Industry Insiders Q&A MASTER THREAD [separate thread for Xbox/Add On & PS3]

[RealEstateWagon]

Amir/Ben

I have yet to experience HD video from the Xbox Video Marketplace where I live, although I've heard nice things about it. Still, I wonder if Marketplace must use well-known codecs only or if you are able to experiment with newer iterations of WMV that go beyond VC-1?

[amirm]

Originally Posted by Head Shot

Amir,

For the upcoming evolution of HD disks to go through the process of approval, which of the two ( TotalHD and Triple Layer 51) will have a better chance ?

I don't think TotalHD would be going through the DVD Forum process. There is no way DVD Forum would be in a position to approve the BD side in it. Historically, the forum has taken a dim view of such things. For example, there used to be a twin disc format with CD on one side, and DVD on the other side (promoted by the record companies). It was too thick and violated DVD spec so there were a lot of fireworks in DVD Forum about it. I don't expect such a contention here but the resistance will definitely be there. I will double check with our forum experts and report back if I hear anything different.

TL-51 can of course move through the forum. How easy it would be to do that, depends on the spec. I have not seen the TL-51 specifications so I have no real data to give you. But the forum has approved many variations of HD DVD format already so there is pretty good structure for evaluating and approving it.

[H9K_]

Originally Posted by paidgeek

I would be interested in your further explanation of why you very much like something that you cannot decode. Please PM me and we can discuss...

Because we can do that in near future and plan to re-watch films many times as the years go by.
Seems some one needs to start looking farer ahead than the tip of their nose.
Why not put the best track you got on it at now?

[amirm]

Originally Posted by dialog_gvf

The BD-25 yield INCLUDES the cover and protective coating yield reductions. You can't put it through as a layer yield.

Not quite Gary . By your logic, the second layer would have the same yields as the first one which we know, and you seem to agree, it does not. Reason is that there is a fixed tolerance for all the layers of the cake in a BD disc. When you have a single layer disc, you get to have more freedom in how flat and perfect the layers need to be. However, when you add layers, the job become progressively more difficult because the lower layers rob you of your available tolerances. I don’t recall the exact number but I believe the flatness is in the order of 8 microns. So let’s use this to walk through an example:

Let’s say the first layer is flat within 4 microns and the cover layer is within 2 microns. Add these two together and you have 6 microns of inaccuracy which is still within 8 micron total allowable and a single layer disc meets the spec requirements. But now, you only have 2 microns left. The second layer now needs to be flat within these 2 microns, not 4 as was the case with the first layer. So your yields drop drastically as if you could make layers that flat, you would already be doing it for the first layer . To compensate, you may now need to have a 1 micron tolerance for the cover layer which drops the yields for that layer. If you did this, then you have 4 microns for the first layer, 3 microns for the second, and 1 micron for the cover layer. As you see, the last two steps are harder to manufacture now and you have to deduce some yield loss because of it. This is why you can’t just use the same cover layer yields as you did for BD-25.

The above is the reason there are no combo BD discs. On paper, one can dream 0 micron, perfectly flat layers. But in reality, sticking a DVD under BD layers with its high tolerances plays havoc with flatness you need to maintain for the upper BD layers. Put another way, the DVD layer underneath the BD layers uses up all of your tolerance allowance, making the discs possible on paper but nearly impossible to manufacture in practice. Or you could produce them with such low yields as the cost to be way outside of the value it brings.

Second manifestation of above is why there used to be a lot of talk about more than 2 layers for BD discs, but in reality, those did not materialize for ROM movie format. The yields drop so much for layers above 2 that in practice you can’t make them without changing the core spec and relaxing some of the tolerances (and in process, make the drive more difficult to make). Recordable discs enjoy other freedom which comes from not having to be produced in a ROM replication process and of courses, the discs can and do cost a lot more than movie discs.

