or Ultimate to begin with.

Second, when you say that Microsoft decided to help third-party software makers with compatibility suites and other assistance, does this include integration into the Vista Media Center interface?

To all Studio insiders,

I think that quite a lot of people here would be interested to know how Hollywood views Digital Intermediate for film and HD, and how they're going to evolve -- both for upcoming movies and more accurate prevervation of catalog classics.

I remember that some years ago Sony DADC (and definitely other Studios) was on the forefront for doing HD telecine of almost every new films being made. And then they abruptly stopped doing that.

According to this story in AV Guide (http://www.avguide.com/the-perfect-vision/76/bluray.php) , there are more and more films which are mastered in HD through a DI. According to Don Eklund, The Pursuit of Happyness, All the King's Men, and Gridiron Gang were scanned at 2K, while Stealth and Spider-Man 2 were scanned at 4K.

This brings some questions:

- Were the initial hi-def DADC telecine drastically less advanced than today's DI? Have they been completely "retired", or this material can still be used today?

- What is the process that leads Studios to say that film A and B will be scanned at 4K, while film C and D will be scanned at 2K?

- And most important of all, how confident are the Studios that 4K material will still be satisfying in 10 or 15 years? How long are you thinking ahead?

I'd also like to ask the same questions in the perspective of catalog titles and film preservation. The more time it passes, the more we're losing forever ancient film negatives, and perhaps grade C or D movies which have never been released on DVD.

- Are we getting closer to a film digitalization which retains the exact same information than a 35mm?

- And what kind of deference treatment will the Studios give to "old" catalog titles when they plan to release them on HD? Paidgeek, if you're reading this, I'm talking about Lawrence and River Kwai, but also about the Brando, the Sidney Poitier, the Ray Harryhausen and so on.

Thank you!

Hi Amirm, I have two more questions. First, according to this Engadget article (http://www.engadget.com/2006/12/30/never-mind-vista-heres-fiji-and-vienna/) from December 2006 HD DVD playback support would be added to "Fiji" or "Vienna". Would this be SP1 or SP2 for Vista? There are many who (http://www.avsforum.com/avs-vb/showthread.php?p=9932369&&#post9932369) really want to use (http://www.avsforum.com/avs-vb/showthread.php?p=9924214&&#post9924214) the Media Center interface w/ EVR to watch our HD DVD collection, which is one of the reasons why many of us opted to upgrade to Vista Home Premium or Ultimate to begin with.
That article is speculative. We have not said anything about follow on updates to Vista.

I appreciate the fondness for the MCE UI and the desire to use it to watch HD DVD. Alas, as you have seen from recent breaches of the formats, writing such software, even on top of Vista's more robust foundation, is very challenging. Third-parties can use different standards and ship products which we simply cannot do.

Second, when you say that Microsoft decided to help third-party software makers with compatibility suites and other assistance, does this include integration into the Vista Media Center interface?
At disc insertion, yes. Beyond that, it is difficult to integrate into the menu system. What third-parties can do however, is mimic the MCE UI. We did this for WMV-HD discs btw. When you played these, it was actually a different player underneath, but with similar control to MCE full screen UI.

Yes, I know none of this is "pretty." :) We will look at integrating HD DVD at the right time into the system.

In another thread you said that the glass master cost the same for each format - about $3k. You are the first person I know that has said that. So apprearently Bd has brought down the cost of creating the glass master. There were some advances we had talked about before, and maybe those are coming onto the market now.
One needs to give credit when it is due. And here, Sony deserves some kudos. In order to make it possible to make the finer pitch masters for BD, they developed a new machine called "PTM." This is a phase change system and uses a similar system to semiconductor etching to create the pits. Result is a simplified process which costs less.

Fortunately, the same machine can be and is used (in at least in one replicator) for creation of HD DVD. So both formats benefitted the same here.

I have no idea if the above is the reason for any mastering cost reduction, as a lot goes into that including level of competition, the backlong for the plant, etc. But thought to put in the good work Sony did anyway.

BTW, this was the one thing I was impressed with when I first visited Sony BD factory....

Thanks Amir.

Have you happened to hear anything about yields on BD50 or other replication cost?

I understantd that people in the BD camp would be in a better position to discuss this, but if they pass on the opportunity to comment on it, I would appreciate your thoughts.

Thanks in advance.

Paidgeek -

Thank you for your answer. It amazes me that suggestions from this forum wind up in a final product!

Would anyone from the HD-DVD camp care to comment on this question as well?

Laters,
Jeff

Quote:
Originally Posted by eq_shadimar
Question about 2.35 subtitles and menus to all insiders -

Many pages ago there was discussion on how to accommodate users with constant height systems by adding some code (HDi and BD-Java??)that would automatically allow the menus and subtitles to fit in the 2.35 frame vs the letterbox or black bar portion of the 1.78 frame.

Have any studios expressed interest in doing this or should we just give up hope on this front? If there was interest are there any US titles that are released or soon to be released that will have this feature?

Thanks for your time,
Jeff


We are looking into this... don't give up hope.

__________________
Sony Pictures BD Insider

To all Studio insiders,

I think that quite a lot of people here would be interested to know how Hollywood views Digital Intermediate for film and HD, and how they're going to evolve -- both for upcoming movies and more accurate prevervation of catalog classics.

This brings some questions:



I'd also like to ask the same questions in the perspective of catalog titles and film preservation. The more time it passes, the more we're losing forever ancient film negatives, and perhaps grade C or D movies which have never been released on DVD.


Thank you!


- Were the initial hi-def DADC telecine drastically less advanced than today's DI? Have they been completely "retired", or this material can still be used today?
No I would not say the HD Center transfers were drastically less advanced. The capabilities have improved as a function of storage and other technical advances, but many of the transfers from the HD Center are acceptable or good by todays standards and are used to service clients.

- What is the process that leads Studios to say that film A and B will be scanned at 4K, while film C and D will be scanned at 2K?
It is a function of what the post facility can do and what the budget for the work is. 4K is still expensive to deal with because of its storage and bandwidth requirements, but I expect the industry will likely standardize on it in the coming years.

- And most important of all, how confident are the Studios that 4K material will still be satisfying in 10 or 15 years? How long are you thinking ahead?
There have been some spirited discussions about 4K in the pursuit of digital cinema standards. Your eyes can resolve about 1 minute of arc. If you do the math you will find that you need to be at less than 2 picture heights to begin to resolve a 4K image. If your question is whether or not 6K or 8K is important, you have to imagine a reason to view the screen from less than 1 picture height. For pure archival purposes, we can always argue for saving the highest possible resolution images, but if costs or practicality lead studios to take no immediate action to digitize older film libraries, then I say it is better to start scanning at 2K today, rather than wait another several years to start going through the entire library at 4K or higher.

- Are we getting closer to a film digitalization which retains the exact same information than a 35mm?
The mediums are different, so I don't think we can ever say we are capturing the "exact same information". The process of digitally sampling anything analog will lead to some finite level of quantizing error but consider this, If you view the picture from a large format negative (65mm) scanned at 1920 x 1080 and compare it to a 35mm negative scanned at the same resolution, you will see a dramatic difference in picture quality. From this simple observation it goes to follow that current HD has sufficient performance to adequately capture the image from a 35mm source without loss of observable picture resolution. This is valid for distribution formats (what the consumer will view) not production environments where higher sampling resolution allows for more accurate manipulation of data.


- And what kind of deference treatment will the Studios give to "old" catalog titles when they plan to release them on HD? Paidgeek, if you're reading this, I'm talking about Lawrence and River Kwai, but also about the Brando, the Sidney Poitier, the Ray Harryhausen and so on.
We have an industry leading department dedicated to the care and archives of our catalog titles, not just the ones that everyone is waiting for, but all of them. I think you will be very pleased to see the condition these titles are in when the are released to the market. Dirt and scratch removal technology has come along way since the days of an artist paint out dirt in Photoshop.

We have an industry leading department dedicated to the care and archives of our catalog titles, not just the ones that everyone is waiting for, but all of them. I think you will be very pleased to see the condition these titles are in when the are released to the market. Dirt and scratch removal technology has come along way since the days of an artist paint out dirt in Photoshop.

Paid (and other insiders),

Two questions:

1) What media and/or storage systems are used for the digital masters, and are they considered viable for long term storage?

2) How are animation movies stored, transfered, and encoded to HD? Such as Lion King, Toy Story, Ice Age, etc.

Amir,

Thanks for the detailed explanation on VC-1/AVC HP. Regarding the loopfilter, I do understand what you mean by AVC softening up the picture when encountering complex scenes. Two good examples of this would be XMEN 3 and Lucky Number Slevin. On close up shots, both discs using AVC HP do a great job with detail and sharpness but with distant grainy scenes with lots to cover, the picture becomes rather flat/soft. Now many will state "director's intention" but this output is specific to the aforementioned specific style of shots in both movies.

Also, please pass on my congrats to your team and the Warner post house that did "The Departed." To me, that is a perfect example of a HD movie. Great detail and sharpness throughout along with beautiful colors. Can't beat that!

I hear Panasonic now has a preference for doing the encoding for studios, rather than selling their AVC encoder to others. If so, that shows they are uncomfortable supporting users of their encoder. Or they think they can do a better job than others. So if are more confident in AVC being used by others, what’s up with this position by Panasonic?Perhaps that it's more profitable to operate a service than it is to sell a product, especially an expensive, niche product with a small potential market upon which to amortize your customer support costs? It's not like their AVC encoder is simply a software product with a potential market of thousands. It's a ton of hardware with a potential market of probably dozens or less. There's absolutely nothing in your characterization of Panasonic's business model selection which can possibly be reasonably interpreted to suggest a deficiency in AVC.

- Talk

Is AVC is more flexible than VC-1 in this respect (since VC-1 always use only 1 pixel) ?

Also *assuming* 4 tap is better than 6, why not go lower ? If it's not possible/not a good idea to go lower, what objective measurements can be made to say 4 is not too low ?

Since AVC uses more computational power, what other advantages does it bring to the table (compared to VC-1) ?

Good catch ! Does that mean the AVC is more flexible than VC-1 in this respect (since VC-1 always use only 1 pixel) ?
Unfortunately it does not mean that. AVC will either use 2 pixels on either side of a block edge or 3 pixels. It picks one or the other based on level of gradient. In no case does it go down to a single pixel as VC-1 does. So you wind up filtering a lot more pixels with AVC, regardless of which mode is selected. And filtering is not good when you want to preserve detail.

Also *assuming* 4 tap is better than 6, why not go lower ? If it's not possible/not a good idea to go lower, what objective measurements can be made to say 4 is not too low ?
It is a balance thing. The filter gets "less accurate" as the taps go down. So you wind up choosing between accuracy and softness/smearing/ringing. Objective measurements come from a lot of research with real material. Per my note, we got plenty of that with HD material in addition to huge amount of content on the internet using our codec (probably 1000X of what is in AVC).

Since AVC uses more computational power, what other advantages does it bring to the table (compared to VC-1) ?
There you go assuming if something is expensive, it must be better :). It is not always the case. I already gave one example, namely the higher order filter. The other example is CABAC or “arithmetic coding.” This is a technique for squeezing the bits in the entropy coder (the lossless part). We use multiple “hufman tables” each optimized for different data rates and resolution. CABAC instead, uses a floating point computational method to achieve the same. Because of its complexity, it requires a lot of MIPS to decode. In some MPEG test clips, CABAC can yield up to 10% compression efficiency if my memory is right. In practice though, we feel that the real advantage is much lower making it a tough pill to swallow given its complexity.

Another difference is multiple reference frames. In theory, this gives the encoder more flexibility in that it could compute the motion vectors based on more than one frame. The downsides are twofold: 1) you have to have more memory in the decoder (and encoder) as you have to hang on to the older frames to reference from. And 2) in “short-GOP” formats such as HD DVD/BD and IPTV scenarios where you want to have quick channel switching, you wind up limiting the number of frames before restarting (i.e. sending an I-frame), so multiple reference frames loses its usefulness.

Hopefully this gives you some idea of where the complexity goes. If you want more than this, maybe one of the advocates of AVC can provide it. It is not easy to explain this stuff :).

I think that Universal has demonstrated repeatedly that VC-1 is capable of producing outstanding results. But I assume you would agree that using VC-1 doesn't guarantee outstanding results.







Actually, Universal does not do any encoding themselves. They use a post house that others also utilize. But I do agree with your point on VC-1. Feed us VHS, and we can only make it look like VHS :).



Yes, that could provide some interesting insight on how the the two codecs (and compressionists) perform on these titles.




I think you are saying this but to be clear, even that doesn't do the job (compare "two codecs"). We are not talking about codecs that auto-encode the movies and as such, the only difference would be the technology. This is a manual process and with different operators, equipment, budget for encoding, and skill of the operator, you get different results. Worse yet, one scene may look better in one, and another in the alternative version.




Amir, thanks for the very detailed comparative history of VC-1 and AVC. That is the sort of information that makes this thread worthwhile.

To follow up on your earlier response to my earlier question, and then your comment on paidgeek's response on one of the same points, you suggested in your response to me that the only time when VC-1 would produce less than outstanding results would be if the source material was inferior. However, in your response to paidgeek, you seem to be acknowledging that there are many other factors that affect the quality of the results. I assume that your response to paidgeek on this point represents a more complete presentation of your views on this subject? That is, the use of VC-1 does not guarantee outstanding results even if the source material is outstanding, in light of the many other factors that affect the result?

To follow up on your earlier response to my earlier question, and then your comment on paidgeek's response on one of the same points, you suggested in your response to me that the only time when VC-1 would produce less than outstanding results would be if the source material was inferior.
I don't recall saying "it was the only time." I said if you feed us bad content, we produce bad content. That is not the same as what you are saying.

However, in your response to paidgeek, you seem to be acknowledging that there are many other factors that affect the quality of the results. I assume that your response to paidgeek on this point represents a more complete presentation of your views on this subject? That is, the use of VC-1 does not guarantee outstanding results even if the source material is outstanding, in light of the many other factors that affect the result?
Of course that is the case. Per my earlier posts, these are professional tools, requiring training, and proper usage and experience. If these are lacking, you get less than perfect results. But note that the same is true of other codecs. So this distinction by itself, is not very meaningful in the context of comparing codecs.