Third manifestation comes from the challenge of using spin coating instead of expensive film. Spin coating is a process where you drop some resin on the inside ring of a BD disc, and let centrifugal force spread it across the surface. Sounds neat on paper and looks exciting to look at it work on a machine. Problem is, the same gravity force means that some of the resin pools at the outside edge of the disc (called “damming effect”), and thins out on the inner circle, making the thickness non uniform, taking away the precious tolerances. Hence the reason expensive but ultra smooth plastic film appears to still be in use for BD-50 discs. The film unfortunately also has a set of issues with contaminations under it and hence the deductions for it when computing the total yeilds.

As for using a rule that double the cycle time means the replicator charges twice the price, I'd really love to see some evidence that is true.

We are not talking about what people “charge.” We are talking about all else being equal, what is the true cost of a BD disc to manufacture (which is what the original question was). If I have two DVD plants, both of which cost identical amounts to build, the one that produces twice as many discs per day, can clearly make me twice as much money for the same investment, given a competitive marketplace. If BD is to compete with DVD, and replace it in three years as some are claiming, then they have to live by these rules. A large studio distributes half a billion DVDs a year. With those kinds of volumes, one cannot look the other way on mass volume production considerations such as this, nor can a replicator try to stay in business having to buy twice as many lines yet sell the discs for the same price as before.

Besides, if they were to do it they would double on their cost, not on the price they charge which includes profit.

Speaking of "profits", if you want to maintain the same gross margin percentage for your business, you would have to mark up the incremental costs by that amount. This would make the equation worse for BD, not better. Put another way, if I sell a disc for 50 cents and it costs me 25 cents to make, then I have 50% gross margin. If the cost goes up to $2 and I sell it for the same "profit" (25 cents) at $2.25, my gross margin just shrank down to 10%! If I am a public company, my stock crashes on that day as there are a lot of companies with much better profit models, given the same cost structure. But we digress as economic discussions can become argumentative quickly....

What is the TRUE yield for DVD?

Upwards of 96%+. And these numbers are verified by independent replicators who do not have an IP stake in DVD format. BD numbers that you see thrown around in press releases not only lack verification by third-parties, but are often from “pilot” lines which are nothing but test/R&D labs. Even if one assumes their yield numbers are right and can somehow be verified, when you take such processes and install them in volume replication plants, you wind up with very different numbers (e.g. labor is free in a lab so people can spend far more time tuning a machine than in a real factory where you have to pay for such expert skills). This is why the industry puts little value on announcements such as one from Panasonic circa 2005, claiming 80% yields in their Torrance “pilot line” for BD-25.

[AlexBC]

Paidgeek,

thank you for the explanations, your care and attention to us here at AVS has made me even more eager to SPHE content.

Regarding CR and KFH, I still have a little doubt, you mention they come from the same transfer, that's fair enough. But is the bitrate of the encoding similar as well?

I ask this because I've been importing SPHE DVDs from various countries (mostly US, Jp and Korea) for a number of years as well as purchasing our local Brazilian version (all of them NTSC versions) and even though one can clearly notice they come from the same transfer, the video bitrate budget sometimes varies a lot from version to version (or region to region). Kinda like some of them would be superbit versions with much higher butyrates.

Now I guess the question is, will this happen to the SPHE BD versions as well? Or we can expect the same or very similar encodings?

It's really frustrating going on a quest to find the very best version of our beloved movies even though it's being distributed by the same company worldwide.

Originally Posted by paidgeek

I would be interested in your further explanation of why you very much like something that you cannot decode. Please PM me and we can discuss...

I second Dan's comments and I can say my reasoning for doing it is just that we can be future safe. The capable hardware will come along (very shortly I guess) and our collection will gain a whole new perspective once it's available to us. I don't think it makes much sense to limit the software now just because the hardware hasn't matched it yet, since it's perfectly possible to keep the software at this higher level and still be backwards compatible (the 48/24 tracks will just be downsampled).

[MidnightWatcher]

To any insider:

I was at a major DVD rental chain up here in Canada earlier this week and asked them if they would be carrying HD DVD any time soon. Their answer was that nothing was confirmed at this time, but he did know that Blu-ray was lobbying the company's head office to carry BD. The head office was reluctant, however, because BD wanted to charge them $100 per disc. Assuming this is accurate, why the huge price variance between rental and retail discs? Is Blu-ray trying to recoup losses somehow, or have they opted to make discs intended for the rental market more durable somehow for this purpose, hence the higher cost?

[paidgeek]

Originally Posted by AlexBC

Paidgeek,

thank you for the explanations, your care and attention to us here at AVS has made me even more eager to SPHE content.

Regarding CR and KFH, I still have a little doubt, you mention they come from the same transfer, that's fair enough. But is the bitrate of the encoding similar as well?

I ask this because I've been importing SPHE DVDs from various countries (mostly US, Jp and Korea) for a number of years as well as purchasing our local Brazilian version (all of them NTSC versions) and even though one can clearly notice they come from the same transfer, the video bitrate budget sometimes varies a lot from version to version (or region to region). Kinda like some of them would be superbit versions with much higher butyrates.

Now I guess the question is, will this happen to the SPHE BD versions as well? Or we can expect the same or very similar encodings?

It's really frustrating going on a quest to find the very best version of our beloved movies even though it's being distributed by the same company worldwide.



I second Dan's comments and I can say my reasoning for doing it is just that we can be future safe. The capable hardware will come along (very shortly I guess) and our collection will gain a whole new perspective once it's available to us. I don't think it makes much sense to limit the software now just because the hardware hasn't matched it yet, since it's perfectly possible to keep the software at this higher level and still be backwards compatible (the 48/24 tracks will just be downsampled).

As a rule, we are going to use the same encoded video file for all versions worldwide unless we cannot do so because of censorship. When censorship is an issue, we will keep the same bit rate specs as the NA version. The situation with DVD was different in that disc capacity required for some versions might force a different bitrate, particularly for PAL.

For audio, I am clear on the benefits of using one of the lossless codecs and the attraction this has for some consumers. What I did not understand was the preference for DTS lossless versus Dolby because, being lossless, the only other consideration should be the availability of decoders... Dolby has some out there with more coming and DTS, at present, does not.

[Kris Deering]

Originally Posted by paidgeek

Kris, in planning at our releases, we are considering all markets. How we support audio in Europe is one of the key factors we will based encoding decisions on. With multiple languages in 20 bit DD THD we will easily use more than 20% of a BD-50 on some titles.

My point on HD-DVD is was not whether there are any titles with DD THD, but rather that it is very uncommon.

I understand that but it isn't like Sony is releasing their present titles with uncompressed audio for every language now. What I am saying is there is no reason that I can see for Sony not to include a full resolution PCM OR DDTHD soundtrack with every release and still include all the languages you want at 640DD or 1.5DTS. Adding both seems like a waste of space. With BD50 and BD's bandwidth there doesn't seem to be a reason you couldn't and still deliver top notch video.

[Azumi]

Originally Posted by paidgeek

As a rule, we are going to use the same encoded video file for all versions worldwide unless we cannot do so because of censorship. When censorship is an issue, we will keep the same bit rate specs as the NA version.

To the forum mods,

I hope you'll allow this little follow-up on censorship, which could be helpful for people interested to buy imported discs from Europe:

The UK and Germany are the two european countries with consistent censorship requirements. In Great Britain, a stringent censorship has been in place for quite a long time, while in Germany a new set of regulations has been put in place a few years ago. It's important to remember that some of the cuts are made on a voluntary basis by the studios themselves, in order to obtain more "favorable" ratings.

Therefore, as a rule, if you're interested to buy discs like "Kung Fu Hustle" and are unsure if the UK/German versions are untouched, please remember that the French discs are always presented in their uncut/uncensored versions.

If you wish to import from the UK/Germany, you'll have to make the appropriate research (BBFC, local DVD/hi-def websites) to make sure that the film hasn't been altered.

[paidgeek]

Originally Posted by Kris Deering

I understand that but it isn't like Sony is releasing their present titles with uncompressed audio for every language now. What I am saying is there is no reason that I can see for Sony not to include a full resolution PCM OR DDTHD soundtrack with every release and still include all the languages you want at 640DD or 1.5DTS. Adding both seems like a waste of space. With BD50 and BD's bandwidth there doesn't seem to be a reason you couldn't and still deliver top notch video.