Unfortunately it does not mean that. AVC will either use 2 pixels on either side of a block edge or 3 pixels. It picks one or the other based on level of gradient. In no case does it go down to a single pixel as VC-1 does.
That's incorrect. H.264 has the option of only filtering a single pixel on each block edge. Just read the specification (I've used bold red text to show the case where only one pixel on each side of the block is filtered):

************************************************************ ****
8.7.2.3 Filtering process for edges with bS less than 4

Inputs to this process are the input sample values pi and qi (i = 0..2) of a single set of samples across an edge that is to be filtered, chromaEdgeFlag, bS, β, and indexA, for the set of input samples, as specified in 8.7.2.

Outputs of this process are the filtered result sample values p'i and q'i (i = 0..2) for the set of input sample values.

The filtered result samples p'0 and q'0 are derived by
∆ = Clip3( –tC, tC, ( ( ( ( q0 – p0 ) << 2 ) + ( p1 – q1 ) + 4 ) >> 3 ) )
p'0 = Clip1( p0 + ∆ )
q'0 = Clip1( q0 – ∆ )
where the threshold tC is determined as follows.

– If chromaEdgeFlag is equal to 0, tC = tC0 + ( ( ap < β ) ? 1 : 0 ) + ( ( aq < β ) ? 1 : 0 )

– Otherwise (chromaEdgeFlag is equal to 1), tC = tC0 + 1

Depending on the values of indexA and bS the variable t'C0 is specified in Table 8-17. Depending on chromaEdgeFlag, the corresponding threshold variable tC0 is derived as follows.

– If chromaEdgeFlag is equal to 0,
tC0 = t'C0 * (1 << ( BitDepthY – 8 ) )

– Otherwise (chromaEdgeFlag is equal to 1),
tC0 = t'C0 * (1 << ( BitDepthC – 8 ) )

Let ap and aq be two threshold variables specified by
ap = Abs( p2 – p0 )
aq = Abs( q2 – q0 )

The filtered result sample p'1 is derived as follows

– If chromaEdgeFlag is equal to 0 and ap is less than β,
p'1 = p1 + Clip3( –tC0, tC0, ( p2 + ( ( p0 + q0 + 1 ) >> 1 ) – ( p1 << 1 ) ) >> 1 )

– Otherwise (chromaEdgeFlag is equal to 1 or ap is greater than or equal to β),
p'1 = p1

The filtered result sample q'1 is derived as follows

– If chromaEdgeFlag is equal to 0 and aq is less than β,
q'1 = q1 + Clip3( –tC0, tC0, ( q2 + ( ( p0 + q0 + 1 ) >> 1 ) – ( q1 << 1 ) ) >> 1 )

– Otherwise (chromaEdgeFlag is equal to 1 or aq is greater than or equal to β),
q'1 = q1

The filtered result samples p'2 and q'2 are always set equal to the input samples p2 and q2:

p'2 = p2
q'2 = q2
************************************************************ ****

For those interested in how H.264 really works, and not Amir's interpretation, the specification can now be freely downloaded from ITU:

http://www.itu.int/rec/T-REC-H.264-200503-I/en

and a synopsis of the loop filter can be found in this e-book (Chapter 6.4.7):

http://neuron2.net/library/avc/H.264.And.MPEG-4.Video.Compression.Video.Coding.For.Next.Generation.Multime dia.pdf

Ron

CE industry insider, compression engineer and Le Zouave (the guy on the cover of Zig-Zag papers) look-alike.

I don't recall saying "it was the only time." I said if you feed us bad content, we produce bad content. That is not the same as what you are saying.


Of course that is the case. Per my earlier posts, these are professional tools, requiring training, and proper usage and experience. If these are lacking, you get less than perfect results. But note that the same is true of other codecs. So this distinction by itself, is not very meaningful in the context of comparing codecs.

Thanks for the clarification. I am not really all that interested in comparing codecs. I am just a consumer. I am not in the business of marketing codecs. I am much more interested in comparing results, and trying to determine how closely the end product resembles the source material. Would you agree that there are other factors besides the quality of the source material, and the skill of the compressionist that affect the quality of the results, assuming VC-1 is used? Such as bit rate budgets, for example?

I think you will be very pleased to see the condition these titles are in when the are released to the market. Dirt and scratch removal technology has come along way since the days of an artist paint out dirt in Photoshop.

Paidgeek, thank you very much for your kind answer.

Hi,

Yes, as a small independent studio my perspective is very different. I don't have a lot of politics and partners to deal with nor do I stand to gain from any winners or losers in this game. As a matter of fact, by supporting both formats at such an early stage, I am at far more risk of losing than gaining.

If you look at the companies that stand to make fortunes off the Intellectual Property rights and Royalties from Licensing of these formats, you begin to understand why people fight so hard for their position.

Sony may be attempting to support some of the replication manufacturers, to some degree, but I don't know anything about that.

To my knowledge, Sony is certainly not supporting anything having to do with authoring, because it is clearly more expensive to author a BD format than it is HD-DVD. A lot of it has to do with the menus. They are handled differently on BD verses HD-DVD.

This whole encoding issues is along the same lines. Certain encoders are better for certain jobs. H-264 looks really good if you are compressing the heck out of something to fit on an Video i-Pod. MPEG-2 looks really great if you have a LOT of bandwidth. AVC and VC1 both look really good if you need your encode optimized for HD-DVD or BD.

There are differences between all of the encoders and they all have their application. I would go as far as to say that VC1 starts looking better than AVC once you get down to around 13mbps to 16mbps. This is my opinion of course, but really do believe VC1 may be more efficient at lower encode rates.

AVC on the other hand, may perform better given the headroom and encoding bandwidth.

All of these questions are what is driving me to do a "Real World" encode test. I want to know for sure what is the best approach for producing the best encodes.

Also, at this point in time, it is too early for us to say which format is selling better. Due to the recall on our first title, CHRONOS, I don;t have fair comparison sales figures yet. Sales impacts our retail pricing as well, since we need to amortize costs over the number of units we sell.

All of these titles are a wash for us at best...and perhaps even a loss. We are not sure yet. Time will tell, and we will continue to release new titles until we see the writing on the wall regarding the potential of the current HD formats.

In the end, I am not interested in the politics, I am only interested in turning out the best product possible. If I worked for Sony or Microsoft, I might have a different perspective, but I don't...I just work for me and my customers....:)

Best regards,

Rich



Would you say there was a difference in cost for you to produce Chronos on Blu-ray versus HD DVD? Common perception on these forums says it's more expensive for BD but that Sony is subsidizing...

Do you have any thoughts to ofter on that? Being a smaller studio, I'd find your perspective on the whole thing to be very illuminating for any studio in a similar position.

OK...fair answer..I get the analyzing thing...I am the same way...:)

I am in a different position that most small studios. My label is a labor of love, it is not my primary business. I do not reliant need my label to be profitable on ever release we do....albeit the businessman in me does a pretty good job at that. However, it is not my #1 priority.

What is a priority is learning how all these new formats work and knowing every detail I can about how this all works....from authoring to placing the product on a shelf at retail.

Why do I do this?

Because my true source of funding for my label comes from doing new business development for major corporations. Part of my business revolves around content acquisition. That means I find and secure content for labels.

I also work in IP Management and serve as an Agent / Representative for Producers. The products that makes sense for me end up on my label.

However, more than 80% of my revenue comes from Consulting, independent New Business Development. and Sales Representation. I work in the areas of new media technology, production, authoring, replication, packaging, warehousing, distribution, and fulfillment.

In the end, having the label as a test bed and being the bleeding edge early adopter I am ... provides me with the product knowledge and experience I need to provide quality consulting services and generate revenue via sales and new business development.

So why do I bore you with all this? Because my prices are as good and in most cases much better than any Major Studio. If I were to quote them, I would probably violate NDA's with companies I work for that could crush me with their team of lawyers. Based on my long standing great relationships with my partners, I doubt they would really do that, but I would certainly be in seriously hot water a the very least.

To be honest, I could probably quite the pricing of almost ever Major Studio and Independent Studio within 5% of their actual cost...some I could quote spot on. However, after my long winded explanation, perhaps you can understand why I need to be very careful about what I say in a public forum.

However, I can say whatever I like about encoding formats, HD Formats, DVD Authoring Systems, and any related Technology as I currently have no consulting agreements or contracts in place with any company that prevents me from speaking my mind.

Best regards,

Rich





Because we analyze everything to the Nth degree around here :)

One of the advantage of HD DVD was lower cost of replication. Obviously, that cost is not always passed on to smaller studios, because your price may be dictated by supply and demand rather than cost to replicate - until you get high volume. But there was some evidence that a BD disc cost about 30% more to replicate than an HD DVD.

In another thread you said that the glass master cost the same for each format - about $3k. You are the first person I know that has said that. So apprearently Bd has brought down the cost of creating the glass master. There were some advances we had talked about before, and maybe those are coming onto the market now.

But I haven't heard anything about replication cost around here in a long time. That could be because they have made advancements in yields, and there is no longer a big difference. Or maybe people just got tired of talking about it. Still, i'm particularly interested in the experience of small studios in replication.

Without getting into specifics, can you give me a percentage of how much more BD cost to replicate? Is it 0%? 30%? These are the kind of things we like to talk about around here.

And thank you in advance. I really appreciate the insiders on this board giving us this kind of insight.

...

To my knowledge, Sony is certainly not supporting anything having to do with authoring, because it is clearly more expensive to author a BD format than it is HD-DVD.

...


Without getting into numbers that he admitted could get him into hot water, there seems to be little ambiguity in this assertion.

Any counters?

Well, we don't work in anything but 96/24 if at all feasible. On occasion, we are forced to work in 48/24...but it is rare these days.

Moving forward, we are ONLY releasing HD Content in 96/24 unless the original is not available in that format. This was the case of some of the Monster Music releases we produced.

If the original is analog, we will convert to 96/24 using only Apogee or DCS Converters. Only DCS for 192/24 if that opportunity should arise. The DCS Converters are the only ones we know of at the time that do not add a lot of artifacts at that high a sampling rate.

(BTW - One main reason why we never supported SACD had to do with very high sampling rates and the "edginess" this created)

If it is in 48/24...we will ALWAYS stay in the native format to avoid any dithering....which is not also a good thing.

(NOTE: I stand corrected on Dolby Digital Plus. We do not use this so I stated an incorrect technical spec earlier in this post.)


Having worked as a consultant to DTS Entertainment for some time, I became a huge fan of DTS. I have been inside both companies, and if you were to experience the corporate culture of DTS verses Dolby, they are very different organizations. DTS truly lives and breathes audio quality...they are very passionate about it and everyone that worked for the company when I was there was a true audiophile.

To answer your questions:

1 - I can tell you that we needed 6.4mbps average and 8.2 peak to encode a 96/24 / 5.1 Soundtrack using DTS Master Audio Lossless. You would less than that for a 48/24 / 5.1 Soundtrack.

2 - DTS Master Audio Lossless and True HD (which is really MLP Audio) is much better than any lossy compression audio Codec. However, the term spectacular depends on your source material, your equipment, and your ears. If I did not think it was better I would not have used it on our releases. If I did not think it was MUCH better, I would not insist on using it on ever release. So to me, yes, I would say it is a spectacular difference in the ideal listening environment.

3 - I am a guy that still listens to tube equipment....:) Until PCM reaches the 192/24 resolution level...I can hear the difference verses a superb analog recording. The lower noise and increased dynamic range of digital is a given. However, the "stair stepping" of a PCM encode is still apparent to me and is most noticeable in the extreme highs and lows. I can hear it it in the detail and imaging. However, we are talking esoteric differences once you get to 96/24. At 192/24, the difference is dramatic and it is a world apart from anything I have ever heard. A properly recorded, mixed, and mastered original 192/24 soundtrack is a phenomenal listing experience.


However, I would hardly say that DTS at 1.54 or Dolby Digital Plus (<--corrected) sounds transparent to the source. They absolutely do not. However, they both do sound great and even I am satisfied listening to music encoded at 96/24 in DTS. Could it be better? Yes, but unless you have the original source to A/B with, the difference is not apparent and even I do not realize what I am missing...until I hear it side by side.

They are both very good. However, DTS always sounds better to me....regardless....and that is not just because I used to work for them. I just like the sound quality of their Codec...always have.


NOTE: Whenever possible, we encode from the original source from the original mastered audio file from the original hard drive to reduce jitter. We strip the clock from the signal when we master and use a Big Ben clock to regenerate a new clock signal so we have file without "jitter" issues. Jitter is equivalent to "smear" or "time misalignment"


NOTE: The digital 96/24 AIFF or 96/24 .WAV files we encode from are PCM. DTS Master Audio Lossless is transparent to the source.

Best regards,

Rich

But how much does HD-DVD need lossless (especially when it's only 16/48) if 24/48 1.5mbps Dolby Digital Plus is virtually transparent? This seems to be a question that a lot of people don't want to address, as
1. BD supporters don't want to bring it up because BD can't do 1.5mbps Dolby Digital Plus.
2. Certain people want to keep up the illusion that DTS Master Audio and True HD are something spectacularly better, even at 16/48.
3. Certain people want to keep up the illusion that PCM is better than everything else, even when it is only 16/48.

Good point...and it could be. I will have to call Ron Fricke (DP for Chronos, Baraks, etc.) and see if he can lend us come content for this application. I know he has shot in 24p and loves the results.

Thanks for the suggestion.

Rich

Richard -

Electronic cameras are increasingly able to offer 1080p/24 (or faster) material that might on good days approach 'squeaky clean'. Is that also a criterion?

Dunno how to phrase this as a question (what was that old quiz show?) but it seems the very existance of profitable electronic media distribution will exert a force to create movies that look good after being electronically coded. So wouldn't you also want to test some of these?

- Tom

Amirm, thanks for the run down and thanks to dr1394 for the clarification on some of his points ;)

Would love to hear more from you, dr1394 about AVC. I find the discussion here fascinating :D

so when is 360 add on getting its upgrade? and why cant hd dvd players do 1080P upscaling on SDVD?

Several posts that came across as a Senate confirmation hearing have been removed. I highly advise all non-insiders to read my post a page or so back about decorum on this particular thread. Either that, or I'm renaming it "Defendant's Question and Answer and Cross Examining Thread."