If we dropped LPCM today in favor of a lossless codec, I would expect the customers who are currently enjoying those streams to be upset. It would be acceptable to use 20 bit LPCM, but the nature of the way the data is packed requires that it occupy the same space as 24 bit LPCM. Those 8 bits translate into a significant amount of disc space and bandwidth, even for BD 50.

[abr27440]

Originally Posted by paidgeek

If we dropped LPCM today in favor of a lossless codec, I would expect the customers who are currently enjoying those streams to be upset. It would be acceptable to use 20 bit LPCM, but the nature of the way the data is packed requires that it occupy the same space as 24 bit LPCM. Those 8 bits translate into a significant amount of disc space and bandwidth, even for BD 50.

Paid, could you explain why 20bit LPCM would occupy the same amount of space as 24bit?

[darinp2]

Originally Posted by amirm

TL-51 can of course move through the forum. How easy it would be to do that, depends on the spec. I have not seen the TL-51 specifications so I have no real data to give you. But the forum has approved many variations of HD DVD format already so there is pretty good structure for evaluating and approving it.

I would love to see HD DVD require TL-51 with 1.5x spin rate support for all future players (Gen1s being allowed an exception if required). There was a press release recently about a Chinese format where the only major difference was said to be the codecs used, with the same discs used for it as HD DVD. Looks to me like the same drives could be used for both formats if things stay like that. Do you expect the requirements for players for both formats to change so that they have to support TL-51s, the requirements for only HD DVD players to change so they have to support TL-51s (thus creating another difference between the 2 formats), or for neither format to require players to support TL-51s?

--Darin

[AlbertA]

Originally Posted by abr27440

Paid, could you explain why 20bit LPCM would occupy the same amount of space as 24bit?

Because it probably needs to be packed in 3 bytes and probably cannot be bitpacked.

[paidgeek]

Originally Posted by abr27440

Paid, could you explain why 20bit LPCM would occupy the same amount of space as 24bit?

I have not looked into this deeply. Our engineering staff has though and they tell me that there is no way around this as it is part of the specifications.

[paidgeek]

Originally Posted by darkedgex

As I understand it the Playstation 3 should be adding DTS HD MA decoding in this months firmware, and the Panasonic(?) 2nd gen BD player will include DTS HD MA decoding out of the box. Other devices could, conceivably, add DTS HD MA decoding via firmware updates as well. Alarmingly though, Sony's 2nd-gen BD players don't include DTS HD MA or Dolby TrueHD. If you have any contact with the hardware side of Sony, it would be nice to know what their thinking is regarding next-gen audio CODECs.

In any event, I'd like to think both Dolby TrueHD and DTS HD Master Audio decoding will sneak into A/V Receivers as HDMI 1.3 is adopted (allowing even first-gen BD players (assuming they all support HDMI 1.3) to send the raw bitstream out).

My information is different than yours, but I hope yours is right. I am not aware of any updates coming for PS3 or the Panasonic player that can support DTS lossless. You should also know that the first gen players do not support HDMI 1.3. This is due to the fact that the required chips were not available at the time the players were developed.

[darkedgex]

Originally Posted by paidgeek

My information is different than yours, but I hope yours is right. I am not aware of any updates coming for PS3 or the Panasonic player that can support DTS lossless. You should also know that the first gen players do not support HDMI 1.3. This is due to the fact that the required chips were not available at the time the players were developed.

The Panasonic support was just announced in the last day or so at CeBIT—

This week at CeBIT Panasonic introduced the DMP-BD10A, which is a revision to the DMP-BD10. Among other things the A revision brings with it advanced CODEC support including Dolby TrueHD and DTS-HD.

Of course they don't say DTS HD MA, but I'd be really annoyed if they only included support for DTS HD. As for the PS3, I'll have to go spelunking for a source for that later. =)

Anyways, was not aware that first-gen players lacked HDMI 1.3 capabilities.

[Reginald Trent]

Originally Posted by Reginald Trent

Amir, can you name HD DVD exclusive studios that provide concert DVDs? Like Image for example, are they HD DVD exclusive?

Amir, I asked this earlier do you have the answer?

[amirm]

Originally Posted by Reginald Trent

Amir, I asked this earlier do you have the answer?

I am sorry but I don't have an answer. You might want to ask in the HD DVD software section. I am sure someone else does .

[trbarry]

Originally Posted by amirm

.
.
.
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.
.
.
.

The quote is from Amir, on the loop filtering test, but my question here is really for paidgeek, dr1394 or one of the AVC encoder insiders. How common is Qp 17 (or higher) in typical AVC encoded highdef movies? That is, given the large size of highdef discs does Qp 17 or higher typically occur often, or only in extremely hard to encode scenes?

- Tom

[RobertR1]

Amir,

Any comment on the Studio Canal delays?

[Matt_Stevens]

I asked that two days ago and he has not responded.

Amir, if you are not aware of reasons or as to what happened with the audio, please just say so. That is a better answer than no answer at all.

Thanks.

[Krobar]

Hi Amir,

(Maybe this got missed first time round)

According to a Major retailer here in the UK we are set to get a load of the Studio Canal titles on the 26th of this month (Cover art is rather nice), do you know if the audio speedup issue will be fixed for these?

[Matt_Stevens]

xploited cinema just reported that those titles are now indefinitely delayed.

[RBFilms]

Hi Darin,

AACS charges a $3,000 to $10,000 initial contract administration fee ... with the cost for each studio being determined by certain factors I can't recall offhand.

I paid $3,000 for the privilege of being tagged another $1,500 plus $200 for e-mail delivery = $1,700 for every check disc we make that does not work.

So, in the case or an early adopter release like CHRONOS, which took at least four (4) Check Discs to get right, AACS charged $1,700 x 4 = $6,800 for the "AACS Encyption Keys."

Add the $3,000 initial fee for the AACS Contract, and we got banged for almost $10,000 by AACS just to put out a Blu-Ray Disc. Nice.

We are forced to use AACS for BD but have the option with HD-DVD. You may notice that we do NOT use AACS on any of our HD-DVD releases nor do we plan to in the future.

Some people have asked me why the BD Version of CHRONOS cost more...this is one reason. If we could sell 10,000 to 20,000 units, these costs could be absorbed. However, we will be lucky if we sell maybe 25% of that number in year #1.

For some reason, the HD-DVD Version of CHRONOS is outselling the BD Version 2 to 1. This could be temporary as the HD-DVD has been out longer.

This is probably more information than you asked probably care to know, but I cannot help myself once I get rolling..

Originally Posted by darinp2

paidgeek,

Can you please comment on this from http://www.hdtvuk.tv/:
Do you know if the fees referred to here include AACS fees? Also, is that $3000 per release, per company, something else?

Thanks,
Darin

[Azumi]

To all Insiders:

Studiocanal just issued an interesting press release stating that their 2nd HD-DVD wave will now be released on May 21, and that the reason of the delay was to integrate a new version of AACS which is due in April.

Amir, if you're planning to respond, it would be great if you could answer both on BD and HD-DVD on equal ground, because this is a global issue, and because both the BD and HD-DVD versions of Canal's 1st day-and-date release (Prête-moi ta main) have been affected as well and have yet to receive a new release date.

- What do we have to expect from the new set of AACS? Does it require a firmware update of all the existing BD and HD-DVD players in order to read the new discs?

- Does anybody know if this is the reason that led Fox to delay some of their BD titles?

- Assuming that new software has to be installed in our player, will the existing discs continue to be played, even if some of our copies have keys that had been compromised?

[amirm]

Originally Posted by Matt_Stevens

I asked that two days ago and he has not responded.

Amir, if you are not aware of reasons or as to what happened with the audio, please just say so. That is a better answer than no answer at all.

Thanks.

I frankly have yet to get to the bottom of the complete answer. I have had probably dozens of communications on this and don't have definitive answers to give you across all the titles (this is complicated a bit given the larger set of post production houses involved).