However, I would hardly say that DTS at 1.54 or Dolby Tru-HD sounds transparent to the source. They absolutely do not.

You are here saying that lossy DTS and lossless Dolby TrueHD are absolutely not transparent to the source. Well, I would expect that for the DTS (though I always appreciated their 1.5Mbps sound tracks :)), but I would like to know why that would be the case for TrueHD. Are you basically saying that MLP is a flawed "lossless" codec?

Anybody have any information as to what the real problem is here?

Ben,

Just watched "The Illusionist" on XBox Live. Even though it's not an "HD" that movie could fool many. After watching the movie, I put on Gladiator SD DVD (my best DVD PQ wise) upscaled on the HD-A1 and I have to say the Illusionist blows it away. The detail in the picture, the preservation of the black and shadow detail was really good.

Now get broadcast people do use VC-1 at the same bitrate as the Illusionist encode and we'd all be damn happy! Is that bitrate possible on cable/sat? Look like 480 has plenty of life left given the proper treatment :)

so when is 360 add on getting its upgrade? and why cant hd dvd players do 1080P upscaling on SDVD?

Spring, and who says they can't?

You are here saying that lossy DTS and lossless Dolby TrueHD are absolutely not transparent to the source. Well, I would expect that for the DTS (though I always appreciated their 1.5Mbps sound tracks :)), but I would like to know why that would be the case for TrueHD. Are you basically saying that MLP is a flawed "lossless" codec?

Anybody have any information as to what the real problem is here?

Trevor... If I can speak for him, I think he meant to say DD+ and not True HD. If not, I would be confused also.

I meant Dolby Digital Plus.....sorry.

I don't use Dolby Tru-HD ... it is basically MLP Audio. Dolby is not flawed...I just prefer DTS.

Rich

You are here saying that lossy DTS and lossless Dolby TrueHD are absolutely not transparent to the source. Well, I would expect that for the DTS (though I always appreciated their 1.5Mbps sound tracks :)), but I would like to know why that would be the case for TrueHD. Are you basically saying that MLP is a flawed "lossless" codec?

Anybody have any information as to what the real problem is here?

My costs range from 20% to 50% more depending on the complexity of the menus...if that helps.

Without getting into numbers that he admitted could get him into hot water, there seems to be little ambiguity in this assertion.

Any counters?

My costs range from 20% to 50% more depending on the complexity of the menus...if that helps.

That was very helpful, thank you.

Ron, appreciate you chiming in. It is great to have another compression expert to interact with on these topics :). And sorry for the late reply. Had to go back to getting some business done while here.

Anyway, your objection is somewhat valid wrt to wording of the last statement I made, but not relative to the actual argument :). So let me state that more clearly. There is no way to force AVC to use single pixel filtering as VC-1 uses. And that is factual and evident from the pseudo code. I can’t quite tell from the wording of your statement if you are disputing this or not. I could read your statement as “sometimes AVC can pixel filter just one pixel” as opposed to it having a fixed one-pixel mode. Just in case you mean the latter, and for the benefit of others, please allow me to explain why this is not correct.

What the AVC spec allows, as with VC-1 or any other codec with a loop filter, conditions for not filtering content based on certain criteria. This includes things like whether this is an intra/inter block, or the block is a boundary block in a macroblock, or if this pixel lies on an edge (think of left side of your screen – you can’t go past it to filter anything), and the quantization step size (how much you are compressing the video – the adaptive part of the filter). But when these special conditions are not met (say, a block encapsulating background grain in the somewhere away from the edge), the algorithm goes out and filters three pixels on each side of the block, even in a small 4x4 block! As a result, some/many pixels will be filtered multiple times!

You can verify the above from the pseudo code. The first part clearly lists that there are three pixels potentially being operated on (“input sample values pi and qi (i = 0..2)” which means pixels corresponding to i=0, 1 and 2, or three pixels). So we know this code is not talking about single filtering mode at all.

Then there are conditional “if” statement which examine the above factors and if they are not triggered, the filter goes to operate on the three pixels. If those conditions are met, the “otherwise” part kicks in and filters the one pixel which you highlighted (classic “if then, else” statement for you programmers :)). Even when the “otherwise” part is triggered, it is only for that block. The next block can easily get the reverse treatment and get filtered all the way (which again can mean multiple filtering of the same pixels in a 4x4 block). And the block after that and so on.

Net, net, other than drilling into details of how this algorithm works (showing the complexity behind even seemingly simple things), the assertion I made remains valid. AVC will filter and reduce grain and fine texture as a result of this aggressive filtering as compared to VC-1. It will attack many blocks and pixels within them as it runs into difficulty coding them without artifacts. Not only is this evident from the code/spec, but from many encodings and shoot outs we have participated in. The loop filter without a doubt is a major factor in how AVC softens detail in difficult material.

Talking about real pictures, have you had a chance to encode any content with AVC for these HD formats? If so, have you seen an encoder with a single pixel loop filter? If so, who makes it and what title did you work on?

Thanks again for commenting. As Ken Watenabe said to Tom Cruise’s character in The Last Samurai, “I enjoy our conversations!” (You can tell I am still in Japan, can’t you? :)).

If you're taking less samples, what about aliasing effects then? Surely, the values being filtered will be more vulnerable to aliasing effects, lowering the overall prediction efficiency.

Ron, appreciate you chiming in. It is great to have another compression expert to interact with on these topics :). And sorry for the late reply. Had to go back to getting some business done while here.

Anyway, your objection is somewhat valid wrt to wording of the last statement I made, but not relative to the actual argument :). So let me state that more clearly. There is no way to force AVC to use single pixel filtering as VC-1 uses. And that is factual and evident from the pseudo code. I can’t quite tell from the wording of your statement if you are disputing this or not. I could read your statement as “sometimes AVC can pixel filter just one pixel” as opposed to it having a fixed one-pixel mode. Just in case you mean the latter, and for the benefit of others, please allow me to explain why this is not correct.

What the AVC spec allows, as with VC-1 or any other codec with a loop filter, conditions for not filtering content based on certain criteria. This includes things like whether this is an intra/inter block, or the block is a boundary block in a macroblock, or if this pixel lies on an edge (think of left side of your screen – you can’t go past it to filter anything), and the quantization step size (how much you are compressing the video – the adaptive part of the filter). But when these special conditions are not met (say, a block encapsulating background grain in the somewhere away from the edge), the algorithm goes out and filters three pixels on each side of the block, even in a small 4x4 block! As a result, some/many pixels will be filtered multiple times!

You can verify the above from the pseudo code. The first part clearly lists that there are three pixels potentially being operated on (“input sample values pi and qi (i = 0..2)” which means pixels corresponding to i=0, 1 and 2, or three pixels). So we know this code is not talking about single filtering mode at all.

Then there are conditional “if” statement which examine the above factors and if they are not triggered, the filter goes to operate on the three pixels. If those conditions are met, the “otherwise” part kicks in and filters the one pixel which you highlighted (classic “if then, else” statement for you programmers :)). Even when the “otherwise” part is triggered, it is only for that block. The next block can easily get the reverse treatment and get filtered all the way (which again can mean multiple filtering of the same pixels in a 4x4 block). And the block after that and so on.

Net, net, other than drilling into details of how this algorithm works (showing the complexity behind even seemingly simple things), the assertion I made remains valid. AVC will filter and reduce grain and fine texture as a result of this aggressive filtering as compared to VC-1. It will attack many blocks and pixels within them as it runs into difficulty coding them without artifacts. Not only is this evident from the code/spec, but from many encodings and shoot outs we have participated in. The loop filter without a doubt is a major factor in how AVC softens detail in difficult material.

Talking about real pictures, have you had a chance to encode any content with AVC for these HD formats? If so, have you seen an encoder with a single pixel loop filter? If so, who makes it and what title did you work on?

Thanks again for commenting. As Ken Watenabe said to Tom Cruise’s character in The Last Samurai, “I enjoy our conversations!” (You can tell I am still in Japan, can’t you? :)).

Aren't both encoders smart enough to avoid the filtering completely when there is siffucient bit rate to avoid blocks anyway?

- Tom

Aren't both encoders smart enough to avoid the filtering completely when there is siffucient bit rate to avoid blocks anyway?

Sure, but that's not a given at HD optical bitrates.

Sure, but that's not a given at HD optical bitrates.Ben, are you referring to "inloop filter" specifically, or something another side effect in general caused by compression itself, like high QP and high numbers used in quant matrix?

Aren't both encoders smart enough to avoid the filtering completely when there is siffucient bit rate to avoid blocks anyway?

Sure, but that's not a given at HD optical bitrates.


Hi Ben,

So are you saying that 'optical disc rates' are not high enough to guaranty optimal compression?

And if so, having more bitrate is actually beneficial to the encode?

What bitrate would make that a given?

If you're taking less samples, what about aliasing effects then? Surely, the values being filtered will be more vulnerable to aliasing effects, lowering the overall prediction efficiency.
Good observation :). Fortunately, it is not a concern here because we are not talking about the filter taps (samples) in the loop filter. Rather, how many samples are modified/filtered. The actual filter length is fairly comparable in the two codecs.

Even if you were right, surely you don't want to filter a pixel two or more times, in the name of anti-aliasing, do you? Because that is exactly what happens with AVC if you filter a 4x4 block, with 3 pixels from the sides.

If you are talking about the motion compensation, and not the loop filter, then I already answered that. Higher order filters ring, and that is not good. Codec design is not just science. It is also art. You have to balance things based on real experience with real content. And here, we had the upper hand, with a lot more HD experience, allowing us to play with different filters and see which one worked better. It is not like a higher order filter was an unknown science to us or difficult to do :). It was the result of our testing with high-res content that led us to our choice there.

BTW, a bicubic 4-tap filter is quite nice by video standards. When I managed engineering of broadcast equipment, we couldn't afford the hardware cost of bi-cubic so the common choice was bilinear filtering!

So are you saying that 'optical disc rates' are not high enough to guaranty optimal compression?
Not without triggering the loop filter, no.

And if so, having more bitrate is actually beneficial to the encode?
For AVC, yes. VC-1 hits the detail bar at a lower bitrate.

Hi Ben,

So are you saying that 'optical disc rates' are not high enough to guaranty optimal compression?

And if so, having more bitrate is actually beneficial to the encode?
Ben may have his own answer to this. But I don’t believe he was not talking about what is "optimal" and what is not. Rather, he was answering Tom’s question as to whether a loop filter would ever kick in. And the answer is that it does. This is evidenced by blocking artifacts in some BD titles using MPEG-2 despite using the high peak rate. Lack of loop filter makes them visible and hence the reason it is part of our codec and that of VC-1.

Now whether the artifacts in case of MPEG-2 or the effect of loop filter, make the format less “optimal” is a subjective thing for people to go and argue to death, as we have done here at length :). Fact is that the performance curve of advanced codes is even more non-linear than MPEG-2. They approach the quality of the source much faster (i.e. at earlier data rates) than MPEG-2 does, and as you increase the data rate, the incremental improvements can be very small. Would you be able to still see them? Only you can answer that. But if you are happy with MPEG-2 at BD rates, then you should be just as happy with VC-1 at HD DVD rates because of its increased efficiency and distortion mitigation. If this is not believed, then I can do another article on differences between VC-1 and MPEG-2.

Oops. Didn't missed your post Ben while I was answering the same question :). Sorry about that.

Oops. Didn't missed your post Ben while I was answering the same question :). Sorry about that.
Don't let that dissuade you from writing that MPEG-2/VC-1 comparison, though!

Amir,

in light of your recent discussion on codecs, one question:

In your opinion, would you say that neither MPEG2 and AVC cannot reach visual transparency using the bitrates avialable to it on Blu-ray?

Amir,

in light of your recent discussion on codecs, one question:

In your opinion, would you say that neither MPEG2 and AVC cannot reach visual transparency using the bitrates avialable to it on Blu-ray?
It would be unfair to make a general characterization that way wrt to the experience viewers have. Clearly BD has some really good looking titles now, gaining praise from reviewers and consumers alike. So for these titles, and using the eyes of enthusiasts, I would say BD has achieved visual transparency.

If you are asking me what I personally think when looking at the overall portfolio of titles in BD, then I am left wondering how many of them could be improved with VC-1. It is just our nature to analyze the picture, thinking like a “human codec” (:)), spotting things that we know would stress MPEG-2 or AVC when it comes to visual transparency. Having said this, I have to be frank and say that it is unclear whether people could see the visual improvements from VC-1 over these other codecs in every title. What we see should not always be confused with what others see.

xboxboi is quoting Amir insinuating that Universal is dropping SD DVD and going combo only for new releases and HD DVD only for catalog? What's the time frame for the change over?

Rather, he was answering Tom’s question as to whether a loop filter would ever kick in.


Ben or Amir (or both :)),

If a loop filter is needed for a certain segment to mask blocking, is the image still 'transparent to the master', or does the act of filtering, while less distruptive to the viewer than blocking would be, cause the image to be no longer transparent to the master?

I am enjoying this informative discussion :)

If a loop filter is needed for a certain segment to mask blocking, is the image still 'transparent to the master', or does the act of filtering, while less distruptive to the viewer than blocking would be, cause the image to be no longer transparent to the master?

Yes, the output of a loop filter can be visually transparent the master, and certainly a lot closer than without using the loop filter. Part of why loop matters is that it can eliminate a small error which would then propogate into future frames, and so on down through the GOP.

The key is to get enough of an effect to be useful, but not so much as to overly soften.

Aren't both encoders smart enough to avoid the filtering completely when there is sufficient bit rate to avoid blocks anyway?

- Tom
For H.264, the loop filter is controlled by quantization level Qp. Here's the table for the decision on whether more that a single pixel per block edge is filtered.

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2 2 2 3 3 3 3 4 4 4

26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51
6 6 7 7 8 8 9 9 10 10 11 11 12 12 13 13 14 14 15 15 16 16 17 17 18 18

Pixels are labeled like so, with the block edge in the middle.

p3 p2 p1 p0 | q0 q1 q2 q3

The top value is Qp and the bottom value is the level that the amount of pixel difference between p1 and q1 must be less than for the filter to engage.