If there is a partial answer, it is the fact that it seemed no one had tried to make 24p encodes in EU for optical discs. We thought it would be a walk in the park, using US masters but apparently not.

I can also tell you that I am putting extreme pressure on people to fix these things movig forward. But am having a hard time getting full confirmation of what is fixed and what is not. But I know the information is there now. As I learn more, I will reply.

BTW, I know this might be annoying but my preference is to only answer questions when I have something concrete to say. In the past, I just get beat up and get more arguments and hence not replying .

[Ja Phule]

Richard,

Has R&B Films decided how they will use managed copies. Since AACS will not be used for HD DVD, how will this affect the use of managed copies?

Originally Posted by RBFilms

Hi Darin,

AACS charges a $3,000 to $10,000 initial contract administration fee ... with the cost for each studio being determined by certain factors I can't recall offhand.

I paid $3,000 for the privilege of being tagged another $1,500 plus $200 for e-mail delivery = $1,700 for every check disc we make that does not work.

So, in the case or an early adopter release like CHRONOS, which took at least four (4) Check Discs to get right, AACS charged $1,700 x 4 = $6,800 for the "AACS Encyption Keys."

Add the $3,000 initial fee for the AACS Contract, and we got banged for almost $10,000 by AACS just to put out a Blu-Ray Disc. Nice.

We are forced to use AACS for BD but have the option with HD-DVD. You may notice that we do NOT use AACS on any of our HD-DVD releases nor do we plan to in the future.

Some people have asked me why the BD Version of CHRONOS cost more...this is one reason. If we could sell 10,000 to 20,000 units, these costs could be absorbed. However, we will be lucky if we sell maybe 25% of that number in year #1.

For some reason, the HD-DVD Version of CHRONOS is outselling the BD Version 2 to 1. This could be temporary as the HD-DVD has been out longer.

This is probably more information than you asked probably care to know, but I cannot help myself once I get rolling..

[amirm]

Originally Posted by Azumi

To all Insiders:

Studiocanal just issued an interesting press release stating that their 2nd HD-DVD wave will now be released on May 21, and that the reason of the delay was to integrate a new version of AACS which is due in April.

Amir, if you're planning to respond, it would be great if you could answer both on BD and HD-DVD on equal ground, because this is a global issue, and because both the BD and HD-DVD versions of Canal's 1st day-and-date release (Prête-moi ta main) have been affected as well and have yet to receive a new release date.

- What do we have to expect from the new set of AACS? Does it require a firmware update of all the existing BD and HD-DVD players in order to read the new discs?

.....

- Assuming that new software has to be installed in our player, will the existing discs continue to be played, even if some of our copies have keys that had been compromised?

I do have more information on this one but don't know if I can fully disclose it all. But I can address the one point above. If you are using PC software players (at least the ones I know about), then you would need a new player to play some content moving forward (and at some point, all new content). No other product is impacted so you are good to go with any BD/HD DVD player outside of PC software and don't need a firmware upgrade for them.

And yes, regardless you can continue to play the discs you own.

[Matt_Stevens]

Amir, thanks very much for the reponse on the Studio Canal situation. I can understand your hesitating to comment when you are still waiting for confirmation of facts, etc, so I am grateful that you could share with us what you know so far.

[rwestley]

Amirm, I know you can't or will not talk about the rumor of the Xbox 360 with HdMI output. I will say that you should use any pull you might have to get them to include a HD drive in the system when they come out with the upgrade with HDMI. Some rumors do not mention the inclusion of a HD drive and If there is no HD drive in the unit it could be the death of HDDVD. This is the time for Microsoft to show its true support for the format. Give us a small hint.

If Microsoft is woried about alienating the current owners it could offer a special upgrade price to current Xbox 360 owners. Or come out with a new lower priced HD add on drive with HDMI outputs. If one came out for under $100 with HDMI outputs I think it would really sell and would add to the Xbox and HD base. Watch the studios react if Microsoft would do this. I wonder how Andy would spin this?