If the average Qp between the two blocks is 15 or less, the filter does not engage. For Qp 16, the pixel level difference between p1 and q1 must be less then 2. In other words, if the difference between p1 and q1 is larger than 2 (meaning there's some detail in the block), the filter is not used on p1 and q1 (and p2 and q2).

Amir likes to characterize the H.264 loop filter as a "blunt eraser". Looking at the numbers above, I would say that filtering of anything but the block edges is done with a "light touch".

Ron

Ron -

I don't even know if there is any industry consensus that filtering only 1 vs 3 pixels or doing it with only 4 vs 6 taps is really better anyway. It may be one of those religious issues. But I do know that on the popular open source X264 codec you can enter about 4 different levels for relatively how much loop filtering you want. I'm assuming the commercial codecs can do similarly.

To my questions ... there have indeed been some comments that some AVC encodes tend to look softer than we might expect. Do you feel this is really from the AVC loop filtering, some other reason (other filtering?), or just false?

Also, do you know if most AVC highdef disc encodes are using an 8x8 transform? It seems they would be since that was put into the profile for HD. If so, that would avoid the double loop filtering that has been discussed.

- Tom

Ron -

I don't even know if there is any industry consensus that filtering only 1 vs 3 pixels or doing it with only 4 vs 6 taps is really better anyway. It may be one of those religious issues. But I do know that on the popular open source X264 codec you can enter about 4 different levels for relatively how much loop filtering you want. I'm assuming the commercial codecs can do similarly.

To my questions ... there have indeed been some comments that some AVC encodes tend to look softer than we might expect. Do you feel this is really from the AVC loop filtering, some other reason (other filtering?), or just false?

Also, do you know if most AVC highdef disc encodes are using an 8x8 transform? It seems they would be since that was put into the profile for HD. If so, that would avoid the double loop filtering that has been discussed.

- Tom
Loop filtering "strength" is controlled with an offset for the table (and another table for the block edge pixels) in my previous post. The offset can be up to plus or minus 12 (in steps of two) added to the Qp value. For example (with the posted table), if you set the offset to -12, then the filter won't engage until Qp = 28. This offset can be changed every picture.

I'm working on a trick to show exactly which pixels are being loop filtered on decoded images. Stay tuned.

I don't have much H.264 movie content. On the one clip I do have (Equilibrium), it is High Profile, and the encoder selects 8x8 quite a bit.

Ron

xboxboi is quoting Amir insinuating that Universal is dropping SD DVD and going combo only for new releases and HD DVD only for catalog? What's the time frame for the change over? The direct question was already directly answered by Amirm.

Combos when used will be only used for new releases not previously released on DVD, possibility still of some dual HD DVD and DVD releases still for some of the new titles.

Previously released DVD catalog titles will be released HD DVD only.Originally Posted by brez
Amir,

Have you had any response from the studios on the combos? (on Combos)Glad you asked .

I spoke with Uni's Ken Graffeo this week, who clarified Universal's combo disc strategy. Given that most of us already own DVD catalog titles of our favorite movies, Universal plans to only release new movies in combo disc, and to release catalog titles as stand-alone HD DVDs. That means they'll no longer have tier pricing for combo discs and new releases will be competitively priced with other titles in the market. This is good news for us, since that rationalizes the price points, which I know has been a subject of debate on the Forum.

So you see, the voices of majority do have some power here .
http://www.avsforum.com/avs-vb/showthread.php?p=9886623&&#post9886623

Guys, sorry about causing confusion here. My only job was to find out Uni's HD DVD combo/pricing policy, not DVD. I did not mean to imply anything about them turning off normal DVD production and switching to combos only. I can see how my wording could be construed the way it was though. So hopefully no flogging occurs for folks who were led astray . http://www.avsforum.com/avs-vb/showthread.php?p=9950328&&#post9950328

Ron -

I don't even know if there is any industry consensus that filtering only 1 vs 3 pixels or doing it with only 4 vs 6 taps is really better anyway.

- Tom
BTW, there is no 6-tap filter in the H.264 loop filter. Only 3, 4 and 5-tap.

Ron

For H.264, the loop filter is controlled by quantization level Qp. Here's the table for the decision on whether more that a single pixel per block edge is filtered.

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2 2 2 3 3 3 3 4 4 4

26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51
6 6 7 7 8 8 9 9 10 10 11 11 12 12 13 13 14 14 15 15 16 16 17 17 18 18

Pixels are labeled like so, with the block edge in the middle.

p3 p2 p1 p0 | q0 q1 q2 q3

The top value is Qp and the bottom value is the level that the amount of pixel difference between p1 and q1 must be less than for the filter to engage.

If the average Qp between the two blocks is 15 or less, the filter does not engage. For Qp 16, the pixel level difference between p1 and q1 must be less then 2. In other words, if the difference between p1 and q1 is larger than 2 (meaning there's some detail in the block), the filter is not used on p1 and q1 (and p2 and q2).

Amir likes to characterize the H.264 loop filter as a "blunt eraser". Looking at the numbers above, I would say that filtering of anything but the block edges is done with a "light touch".

Ron
dr,

according to your explanation, there are instances when the filter is not triggered; how often would you say this occurs, satistically?

when the conditions for filtering ARE met, which pixels get filtered? all three on each side of the boundary? are there conditions that would entail a varied number of pixels to be filtered or is it all or nothing?

also, when all ARE filtered, how does the level of filtering vary wrt distance from the boundary?

thx

dr,

according to your explanation, there are instances when the filter is not triggered; how often would you say this occurs, statistically?

when the conditions for filtering ARE met, which pixels get filtered? all three on each side of the boundary? are there conditions that would entail a varied number of pixels to be filtered or is it all or nothing?

also, when all ARE filtered, how does the level of filtering vary wrt distance from the boundary?

thx
I hope to be able to answer these questions graphically with real decoded images by modifying the reference decoder. As I said before, stay tuned.

Ron

[QUOTE=Kosty]The direct question was already directly answered by Amirm.

Combos when used will be only used for new releases not previously released on DVD, possibility still of some dual HD DVD and DVD releases still for some of the new titles.

QUOTE]


Are you an insider? Possibility that there will still be DVD and HD DVD new releases...are you implying that they are dropping DVD for most of their new releases and going combo only? that's not how I read what Amir said at all...

With HD DVD we have Mandatory Managed Copy.

Really, How do I make one?

- Rich

Both formats have Managed Copy. Mandatory is up to the studios not the formats, and AFAIK Managed Copy hasn't been implemented yet.

Not to quibble, but I did not name it "Mandatory Managed Copy" that was done by the HD DVD / BD crowd. Since they named it, I was hoping for an insider to explain this. You would think this would be a good time when there is now an unmanaged copy option.

- Rich

Both formats have Managed Copy. Mandatory is up to the studios not the formats, and AFAIK Managed Copy hasn't been implemented yet.
Not to quibble, but I did not name it "Mandatory Managed Copy" that was done by the HD DVD / BD crowd. Since they named it, I was hoping for an insider to explain this. You would think this would be a good time when there is now an unmanaged copy option.

- Rich

AACS continues to negotiate the details for Mandatory Managed Copy. The Mandatory part of this refers to the fact that per the licensing arangement nearly all titles would be available for copying, and there would not be studio discretion on which titles can be copied.

AACS continues to negotiate the details for Mandatory Managed Copy. The Mandatory part of this refers to the fact that per the licensing arangement nearly all titles would be available for copying, and there would not be studio discretion on which titles can be copied.

Thanks for the answer. Has the recent AACS incident cause any back lash on this MMC negotiation?

Ben or Amir (or both :)),

If a loop filter is needed for a certain segment to mask blocking, is the image still 'transparent to the master', or does the act of filtering, while less distruptive to the viewer than blocking would be, cause the image to be no longer transparent to the master?

I am enjoying this informative discussion :)

Related with this question, I would like to have something clarified.

Negative feedback when related to audio inherently includes a time delay which can potentially have a disruptive effect on the sound.

I'm guessing that in the case of the AVC and VC-1 feedback loop, we are actually talking about corrections made to best approximate what is effectively a static frame. That means that there is no feedback smear/filter effect due to immediate comparison with the next or previous frame. It is purely a correction applied to a specific frame encoding to provide more optimal matching between that encoding and the original frame.

Is this a fair assessment?

Can any BD insider say if there is any significance in the pricing scheme during the current Amazon.com sale, i.e. is this the sign of things to come, Fox putting out reasonably priced BDs. I would love it if were true!

Negative feedback when related to audio inherently includes a time delay which can potentially have a disruptive effect on the sound.

I'm guessing that in the case of the AVC and VC-1 feedback loop, we are actually talking about corrections made to best approximate what is effectively a static frame. That means that there is no feedback smear/filter effect due to immediate comparison with the next or previous frame. It is purely a correction applied to a specific frame encoding to provide more optimal matching between that encoding and the original frame.

Is this a fair assessment?
I am not quite sure what you are asking. But let me first say that all modern video codecs work this way, including MPEG-2. The feedback loop is part of figuring out what to send next. It is an attempt, bit rate permitting, to match the original if that is what your asking.

I am not quite sure what you are asking. But let me first say that all modern video codecs work this way, including MPEG-2. The feedback loop is part of figuring out what to send next. It is an attempt, bit rate permitting, to match the original if that is what your asking.

I thought it was previously mentioned that whereas VC-1 and AVC use loop feedback, MPEG-2 doesn't (associated with age of technology and available computing power at the time) -- you are saying they all use loop feedback?

In any case, I'm trying to understand the potential source of a smear/filter effect resulting from applying feedback to the encoding of the frames. A feedback loop basically has the following model:

Original Image(Ti) --> (Original-Decode) -> Encoder ----> Output (Tij)
...........................................^ (Tij)....................................|
...........................................|------------ Decoder <---------|

Unlike audio where feedback is a continuous function of time, I'm anticipating the CODEC feedback operates entirely on one individual frame at a time (Ti). The number of passes of that discrete feedback signified by Tij. The practical number of discrete feedback passes being inherently limited only by available computation power and the settling speed to an optimal stable itterative approximation.

I would have expected that if the above is correct, then no filter/smear effect should necessarily result.

Potential problems would be if the feedback on one data point was actually derived from more than one frame (say previous and next for example) and/or derived from more than one data point while ultimately acting on an individual data point.

I guess the problem would come from having to use lossy compression where some of the generated correction results from subtracting "approximated" data from original data and including that corrective encoding in the next itteration. There is probably an inherent smear factor in that case and the best one can hope for is an optimized smear (and/or additional bitrate) -- though it's clear a difference had to exist or else the feedback error would be zero. The ability to actively control the pixel depth of the feedback action would surely help minimize that effect. Very interesting problem!

I thought it was previously mentioned that whereas VC-1 and AVC use loop feedback, MPEG-2 doesn't (associated with age of technology and available computing power at the time) -- you are saying they all use loop feedback?
I know the terminology here can get complicated and confusing :). But all three codecs use a feedback loop. But only AVC and VC-1 use a *filter* in that feedback loop. What we have been discussing is the filter, not the loop itself.

I know the terminology here can get complicated and confusing :). But all three codecs use a feedback loop. But only AVC and VC-1 use a *filter* in that feedback loop. What we have been discussing is the filter, not the loop itself.
Perhaps it would be helpful to mention that "loop filter" is really short for "in-loop deblocking filter."

The big change is adding a deblocking filter as a process inside the loop filter. This makes it mandatory, instead of the optional deblocking filters that existed for some MPEG-2 players.

Perhaps it would be helpful to mention that "loop filter" is really short for "in-loop deblocking filter."

The big change is adding a deblocking filter as a process inside the loop filter. This makes it mandatory, instead of the optimal deblocking filters that existed for some MPEG-2 players.

Ah, that's interesting!

Implication is that whereas an inherent compression function of the VC-1 encoding is to organize pixels into blocks (8x8,4x8,4x4,8x4), a "deblocking" filter is intended to minimize the potential visible impact of that organization. Can it assist in recovering any lost detail within the blocks, in addition to minimizing the appearance of boundaries between blocks?

Such a filter would necessarily compare only the current frame encoded output with the actual current frame (no future or previous frames drirectly involved).

Am I close?

I think you meant to say optional.
Thanks, corrected. Voice recognition and I are still training each other :).

Implication is that whereas an inherent compression function of the VC-1 encoding is to organize pixels into blocks (8x8,4x8,4x4,8x4), a "deblocking" filter is intended to minimize the potential visible impact of that organization.
Correct.

Can it assist in recovering any lost detail within the blocks, in addition to minimizing the appearance of boundaries between blocks?
Nope, just block edges.

Such a filter would necessarily compare only the current frame encoded output with the actual current frame (no future or previous frames drirectly involved).
The critical part of it being "in-loop" is that future frames are based on the DEBLOCKED version of past frames.

Am I close?
I think so.

Nope, just block edges.


The critical part of it being "in-loop" is that future frames are based on the DEBLOCKED version of past frames.

Great -- thanks much!

Regarding any possible loss of detail within a block, I gather it's the responsibility of the algorithm selecting the block size and location to handle that optimally.

Are there any possible situations where the deblocking filter might not be able to perform its task properly (namely, blending adjacent block edges)?

Regarding any possible loss of detail within a block, I gather it's the responsibility of the algorithm selecting the block size and location to handle that optimally.
Right. And this is a dynamic process that gets a lot of R&D with PEP, as we use differential quantization, where we compress different 16x16 blocks to different degrees.

Are there any possible situations where the deblocking filter might not be able to perform its task properly (namely, blending adjacent block edges)?
There are heuristics in there where it's applied proportionally to how compressed the adjacent blocks are. Combined with variable block size (using smaller blocks with less compression near those edges), this doesn't wind up being that big an issue

Regarding any possible loss of detail within a block, I gather it's the responsibility of the algorithm selecting the block size and location to handle that optimally.
Right. And this is a dynamic process that gets a lot of R&D with PEP, as we use differential quantization, where we compress different 16x16 blocks to different degrees.
Are there any possible situations where the deblocking filter might not be able to perform its task properly (namely, blending adjacent block edges)?
There are heuristics in there where it's applied proportionally to how compressed the adjacent blocks are. Combined with variable block size (using smaller blocks with less compression near those edges), this doesn't wind up being that big an issue

Hmmm, 16x16 block compression calculations, though apparently applied just for analysis purposes (I'm under the impression your maximum encoded blocksize is 8x8)! And I imagine that the "block" size on occasion can reduce all the way down to single pixels (though obviously not helpful in achieving compression). I'm running into terms I'm unfamiliar with and so a little research is in order. (Differential equations -- yes, differential quantization -- no :)) It seems like an interesting subject and one I would like to better understand. Hope to be back with a follow up.

Thanks Ben -- Trevor :)

PS. Could do with a hint on the meaning of PEP. I'm guessing something related to perceptual encoding.

Hmmm, 16x16 block compression calculations, though apparently applied just for analysis purposes (I'm under the impression your maximum encoded blocksize is 8x8)! And I imagine that the "block" size on occasion can reduce all the way down to single pixels (though obviously not helpful in achieving compression). I'm running into terms I'm unfamiliar with and so a little research is in order. (Differential equations -- yes, differential quantization -- no :)) It seems like an interesting subject and one I would like to better understand. Hope to be back with a follow up.
A 16x16 block is a macroblock, consisting of two-by-two 8x8 luma blocks, and a single 8x8 block each of the Cb and Cr chroma blocks. Since color is subsampled, a macroblock is the smallest "independent" part of a frame.

Quantization is basically a measure of how compressed the image gets (I'll say more once I can type better again...). The "quality" slider making a JPEG is setting the quantization level.

PS. Could do with a hint on the meaning of PEP. I'm guessing something related to perceptual encoding.
Parallel Encoder Program. Not named by marketing :).

Here's my first attempt at showing the effects of the H.264 loop filter. Unfortunately, my only H.264 movie clip (Equilibrium) has the loop filter shut off (at least for the first hundred frames). So instead, I tried my technique on a BBC HD trial bitstream. Note that the BBC streams are Main Profile (4x4 blocks only) and real-time CBR at 1440x1080.

Filtered pixels have been converted to peak white. This picture shows the amount of inner pixels (that is, not block edge pixels) that have been filtered. This is an I-frame (which typically gets the most filtering).

EDIT: Looks like I exceeded the bandwidth at imageshack. Here's the images from my website. loop.jpg is the block edge pixels, loop1 is the inner pixels (the pic I had posted before) and loop.jpg is all of the filtered pixels.

http://www.w6rz.net/loop.jpg

http://www.w6rz.net/loop1.jpg

http://www.w6rz.net/loop2.jpg

Ron

Here's my first attempt at showing the effects of the H.264 loop filter. Unfortunately, my only H.264 movie clip (Equilibrium) has the loop filter shut off (at least for the first hundred frames). So instead, I tried my technique on a BBC HD trial bitstream. Note that the BBC streams are Main Profile (4x4 blocks only) and real-time CBR at 1440x1080.

Filter pixels have been converted to peak white. This picture shows the amount of inner pixels (that is, not block edge pixels) that have been filtered. This is an I-frame (which typically gets the most filtering).



Ron

You couldn't find a picture with less fine detail in it, could you?

Here's my first attempt at showing the effects of the H.264 loop filter. Unfortunately, my only H.264 movie clip (Equilibrium) has the loop filter shut off (at least for the first hundred frames). So instead, I tried my technique on a BBC HD trial bitstream...

It is unfortunate for your demo but it does support the conjecture that loop filtering may often be off (or low) in hidef DVD with lots of bits. It would be interesting to find out the results of more samples.

Either way, interesting demo!

- Tom

Thanks for the answer. Has the recent AACS incident cause any back lash on this MMC negotiation?

Only in the sense that AACS has had to focus attention on the incident which may have added further delay to the completion of the final agreement.

Amir,

Is Paramount sticking with VC-1 on HD DVD for their upcoming releases http://dvd.themanroom.com/dvd-newsview.php?id=498 ?

Thanks,
Robert.

all insiders,

What is the maximum number of buttons (on menu) for HD DVD and Blu-Ray (are there any limitations)?

Thanks

Thanks to all insiders for making this such a great thread.

I've always been curious as to the process involved with studios selecting which movies they are going to release in a HD format. Could any of the insiders comment on the general processes involved, even if it is not studio specific? Are there formal comities used or does it come down to certain individuals calling the shots? Do they have a formula that can quantify how many disks might sell now verses later? For old catalog titles what would separate one from the other? I imagine it could get very complicated but any insight into this would be greatly appreciated,
thank you.

Regarding any possible loss of detail within a block, I gather it's the responsibility of the algorithm selecting the block size and location to handle that optimally.
Right. And this is a dynamic process that gets a lot of R&D with PEP, as we use differential quantization, where we compress different 16x16 blocks to different degrees.

A 16x16 block is a macroblock, consisting of two-by-two 8x8 luma blocks, and a single 8x8 block each of the Cb and Cr chroma blocks. Since color is subsampled, a macroblock is the smallest "independent" part of a frame.

Quantization is basically a measure of how compressed the image gets (I'll say more once I can type better again...). The "quality" slider making a JPEG is setting the quantization level.

Parallel Encoder Program. Not named by marketing :).

So, if I interpret this correctly. There are multiple computers/threads processing consecutive 16X16 macroblocks at varying quantization levels with the purpose of comparing the results and selecting the particular quantization that provides the most "favorable" result for each individual macroblock? (Such a quantization "pass" could occur as a precurser to a luminance analysis that would be looking at potential luma-block assignments.)

I was thinking the encoded luma-blocks can be fully flexibly located, or are they actually restricted to being entirely contained within a sampled macroblock? I'm presuming the 16x16 macroblocks are a fixed matrix overlay that subdivides the original JPEG. Not allowing luma-blocks to cross macroblock boundaries would seem a natural fallout of the macroblock specific subsampled chroma blocks, as well as a possible analysis focus on one macroblock at a time.

Of course, it's different if luma-blocks are permitted to cross macroblock boundaries. And in that case, the chroma applied to a single luma-block could potentially be governed by as many as four separate sets of chroma blocks (both Cr & Cb per set) -- each applied to separate portions of course.

However, given a luma-block is inherently confined to a fixed luminance level (excluding deblocking filter action), recognizing a luminance match between adjacent macroblocks might be awkward. Further, if there is potential for a luma-block extending across macroblock boundaries, an analysis pass would have to be made that simultaneously encompasses up to four adjacent macroblocks -- possibly messy in a multi-cpu environment.

On the other hand, the luminance analysis and subsequent assignment of luma-blocks could be performed fully independently of the macroblock quantization analysis and of the macroblocks themselves.

My guess would be that maintaining the 16x16 macroblock as a fixed boundary for all processing would both simplify the processing and be more conveniently parceled out to multiple quasi independent processors/threads. That would imply luma-blocks are indeed not allowed to cross macroblock boundaries.

Do any of my ramblings come close? :)

Are there any possible situations where the deblocking filter might not be able to perform its task properly (namely, blending adjacent block edges)?
There are heuristics in there where it's applied proportionally to how compressed the adjacent blocks are. Combined with variable block size (using smaller blocks with less compression near those edges), this doesn't wind up being that big an issue

Still got to work on this one :).

Here's my first attempt[...]FYI: the image is gone (well, it's that ImageShack frog thing, anyways). :(

So, if I interpret this correctly. There are multiple computers/threads processing consecutive 16X16 macroblocks at varying quantization levels with the purpose of comparing the results and selecting the particular quantization that provides the most "favorable" result for each individual macroblock? (Such a quantization "pass" could occur as a precurser to a luminance analysis that would be looking at potential luma-block assignments.)
Actually, each computer works on a big indepentant chunk (typically a chapter or more per worker). The details are a little different from what you said, but yes, there's basically a heuristic to determine which blocks need lighter compression compared to others (smooth gradients need to be compressed less, for example).

I was thinking the encoded luma-blocks can be fully flexibly located, or are they actually restricted to being entirely contained within a sampled macroblock? I'm presuming the 16x16 macroblocks are a fixed matrix overlay that subdivides the original JPEG. Not allowing luma-blocks to cross macroblock boundaries would seem a natural fallout of the macroblock specific subsampled chroma blocks, as well as a possible analysis focus on one macroblock at a time.

Correct. Although note the source is uncompressed 4:2:0, not JPEG.

However, given a luma-block is inherently confined to a fixed luminance level (excluding deblocking filter action), recognizing a luminance match between adjacent macroblocks might be awkward. Further, if there is potential for a luma-block extending across macroblock boundaries, an analysis pass would have to be made that simultaneously encompasses up to four adjacent macroblocks -- possibly messy in a multi-cpu environment.
We don't thread down to the macroblock level, for the reasons you suggest.

Do any of my ramblings come close? :)
You're doing well.

FYI, if you're a SMPTE member, you can get a copy of the VC-1 spec.

To Amir or other Microsoft insider:

Is the upcoming firmware update for the 360 HD-DVD player going to fix the inability to play the HD-DVDs "Discovery Atlas: Australia Revealed" and "Discovery Atlas: Italy Revealed"? Thanks, and keep up the good work!

FYI: the image is gone (well, it's that ImageShack frog thing, anyways). :(
Repaired and some additional images added.

Ron

OK, I got the answer. It seems that this is an implicit option, rather than explicit. But it is there.

During first-run setup (or later under settings) on Vista, if you set the resolution to 4:3 (or similar non-16:9 resolution) but that you indicate you have a widescreen display, MCE will assume non-square pixel and set the display mode accordingly.

Thanks for looking into this Amir. It's a bit of a blow that I'll have to upgrade to Vista though. Is there no fix for existing Windows XP MCE 2005 owners?

Only in the sense that AACS has had to focus attention on the incident which may have added further delay to the completion of the final agreement.

The ability to make legal copies should go a long way to reducing the desire to make illicit copies.

Perhaps they should be speeding up the final agreement and MC?

Gary

The ability to make legal copies should go a long way to reducing the desire to make illicit copies.

Perhaps they should be speeding up the final agreement and MC?

Gary

Insiders: Are we talking about "Managed Copy" or "Mandatory Managed Copy"?
As far as I know, HD DVD was always using the "Mandatory" word. Has that changed?

- Rich

Amir,

It appears that at least on first generation Toshiba HD DVD players, the player transcodes plain DD EX and dts ES 6.1 audio to dts 5.1. I can see why the player needs to transcode advanced audio such as DD+ or TrueHD since these are not supported in most receivers, but by transcoding legacy audio formats, you lose the rear surround information. Does this have to do with having to overlay player sounds (menu clicks, etc) in advanced menu? Would it be possible to have an option to bypass this overlay and have the player simply pass audio bitstream to the receiver for legacy audio tracks or overlay the menu sounds without losing additional channel information?

The ability to make legal copies should go a long way to reducing the desire to make illicit copies.

Perhaps they should be speeding up the final agreement and MC?

Gary

Everyone working on AACS feels as you do, on both points.

Insiders: Are we talking about "Managed Copy" or "Mandatory Managed Copy"?
As far as I know, HD DVD was always using the "Mandatory" word. Has that changed?

- Rich

No change. Both are important. Both the ability to make the copies in the first place (requiring technical solutions, significant infrastructure, and product rollouts), and the consumer garuntee of the right to do so.

So, if I interpret this correctly. There are multiple computers/threads processing consecutive 16X16 macroblocks at varying quantization levels with the purpose of comparing the results and selecting the particular quantization that provides the most "favorable" result for each individual macroblock? (Such a quantization "pass" could occur as a precurser to a luminance analysis that would be looking at potential luma-block assignments.) [/QUOTE}

Actually, each computer works on a big indepentant chunk (typically a chapter or more per worker). The details are a little different from what you said, but yes, there's basically a heuristic to determine which blocks need lighter compression compared to others (smooth gradients need to be compressed less, for example).

[QUOTE=TrevorS]I was thinking the encoded luma-blocks can be fully flexibly located, or are they actually restricted to being entirely contained within a sampled macroblock? I'm presuming the 16x16 macroblocks are a fixed matrix overlay that subdivides the original JPEG. Not allowing luma-blocks to cross macroblock boundaries would seem a natural fallout of the macroblock specific subsampled chroma blocks, as well as a possible analysis focus on one macroblock at a time.

Correct. Although note the source is uncompressed 4:2:0, not JPEG.

However, given a luma-block is inherently confined to a fixed luminance level (excluding deblocking filter action), recognizing a luminance match between adjacent macroblocks might be awkward. Further, if there is potential for a luma-block extending across macroblock boundaries, an analysis pass would have to be made that simultaneously encompasses up to four adjacent macroblocks -- possibly messy in a multi-cpu environment.

We don't thread down to the macroblock level, for the reasons you suggest.

Do any of my ramblings come close?

You're doing well.

FYI, if you're a SMPTE member, you can get a copy of the VC-1 spec.

This is functioning as a really good brain teaser for me, as well as an opportunity to learn something new. :)

Unfortunately, I'm not a member of any professional organizations (once a member of IEEE, but that was awhile ago.)

I was apparently thinking in terms of much too low a level of thread granularity. At the chunk/chapter level, each thread would be responsible for the complete differential quantization analysis of each macroblock of each 4:2:0 frame, and would specify/provide an individual quantization result for all encompassed macroblocks.

In addition, since all chunk/chapter macroblocks would be processed locally (within the current thread), there is no inherent reason why a given macroblock must fully contain a luma-block (except at frame boundaries, of course). There can be full independence of the quantization and luminance analysis passes. This would allow maximal flexibility in conforming the luma-blocks to the image (given the defined set of luma-block sizes). The luma-block chroma attributes would be governed entirely by the underlying macroblock boundaries.

I do wonder how it's determined whether the quantization for a given macroblock is "optimal". What "limits" any single test pass would be compared against (I guess I'm actually asking a question here). Further, I'm not sure exactly how perceptual encoding comes to play with visual information.

My experience with perceptual "masking" is limited to Dolby Lab's claims regarding Dolby Digital and it's fairly heavily compressed audio streams. I was very dissatisfied with their results as I heard them on DVD in the late 90's -- yet they claimed them to be indistinguishable from the original (yeah, right!) I was and continue to be a fan of LD 1.2Mbps DTS. Whatever their perceptual encoding scheme, it is very effective to my ears (likewise DVD 1.536Mbps). Howver, when DTS tried to say that 768Kbps was just as good, my ears didn't accept that at all -- it was very obviously a compromise necessary to coax at least some studios to include the cutback DTS on at least some of their releases. I still preferred it to DD at 448 Kbps, but no way was it DTS as I knew it.

When I listen to sound, I tend to hear the whole thing, but I also shift my focus to individual elements in the mix, perhaps a specific instrument or a voice. It seems to me that with video I tend to focus on individual portions even more. I think this is because I can only really focus on one relatively small portion of a screen at a time, everything else is somewhat peripheral. In fact, I feel my ability to evaluate/discriminate sound is actually FAR more capable than my ability to evaluate video -- I suspect that is a survival based fundamental coming from millenia in the past. Sound provides pretty detailed 360 degree coverage plus elevation/declination, whereas vision is extremely limited.

Some say that high quality audio is of little value in home theater -- the emphasis being on the visuals. However, I find the ability of the sound to engage and transport me, to be easily able to offset moderate limitations in the video (though it's certainly nice when both are cooking together :)).

Could you give some idea of what perceptual encoding is really looking at when it comes to video?

Thanks -- Trevor

Paid,
Hello there, could you tell me what subtitles the BR discs bought from Japan have please? I'm looking for BR discs with Chinese subs (for my better half:) ) and am hoping i can get them through the net.
Ta,
Will.

I know this is old news but was just wondering of any new developments?


LAS VEGAS - Jan. 8, 2007 - Today at the 2007 International Consumer Electronics Show (CES), Microsoft Corp. and Broadcom Corp. announced a joint effort to support a hardware and software reference design for more cost-efficient HD DVD playback.


"We've seen strong sales of HD DVD players and movies so far, but using this new platform, companies can produce HD DVD players in record time, resulting in more choices for consumers," said Amir Majidimehr, corporate vice president of the Consumer Media Technology Group at Microsoft. "The power and high integration of BCM7440, combined with the Windows CE 6.0 HD DVD platform, results in even lower cost players for consumers than the high value products already in the market."

The first players incorporating the new platform are expected from manufacturers in the second quarter of this year.

A question for paidgeek,

It was suggested a while back by yourself that Casino Royale would be a region free release. However, it seems it is not which is causing a bit of a problem in the UK.... Any reason for the change in coding policy for this title?

And as an extenstion to Ian-S's question, Sony do realise that regional encoding is a kiss of death for Europe and more so the UK?.

I favour blu ray myself but i know many who wont buy into anything that is region locked. this IS giving HD-DVD an advantage here and is seen a a HUGE plus point over BD. with DVD players they are always made availble to be region unlock and is seen as vital to sell dvd players. We can get our BD players unlocked for DVD but not for BD titles, and this is a huge mistake.

It would not surprise me if the BDA lose europe based on this issue alone. YOU HAVE BEEN WARNED :)

YOU HAVE BEEN WARNED :)
:eek: :eek: :eek:

I’ll give him a verbal heads-up and see if he can respond later today to the questions posted on this page.

What happened to my Gunners this year :( ?

:eek: :eek: :eek:

I’ll give him a verbal heads-up and see if he can respond later today to the questions posted on this page.


Not a question, but an enhancement of the same topic. The continental Europe BD editions of Hellboy, Vertical Limit, Big Fish and Layer Cake are all region free.

The Buena Vista's 1st wave of Euro releases are also region ABC according to their covers.

And as an extenstion to Ian-S's question, Sony do realise that regional encoding is a kiss of death for Europe and more so the UK?. I favour blu ray myself but i know many who wont buy into anything that is region locked. this IS giving HD-DVD an advantage here and is seen a a HUGE plus point over BD.

I can confirm that this is also the case here in Germany. Being completely region free is (besides cheaper standalone players) the number one reason a lot of early adopters/enthusiasts prefer HD-DVD right now.

Paid, if Warner and Paramount can release all their Blu-ray discs region free, there are apparently no legal/contractual reasons why Sony, Disney and Fox couldn't do the same, right? I really feel it would give Blu-ray a huge boost in Europe if these studios would re-consider their region coding policy.

http://www.dvdtimes.co.uk/content.php?contentid=63849

are reporting Casino Royale will be coded A, B, C. How does this fit with the policy of coding in the first year and then removal or is this an error?

Casino Royale has already been released in theaters worldwide. Since the theatrical window for the title is passed, we are coding the disc all region.

Yes I am keen to understand what has gone on with the regional coding for Casino Royale. There are a few 'miffed' customers in the UK who followed the advice from this thread that I was given. I have had now two 'angry' emails to deal with :( What changed? This is just the sort of stuff the UK doesn't need if we are to see BD develop.

Yes I am keen to understand what has gone on with the regional coding for Casino Royale. There are a few 'miffed' customers in the UK who followed the advice from this thread that I was given. I have had now two 'angry' emails to deal with :( What changed? This is just the sort of stuff the UK doesn't need if we are to see BD develop.Coded Region A/B/C is all-region and will play in all players. So where's the problem?

Many users are reporting to me it is not coded as A, B C. The only reason currently given. (See also the posts by Ian S and Gandley above). I have no way of testing a US disc in a UK player atm as my player is US based and my disc is 'still in the post'. If this is just a compatibility thing then we can deal with it as that, if it is a region thing then we can treat it accordingly. The first thing we need to check is the disc was coded as ABC. Then if that was the case, we can then work out for the 'distressed' users what the issue is. They are currently all describing it as a region thing.

Many users are reporting to me it is not coded as A, B C.
How many users? 10, 20, 100?

The first thing we need to check is the disc was coded as ABC. Then if that was the case, we can then work out for the 'distressed' users what the issue is. They are currently all describing it as a region thing.

What msg was displayed when they tryed playing the disc nick? Could it simply be a faulty batch of discs?
Always a risk importing discs for a new tech.

US Casino Royale is coded A and won't play in European region B players. Tested this today myself.

To Amir or other Microsoft insider:

Any word on the 'Babel' subtitle/fast forward issues for the 360 HD-DVD add-on? I rented a copy from Netflix and couldn't watch it without major problems and I found this thread on AVS with other users reporting problems...

http://www.avsforum.com/avs-vb/showthread.php?t=809426

Thanks.

I can confirm that this is also the case here in Germany. Being completely region free is (besides cheaper standalone players) the number one reason a lot of early adopters/enthusiasts prefer HD-DVD right now.

Paid, if Warner and Paramount can release all their Blu-ray discs region free, there are apparently no legal/contractual reasons why Sony, Disney and Fox couldn't do the same, right? I really feel it would give Blu-ray a huge boost in Europe if these studios would re-consider their region coding policy.


Region Coding is strictly for the benefit of the software distributors. Nothing else. Due to the way the film business works these days with multiple distributors globally, and some very staggered release dates, region coding enables them to control distribution.

For instance, Title A opens in U.S. theatres on March 9th. Australia is the same day, but most European markets get it in May and June. Italy not until September.
Japan gets it in August.

Said film gets a video release in the U.S. in July. It hasn't even open theatrically in some major markets yet. So, by Region coding, they can protect the theatrical market in other countries because those folks there can't actually play the U.S. region discs.

There is no "contractual" obligation with respect to the BDA or DVD forum to region code your discs. It's strictly to protect the distributors.

:eek: :eek: :eek:

I’ll give him a verbal heads-up and see if he can respond later today to the questions posted on this page.

What happened to my Gunners this year :( ?


Thanks, that would be good of you. TBH this is really a big deal, if we could unlock the BD players here in the UK to play all region discs just as we can with DVD that would solve the problem. obviously DVD is region encoded for us but all manufactures pass out firmware discs out the back door to allow multi region.

I dont want to go on about it, but the BDA is handing HD-DVD a golden ticket, and the BDA really dont want to do that.
I would suggest that the BDA allow manufactures to unlock there players for multi region use with BD discs. If not, well the result could speak for itself.

As for the gunners, like chelsea, they are a little off the boil. actually they suck this season :D .

But please sort this out BDA, or you will be caught by your own petard. Europe needs quick action.

Region Coding is strictly for the benefit of the software distributors. Nothing else. Due to the way the film business works these days with multiple distributors globally, and some very staggered release dates, region coding enables them to control distribution.

For instance, Title A opens in U.S. theatres on March 9th. Australia is the same day, but most European markets get it in May and June. Italy not until September.
Japan gets it in August.

Said film gets a video release in the U.S. in July. It hasn't even open theatrically in some major markets yet. So, by Region coding, they can protect the theatrical market in other countries because those folks there can't actually play the U.S. region discs.

There is no "contractual" obligation with respect to the BDA or DVD forum to region code your discs. It's strictly to protect the distributors.

Which we understand and all are discs are pretty much region encoded for DVD. but in Europe all manufactures provide firmware to unlock the machine for multiregion use (or the odd few that dont, get there players hacked, but normaly the manufactures will let the firmware discs slip out the back door so you can just flash the firmware or have code to press on the remote). this is pretty much the norm and so now is a requirment for many many people. HD-DVD obviously realised the importance of multi region in europe and made sure it was good to go, and on that alone people will choose HD-DVD, it is that important.

Thanks, that would be good of you. .
I actually just passed on your query today to him..........he's just super busy :eek: .

I assume that some of you guys are mods or what not on AVforum? in England.....i.e. Ian, Nick ???????? those names ring a bell.

Just a word to state that England people are not the only one really annoyed by this region thing...

I thought we (France) were the most annoying country with this video releases (Need to wait six month after theatrical release date)... Looks like we are not alone ;)

This is realy annoying especially at that time where early adopters (most cinephiles) simply want movies... !
So far, beside the lack of entry level BD player (don't want a game machine) it is the only thing that stop my Master Card to become neutral :D

Any insiders inputs on these real shame ?

--Patrice

I actually just passed on your query today to him..........he's just super busy :eek: .

I assume that some of you guys are mods or what not on AVforum? in England.....i.e. Ian, Nick ???????? those names ring a bell.Yeah we are long time members and uber av fans who do are best to help others in England.

We see Blu ray is really getting some stick for region encoding, or more so no region fix on players. those that do buy into blu ray often post somthing like this

"P.s. I have gone dual format and don't regret it at all.BD region coding is my only regret!"

Keep the discs region locked but let the players be altered via firmware as they always have allowed for DVD. PLEASE :) OR let HD-DVD win

Keep the discs region locked but let the players be altered via firmware as they always have allowed for DVD. PLEASE :) OR let HD-DVD win

Couldn't agree more. What does the Blu-Ray insiders have to say about this?

moderator

have we forgotten Insiders rules so quickly?

Post Questions [only questions] directed to and answered only by Industry Insiders who are asked to identify themselves as such in their sig : subject to AVS approval

Industry Insiders only may answer questions or make comments: this is the thread for chat between Insiders as well

any AVS member can post questions [only questions please] for Insiders- but we will not tolerate any bashing

AVS recognizes the special nature of Industry Insiders and values their participation: we ask all AVS members to treat them with respect

Remember: Questions only: no off topic posts: they may be removed: and only Insiders [who have been recognized by AVS moderators] may answer

Insiders: please PM Ken H and myself with your credentials: mods and only mods will make the determination of who qualifys as an insider and all such discussions will be done off line: not to be discussed here

This is a continuation of the original thread Industry Insiders Q&A Thread: only Questions to insiders please which did not end well: any more of this and we will take strong action: Insiders are to be treated with respect:

Please help us make this a flagship thread on AVS by observing these rules

have we forgotten Insiders rules so quickly?


Sorry :o.

Dear microsoft insiders, can you guys tell me if 300 is coming to hd dvd mainly cause i just saw the movie and it is SPECTACULER JUST SO GOOD, so if you could tell me it would be great.

Dear microsoft insiders, can you guys tell me if 300 is coming to hd dvd mainly cause i just saw the movie and it is SPECTACULER JUST SO GOOD, so if you could tell me it would be great.

Certainly expect to see it in HD DVD. It should be a spectacular title as you say.

(P.S. Sorry about not responding earlier to this thread. Been busy with other things, including our AVS Forum meet in Vancouver. Hope to catch up to the other questions soon.)

A question for paidgeek,

It was suggested a while back by yourself that Casino Royale would be a region free release. However, it seems it is not which is causing a bit of a problem in the UK.... Any reason for the change in coding policy for this title?

My apologies.

I mistakenly thought every version was going out coded for all regions. The NA version was already completed as region A only. All other versions worldwide are coded region ABC. I'm truely sorry for an inconvenience this caused to those who purchased the NA version for use in other territories..

Region Coding is strictly for the benefit of the software distributors. Nothing else. Due to the way the film business works these days with multiple distributors globally, and some very staggered release dates, region coding enables them to control distribution.

For instance, Title A opens in U.S. theatres on March 9th. Australia is the same day, but most European markets get it in May and June. Italy not until September.
Japan gets it in August.

Said film gets a video release in the U.S. in July. It hasn't even open theatrically in some major markets yet. So, by Region coding, they can protect the theatrical market in other countries because those folks there can't actually play the U.S. region discs.

There is no "contractual" obligation with respect to the BDA or DVD forum to region code your discs. It's strictly to protect the distributors.

Allow me to respond to your post.

It is correct to say that the content companies have asked for region coding in order to protect their theatrical exhibition windows. That said, your local theater owner also places some value on region coding because if a BD title is being trans-shipped from the U.S. to a country that has not yet seen said title in theaters, it is not good for the their business. In the absence of region coding, staggered theatrical windows will still exist, but to manage them, BD releases may not be available in the U.S. or anywhere else until the major theatrical windows have been served. Accordingly, whereas many customers appreciate the fact that packaged media release timing is closer than ever to theatrical release date, we may end up going back to 4 or 5 month schedules before the titles are released. We know from experience, and sales results, that everyone wants the titles as soon as they can possibly be released, so please understand the need for the studios to utilize region coding.

Paidgeek,

Do you know if Sony will release any of the Zorro movies to Blu-Ray This year?

Asked this question in another forum and no answer so it seems its must be asked to the insiders here.

Does any of the format

1. Import subtitles to purchased discs?

2. Let me import my own written subtitles?

Paidgeek,

Do you know if Sony will release any of the Zorro movies to Blu-Ray This year?

I have not seen any on the schedule yet... This is always subject to change...

Asked this question in another forum and no answer so it seems its must be asked to the insiders here.

Does any of the format

1. Import subtitles to purchased discs?

2. Let me import my own written subtitles?

Blu-ray can support external subtitle data being linked to a purchased disc, but we have used this feature yet.

There is no provision for creating your own subitle files...

Blu-ray can support external subtitle data being linked to a purchased disc, but we have used this feature yet.

There is no provision for creating your own subitle files...
The external subtitle data should be store on the studio web site.
Is the downloading or linking of these external subtitle data constrained by the region coding?

Paidgeek,

A few weeks ago I asked you if there are some good Sony catalogue titles soon to be announced by Sony. I've been a little dissapointed by Sony's latest list, they seem to have dried up lately (as well as other studios).

Where's Spiderman, Bad Boys, Starship Troopers etc?

Is there some problems with pressing the volume of discs at the moment (which has caused no new announcements to arrise).

Wouldnt it be easy to implement so that a movie checks for subtitles from a USB flashmemory unit for ex? There will no doubt be loads of releases individual for various countrys wich will never see a release of subtitles for my specific language, Im thinking of Japanese and French releases in particular, and having the potential to use "fansubs" would be a fantastic thing.

Also a question wich I dont know if anyone can awnser, but its worth a try.
Will there be a lot of TVseries released on any HD format, TVseries wich wasnt recorded in HD or on film, or wich only later seasons got recorded in HD or film? Having yet to collect some TVseries I wonder if its any use waiting or should I just continue looking at the DVDs?

Paidgeek,

A few weeks ago I asked you if there are some good Sony catalogue titles soon to be announced by Sony. I've been a little dissapointed by Sony's latest list, they seem to have dried up lately (as well as other studios).

Where's Spiderman, Bad Boys, Starship Troopers etc?

Is there some problems with pressing the volume of discs at the moment (which has caused no new announcements to arrise).

There are many highly anticipated titles being released this year, just not necessarily every one you are looking for.

The external subtitle data should be store on the studio web site.
Is the downloading or linking of these external subtitle data constrained by the region coding?

There is no constraint by region code. It would be good to know what languages are not currently provided on the disc that our customers are looking for. If another thread is started on this topic, we will study the results.

Allow me to respond to your post.

It is correct to say that the content companies have asked for region coding in order to protect their theatrical exhibition windows. That said, your local theater owner also places some value on region coding because if a BD title is being trans-shipped from the U.S. to a country that has not yet seen said title in theaters, it is not good for the their business. In the absence of region coding, staggered theatrical windows will still exist, but to manage them, BD releases may not be available in the U.S. or anywhere else until the major theatrical windows have been served. Accordingly, whereas many customers appreciate the fact that packaged media release timing is closer than ever to theatrical release date, we may end up going back to 4 or 5 month schedules before the titles are released. We know from experience, and sales results, that everyone wants the titles as soon as they can possibly be released, so please understand the need for the studios to utilize region coding.

Paidgeek,

I think that most people understand the need to protect local industries with region coding. However, it's difficult for a customer to accept that this policy may change from studio to studio. Like The Departed, for instance.

So my question is actually a plea:

We're at a phase where Blu-ray has to grow its installed base, and where the market has yet to start and make numbers in most areas of the world. I'd say that at this precise moment, a transnational sale doesn't really hurt a local market, but it could help convert more people to BD and ultimately turn them into faithful customers in their local market.

Could Sony take into account these considerations and try to be "less pragmatic" in their region coding approach, at least for a limited grace period?

There is no constraint by region code. It would be good to know what languages are not currently provided on the disc that our customers are looking for. If another thread is started on this topic, we will study the results.
Thanks.

Could external subtitle data be downloaded to a local hard disk (i.e. PS3’s hard disk)?

or

Should external subtitle data be downloaded only to the player’s persistent memory when watching the related movie?

Amir,

Sony PlayStation 3 is HDCP-encrypted even on gaming content, which is a puzzling state of affairs to be sure as it effectively locks out owners of older widescreen DVI displays.

Hypothetically, would Microsoft only enable HDCP when playing movie content on any of its gaming platforms? Could you foresee a digital output also supporting the same custom resolutions as, say, those supported with VGA on Xbox 360?

There are many highly anticipated titles being released this year, just not necessarily every one you are looking for.

Hey paidgeek, I was recently watching "Nothing But Trailers" on HDNet, and enjoyed a trailer for "Doublure, La" or "The Valet" which is listed as being a Sony Pictures Classics release here in the US. It's supposed to have a limited run in theaters in the US this April.

The target audience for this film is probably not big enough for a Blu-Ray release, but I feel I have to ask if there's any hope for this film to be released domestically?

~Alan

Hello,

Does any insider know if TMNT (a Weinstein film that is co- distributed by Warner) Will be released on Blu ray? Thanks so much for any help you might be able to offer! :)

Can any insider tell me if it might be on the consideration table of any studio to start releasing some new releases in HD formats BEFORE the DVD release?

I keep thinking this would be a great way to start enticing people to upgrade to BD or HD DVD. What are the ramifications in doing it? Is it even feasible?

Thanks in advance!

Amir, any news on the 24hz playback option for the Xbox 360? Here in Europe the jadder issue seems to be a significant constraint to the HD format acceptation in general. I'd like to hear your opinion on that matter.

Thanks in advance.

Amir, any news on the 24hz playback option for the Xbox 360? Here in Europe the jadder issue seems to be a significant constraint to the HD format acceptation in general. I'd like to hear your opinion on that matter.

Thanks in advance.
Nothing to announce on this front specifically. However, I did learn about another alternative last weekend. We had our usual AVS Forum meet where we demo'ed HD DVD on a 1080p projector, in this case, a Sony Pearl. To our surprise, we saw no motion judder. It seems that the Pearl is able to take the 1080i, perform inverse telecine, and then display the image at the proper 24fps frame rate! For example, the pans in Batman Begins were smooth as silk, matching our VP-50 processor that we use in HD DVD demos.

So kudos to Sony for doing the proper processing in their projector. If more displays do this, the whole notion of 24p output can become moot and people even with current players, can watch judder free images. And that is the beauty of 1080i in a way in that it does not lock in judder like 1080p60 might.

Paidgeek,

Are you guys currently working on Bad Boys and Bad Boys II?

I'd love to see these movies get released on Blu-Ray!

Thanks.

Nothing to announce on this front specifically. However, I did learn about another alternative last weekend. We had our usual AVS Forum meet where we demo'ed HD DVD on a 1080p projector, in this case, a Sony Ruby. To our surprise, we saw no motion judder. It seems that the Ruby is able to take the 1080i, perform inverse telecine, and then display the image at the proper 24fps frame rate! For example, the pans in Batman Begins were smooth as silk, matching our VP-50 processor that we use in HD DVD demos.

So kudos to Sony for doing the proper processing in their projector. If more displays do this, the whole notion of 24p output can become moot and people even with current players, can watch judder free images. And that is the beauty of 1080i in a way in that it does not lock in judder like 1080p60 might.

Isn't the VP50 the Sony Pearl, I thought the Ruby was the high-end VP100. Does the Ruby even do IVTC properly, I thought it had trouble with 1080i sources, though I may very well be wrong (I don't own one, unfortunately :()

Isn't the VP50 the Sony Pearl, I thought the Ruby was the high-end VP100. Does the Ruby even do IVTC properly, I thought it had trouble with 1080i sources, though I may very well be wrong (I don't own one, unfortunately :()
You are half right in that I meant to say Pearl, not Ruby (edited above). The model number is VPL-VW50. VP50 is an outboard processor which we use for conversion from 1080i to 1080p48 for our Marantz.

Amir,

Sony PlayStation 3 is HDCP-encrypted even on gaming content, which is a puzzling state of affairs to be sure as it effectively locks out owners of older widescreen DVI displays.
Oh, I didn't know this. Strange that there would be some concern on capturing gaming output.

Hypothetically, would Microsoft only enable HDCP when playing movie content on any of its gaming platforms? Could you foresee a digital output also supporting the same custom resolutions as, say, those supported with VGA on Xbox 360?
Is there a specific resolution you are concerned about here that might not be supported in the future?

Amir, welcome back to the forum!

Any answer to this question?

Is the upcoming firmware update for the 360 HD-DVD player going to fix the inability to play the HD-DVDs "Discovery Atlas: Australia Revealed" and "Discovery Atlas: Italy Revealed"?

Sony PlayStation 3 is HDCP-encrypted even on gaming content, which is a puzzling state of affairs to be sure as it effectively locks out owners of older widescreen DVI displays.

Anything output HDMI will be HDCP. AFIAK it's required for the devices to handshake.

Paidgeek, now that Sony has "officially" said Lawrence of Arabia & Spider-Man 2 are on the way for Blu-ray (via Casino Royale's preview) has any mastering/encoding work been done as yet?

Hello Armir,

I think tomorrow the CEBIT starts. Wouldn't that be a great opportunity to announce some news about Microsoft... maybe about the xbox.... maybe about the DTS update? ;)

I know, I know, in spring... but spring is such a great time span. Will the update really hit us with surprise? Like one day it suddenly pops up and there it is. I would prefer to know a couple weeks ahead that (an exactly when) it is coming so I can look forward to it. An old Star Trek wisdom says: desiring something is much more pleasent than having something. :)

Greetings,
Caurus

Oh, I didn't know this. Strange that there would be some concern on capturing gaming output.

Is there a specific resolution you are concerned about here that might not be supported in the future?

Hi Amir,

I simply feel that with Xbox 360 and its VGA support, Microsoft has gone to great lengths in providing 1:1 pixel mapping to all plasmas, consumer LCD panels and many PC monitors. It would be a shame if that did not extend to any Microsoft gaming devices which may support a digital output in the future. It would equally be a shame if HDCP were enabled for gaming output as many superb monitors would be incompatible - as they are with PS3.

PeterTHX - development PS3 debug stations can run with HDCP disabled, making them compatible with non-HDCP DVI displays so your assertion that HDCP is required for digital handshaking appears to be incorrect.

Paid,
Hello there, could you tell me what subtitles the BR discs bought from Japan have please? I'm looking for BR discs with Chinese subs (for my better half:) ) and am hoping i can get them through the net.
Ta,
Will.
Pretty pls?
Ta

Amir,
Hi there. Eny news on what else might be in the 2.1 update for the Tosh 1st gen players? If i recal one of your team was on the testing side when 2.0 was ready. I'm hoping they unlock R3 NTSC playback, what are the odds?
Ta.

i don't know if this has any bearing on the region coding, but I was in London last month and Casino Royale was not only still playing in the theaters, it was sold out.

Anything output HDMI will be HDCP. AFIAK it's required for the devices to handshake.

Paidgeek, now that Sony has "officially" said Lawrence of Arabia & Spider-Man 2 are on the way for Blu-ray (via Casino Royale's preview) has any mastering/encoding work been done as yet?

We have not started LOA yet, but work may be underway on SM2.

Paidgeek,

I think that most people understand the need to protect local industries with region coding. However, it's difficult for a customer to accept that this policy may change from studio to studio. Like The Departed, for instance.

So my question is actually a plea:

We're at a phase where Blu-ray has to grow its installed base, and where the market has yet to start and make numbers in most areas of the world. I'd say that at this precise moment, a transnational sale doesn't really hurt a local market, but it could help convert more people to BD and ultimately turn them into faithful customers in their local market.

Could Sony take into account these considerations and try to be "less pragmatic" in their region coding approach, at least for a limited grace period?

I think Sony Pictures is taking a more liberal approach than most in terms of region coding restrictions. Put yourself in the position of a theater owner, a big movie is scheduled to play in your auditorium and you have agreed to show it on particular screens, then before you even receive your prints, people are telling their friends about how great it was on the Blu-ray they bought overseas and would they like to borrow it. Not so good...

Theatrical exhibition is an important part of the life of a film and it is the biggest part of what the director and everyone else on the project works for.

Amir,
Hi there. Eny news on what else might be in the 2.1 update for the Tosh 1st gen players? If i recal one of your team was on the testing side when 2.0 was ready. I'm hoping they unlock R3 NTSC playback, what are the odds?
Ta.
Unfortunately, I don't have any details on what they fixed. The only thing I know is that it improves "disc readability". If I find out more, I will report back although I am sure people here will figure it all out before I do :).

I think Sony Pictures is taking a more liberal approach than most in terms of region coding restrictions. Put yourself in the position of a theater owner, a big movie is scheduled to play in your auditorium and you have agreed to show it on particular screens, then before you even receive your prints, people are telling their friends about how great it was on the Blu-ray they bought overseas and would they like to borrow it. Not so good....
While this is a laudable goal, I have to chime in with the views of other content owners who dispute it. The scenario which you paint is a dream which simply is not backed by realities of the technologies designed in these formats. The DVD region coding was highly ineffective due to bypass methods widely available from even major CE company products (did Sony DVD players not have a bypass mechanism for Europe?). The BD version significantly waters the region selection down, making it even harder to limit playback in specific countries where such theater owners might have these concerns.

More puzzling is that while BD reduces the number of regions as compared to DVD, it has made the implementation much more robust against hacks/bypass. What you gave on one hand, you took away with the other!

Besides, I would think that Theater owner’s bigger problem is the camcorder’ed version of the same movie being available way, way before it comes to BD/DVD (i.e. a day after it opens in US). That impacts their business far more than any cross boarder importation of region free high def movies.

Net, net, while I agree there are some real business concerns here, region coding in the form of BD/DVD did not solve those problems and instead, simply serve to annoy our customers. This view is shared by many and hence the reason HD DVD leaves out region coding…

This view is shared by many and hence the reason HD DVD leaves out region coding…

And also why Fox doesn't support HD DVD.

Amir,

When did Microsoft start distributing theatrically released films? Or are you speaking for Warner/Paramount/Universal?

And also why Fox doesn't support HD DVD.
That is not the reason Fox doesn't support HD DVD. At the time they chose BD, there was still time to add region coding to HD DVD spec.

Amir,

When did Microsoft start distributing theatrically released films? Or are you speaking for Warner/Paramount/Universal?
I am conveying the views of some other studios per my post. If all studios in DVD Forum pushed for region coding, there would be region coding in HD DVD. After all, that is why we have mandatory new audio codecs in HD DVD despite objections of other companies...

I am conveying the views of some other studios per my post. If all studios in DVD Forum pushed for region coding, there would be region coding in HD DVD.Which major studios do support region coding? Also you say that region coding wasn't added to HD DVD since not enough companies wanted it but why than was a technology like FGT added to HD DVD? Wasn't FGT added simply to get Thomson to support HD DVD?

Which major studios do support region coding?
Presumably the ones already using it in BD format :p.

I think you are confusing usage and "support." It is one thing to use region coding after the product is out with that capability, but an entirely different matter convincing the rest of the companies in a standards setting organization to adopt it. For the latter, you have to make a good case for it to convince the majority voting (we know, that is what we had to do with VC-1). If there are companies similarly situated to you who don't agree, then it becomes a tough challenge and you are unlikely to prevail.

Also you say that region coding wasn't added to HD DVD since not enough companies wanted it but why than was a technology like FGT added to HD DVD? Wasn't FGT added simply to get Thomson to support HD DVD?
Majority of the DVD Forum board voted to adopt FGT (not just the companies building HD DVD products). Since the BDA companies hold majority position in DVD Forum board, your are saying that BDA companies adopted FGT as to cause Thomson to support HD DVD :p.

Amir,

This might have been asked before, but I can't find it.

Since last summer, nearly half of the theaters in a 20 mile radius from me are digital projection theaters. I will only go to those theaters now. But I've been really wondering what codec may be used in them. Can you shed any light on what codecs are used in theaters? Do any of them use VC-1?

That is not the reason Fox doesn't support HD DVD. At the time they chose BD, there was still time to add region coding to HD DVD spec.

Can you dig into what their reason for not supporting HD DVD is, then? From my tracking, Fox is the only studio to use Regional coding on all their titles, while the other Blu-ray only studios (Sony, Disney, and Lionsgate) have some titles with it and some titles without it. If they didn't believe it to be a necessity, why do they so consistently use it?

Amir,

This might have been asked before, but I can't find it.

Since last summer, nearly half of the theaters in a 20 mile radius from me are digital projection theaters. I will only go to those theaters now. But I've been really wondering what codec may be used in them. Can you shed any light on what codecs are used in theaters? Do any of them use VC-1?
VC-1 gets used in art houses but not mainstream theaters. For the latter, Ben probably knows more than I do. Last I heard they were using high data rate, I-frame only compression but my information is stale.

Can you dig into what their reason for not supporting HD DVD is, then? From my tracking, Fox is the only studio to use Regional coding on all their titles, while the other Blu-ray only studios (Sony, Disney, and Lionsgate) have some titles with it and some titles without it. If they didn't believe it to be a necessity, why do they so consistently use it?
I can't disclose anything that is not public. Suffice it to say, they appear to highly value the technology that is behind BD+. They announced strong support for BD+ as soon as it was ratified by BDA for example, some 2 years ago. And that was the same time they announced support for BD format. Yet they did not make a big deal out of BD having region coding, and HD DVD not.

On your second question, I have no insight. I just know that it was not a consideration as far as their choice for formats. Maybe if they had adopted HD DVD, they would have then pushed for region coding after the fact. I have heard, but do not know for sure, that BD implements region coding using BD+. If so, they may support region coding for that reason. But again, I have no idea here and just making random guesses :).

Can you dig into what their reason for not supporting HD DVD is, then?

BD+ and ROM mark

Paidgeek: in authoring a BD disc, does BD+/ROM mark have to be encoded or is it just a flag?

I can't disclose anything that is not public. Suffice it to say, they appear to highly value the technology that is behind BD+. They announced strong support for BD+ as soon as it was ratified by BDA for example, some 2 years ago. And that was the same time they announced support for BD format. Yet they did not make a big deal out of BD having region coding, and HD DVD not.

On your second question, I have no insight. I just know that it was not a consideration as far as their choice for formats. Maybe if they had adopted HD DVD, they would have then pushed for region coding after the fact. I have heard, but do not know for sure, that BD implements region coding using BD+. If so, they may support region coding for that reason. But again, I have no idea here and just making random guesses :).


Does Fox have patents on BD+?

This might have been asked before, but I can't find it.

Since last summer, nearly half of the theaters in a 20 mile radius from me are digital projection theaters. I will only go to those theaters now. But I've been really wondering what codec may be used in them. Can you shed any light on what codecs are used in theaters? Do any of them use VC-1?
My belief is that as per adherence to all the specifications that Digital Cinema Initiatives, LLC* has outlined…….digital theatrical exhibitors have the feasibility to play back JPEG 2000-encoded content as well as the MPEG format.

*Digital Cinema Initiative (DCI) includes Universal, Disney, Fox, MGM, Paramount, Sony Pictures Entertainment and Warner Brothers Studios, did I leave out anybody ?

"DCI’s primary purpose is to establish and document voluntary specifications for an open architecture for digital cinema that ensures a uniform and high level of technical performance, reliability, and quality control. DCI will also facilitate the development of business plans and strategies to help spur deployment of digital cinema systems in movies theatres."

But alas, this info may also be a little stale, I suggest you google Digital Cinema Initiatives and their latest draft may be online.

Does Fox have patents on BD+?
From what I know, BD+ is owned and licensed by Sony+Matsushita (Panasonic). The core technology below it though, is from CRI. CRI is a start-up and I don't know their ownership situation. So unless Fox has some interest in CRI, the answer is probably no.

Amir,

Any chance of seeing the Illusionist on HD DVD anytime soon?

In a way yes. By adding these sampling rates, you get the same suite as DVD-A because TrueHD is the same as MLP used in DVD-A compression. But in a way, it is better in that DVD-A had an expensive (to license) navigation system. In the case of HD DVD, one simply makes an audio-only playlist (my preference) and you have a nice CD-type of experience which doesn't cost a penny more in licensing costs. And of course, you can also go crazy with HDi menus and such and blow away anything DVD-A could (not my preference).

Having said this, my personal opinion is that the time has come and gone for any kind of mass market around high-resolution optical formats. I much rather download the same files over the internet when I want it, than to get stuck with a few specialty retailers stocking a couple of discs here and there. And most of the world could care less about any improvements over CD it seems.

amirm, to play DVD-A we use a dedicated or universal player. Will HD DVD-Audio play on any HD DVD player with TrueHD decoder?

amirm,
Do you know anything about the 2nd Wave of Studio Canal releases for HD DVD being Delayed? Any info would be appreciated Thanks!

VC-1 gets used in art houses but not mainstream theaters. For the latter, Ben probably knows more than I do. Last I heard they were using high data rate, I-frame only compression but my information is stale.
Folks like Living Room Theaters just to normal WMV9 Advanced Profile, with IBP encoding. The v11 codecs we released in the fall gave a very nice quailty boost for that market. Lots of them are on 1280x1024 projectors, so you can do a 1280x720 12 Mbps encode that looks incredible, and can be digitially delievered.

Ding, ding! Round two on AVC vs VC-1 :). I finally have some free time to come back to Ron’s defense of AVC loop filter. For those who are joining us now, here is Part 1: http://www.avsforum.com/avs-vb/showthread.php?p=9931723&&#post9931723 And Part 1(a) is here: http://www.avsforum.com/avs-vb/showthread.php?p=9941753&&#post9941753

Here's my first attempt at showing the effects of the H.264 loop filter.
First, I have to thank you Ron on two fronts. One, for thinking of this visual representation to talk about a topic that is very complicated. I was really struggling on how to explain the details behind the complicated logic in AVC filter. Then I saw your simulation using a white pixel in place of a filtered one and that gave me a much better tool to respond. Second, I hope people appreciate what a great forum this is, as to motivate experts in the industry to put aside their day job, and go and modify a decoder and add such logic to it, simply to educate people (OK, so there is ulterior motive to defend one’s opinion but bear with me on this :) ).

Let me start by using words to explain something and then we go to pictures. In compression circles, as Ron did in his response, we use the term “Quantization” or “Qp” as a measure of how much we compress a block of pixels. The encoder dynamically changes Qp as it examines the complexity of the source against the available bit budget. Note that there is simple, “linear” relationship between Qp and visual artifacts (or we would not need humans to optimize these encodes). But Qp is a way to objectively talk about compression of video.

AVC uses a log scale of 1 to 51 for its Qp factor. VC-1, like MPEG-4 Part 2, uses a linear scale of 1 to 31. In both cases, 1 means essentially no compression, and maximum scale means just that, maximum compression. Ron suggested that we look at the case for Qp of 17 for AVC which is not a bad number for the kind of Qp often seen in HD DVD/BD encode so I will do the same. The same Qp roughly corresponds to “2” in VC-1. So we use that there for our codec.

I have gone further than Ron in providing this simulation to you in a few areas:

1. We use a full, 1920x1080 (i.e. 1080p) frame instead of reduced resolution 1440p frame that Ron used. This better reflects the kind of source that is used for HD DVD/BD, as opposed to 1440p which may be sourced from HDCam or fit for lower resolution broadcast.

2. We used the actual frame from movies encoded for HD formats to make this a representative sample. As such, we had to get kind permission of the studio holding the rights to the movies. To make this easier and not pass around a hat for donations (:)), we had to pick a non-descript scene with no actors, etc. The photographer in me is disgusted with the choice of the frame :), but it serves nicely for this purpose.

Since one can easily be accused of hand selecting favorite frames to prove one point or the other, I am showing you two different frames. We have a lot more, and they all reinforce the same point, some even more powerfully than these examples. Of course, if you don’t believe me, then you should not believe Ron either and assume he picked a frame that was favorable to his case :).

3. We changed the simulation to show the effects of a pixel being filtered multiple times, which we believe is maybe a bigger issue than even the larger number of pixels that AVC filters. Here, we use the color of red for a pixel which is being pounded on more than once. White pixels are used when a pixel is only filtered once. So red pixels are worse than whites.

4. We modified our VC-1 decoder to perform the same function. We think it is important to compare the two technologies side-by-side, as opposed to looking at one side, and say, “hey, this is not so bad.” I think it is in this contrast that you see the benefits of VC-1.

5. We also modified the respective decoders to output useful statistics such as number of pixels filtered once, or more. While the picture does prove the point easily, it is still good to have some hard stats to perform some precise comparisons.

OK, with the introductions out of the way, let’s look at the results in the next post.

So here is the part 2 of AVC vs VC-1 loop filter simulation.

Here is the original video frame. http://la.gg/upl/Clip1.jpg
Note that we have compressed it in JPEG to reduce bandwidth needed.
So please don’t examine it for quality comparisons as much as getting an idea of what the original looks like:

Now let’s look at the VC-1 results (also in JPEG so please ignore the compression artifacts -- the block dots next to white ones).
As you see, despite the mild level of compression and a frame that looks easy to encode, there are some block edges which need to be touched.
The reason is that the frame is not as easy to encode as it seems. And keep mind that filter strength can be variable.

http://la.gg/upl/Clip1_vc1_qp2.jpg


Now let’s look at AVC frame:
http://la.gg/upl/Clip1_AVC_QP17.jpg

Wow, what happened there :)? Not a pretty picture it seems. Let’s drill into statistics here:

VC-1 pixels filtered: 54234 (3%)
AVC pixels filtered: 326148 (16%)
VC-1 pixels filtered more than once: 54 (0.00%)
AVC pixels filtered more than once: 115181 (5.51%)
Ratio of AVC to VC-1 pixels being filtered: 6.01:1
AVC pixels touched twice: 102594
AVC pixels touched three times: 10686
AVC pixels touched four times: 1901
Ratio of AVC pixels touched more than once compared to VC-1: 2133:1

As you see, AVC filters 6X more pixels in this example than VC-1 which is even worse than the case I presented before.
Worse yet, while VC-1 hardly touches a pixel twice, that seems to be more of a norm with AVC,
with a whopping 2133 times more pixels touched that way, than in VC-1.

With some 16% of the image now filtered in AVC, you are very likely to see its softening effect.
In contrast, you will have a heck of a time with VC-1 given the fact that it has only touched 3%.

Now, let’s look at a second test case. Because the images are large, I am just going to provide the links.
Again, here is the original in JPEG: http://la.gg/upl/Clip2.jpg

The VC-1: http://la.gg/upl/Clip2_vc1_qp2.jpg

And AVC: http://la.gg/upl/Clip2_AVC_QP17.jpg

The stats are:

VC-1 pixels filtered: 56454 (3%)
AVC pixels filtered: 149143 (7%)
VC-1 pixels filtered more than once: 138 (0.01%)
AVC pixels filtered more than once: 51739.00 (2.48%)
Ratio of AVC to VC-1 pixels being filtered: 2.64:1
AVC pixels touched twice: 49324
AVC pixels touched three times: 2025
AVC pixels touched four times: 390
Ratio of AVC pixels touched more than once compared to VC-1: 375:1

Here, AVC does better in that its ratio of pixels filtered is “only” 2.64 times higher than VC-1. And its ratio of multiple filtering is 375:1.
None of this is anything to be proud of though as this still translates to a softer image in AVC.

Amir likes to characterize the H.264 loop filter as a "blunt eraser".
Looking at the numbers above, I would say that filtering of anything but the block edges is done with a "light touch".

Ron
Hopefully, I have shown that AVC loop filter is anything but “light touch.”
I first described the algorithm at high level, and now we have empirical data to back the same.
And please note that what I have described here is what we see in the field competing against AVC.
Both the algorithm and simulations back what we see in real life, with real HD DVD/BD content encoded in the two formats.

Where does this leave us? The three codecs rank this way: MPEG-2 uses large and visible blocks (8x8).
With no loop filter to help it, and much less efficient algorithm elsewhere,
you are going to see blocking artifacts when MPEG-2 is under stress (or "false sharpness" when it is not).
AVC overcompensates to the other extreme, ruthlessly filtering block edges.
Result is a smoother image, but lacking in very fine detail as opposed to VC-1.
VC-1 uses smaller block sizes like AVC and sports some of the same efficiency advantages of it over MPEG-2.
But its light touch gives just the right amount of distortion mitigation to produce an image that is rich in detail, while not marred by artifacts.

Again, I like to thank Ron for his insightful suggestion on how to visualize these concepts.
Hopefully everyone has a somewhat better understanding of these technologies
and can discuss things using the algorithm behind the codecs as opposed to other subjective evaluations that are hard to defend.