Industry Insiders Q&A Thread: only Questions to insiders please - Page 23  

I think we probably need to get off this topic before we get told to do so :). I should also preface that, I am at least a decade past the time I looked at this kind of thing :). So here goes one response anyway. If it is wrong, someone should create a new thread in the right place to disucss it further.

I am confident the HDMI recievers have a PLL that is used to recover the source signal/clock. The problem we have here though, is not just straight signal recovery as would be typical in these scenarios. But maintaing very precise timing relative to the original signal before it went through the HDMI transmitter. Put another way, the receiver may be able to nicely lock onto the transmitter, but if the transmitter is dancing back and forth, the recovered signal while digitally accurate, is also smeared in time.

Take the same data after the PLL and put it in a simple buffer, and then you can clock that out with precision. Down side is the delay and additional complexity. And you have to figure out what to do if your buffer underruns.

It depends on whether they grant you what we call the "burn right." Publishers charge a few cents a song for every physical copy so this "right" doesn't come for free. And of course, all burning logic these days is for CDs. So new software would need to come about to burn it to HD formats and at higher sampling rate. But then the content owners may require some for of copy protection to be used which off hand, I am not sure what it would be as AACS can not be generated by consumers.

The best solution is what we already have. That is, stream the source from the PC and be done with it. Xbox 360 can do it already. Indeed, all we need is a bit of extra software and every HD DVD player with an Ethernet port can do it. Pioneer BD player has this capability but I am not sure if it can handle > CD resolution though (i.e. WMA Pro).

Per answer above, an network audio receiver would do the job nicely. New receivers should start having this functionality built-in.

Agreed. And there should be some nice solutions on this soon.

So is this a crapshoot with the BD players at least with the 1G?

That has nothing to do with jitter. If the data is so far out of sync that the buffer is exhausted on the receiving device, then you'd lose the signal completely while the HDMI interface resynced. But there's no way you can suggest that minor fluctuations in the clock sending compressed data over a very high-speed link somehow affects the timing of the very low-speed data sent to the D/A converters after it's decompressed, decrypted, buffered, etc. That is purely a concept made up by the companies that want to sell you $250 HDMI cables.

--
Steve

Don't you already have the additional complexity, since there are so many different interfaces running at different speeds inside an HD-DVD player? I mean, the timing of an HDMI signal isn't even in the same ballpark as the timing of digital audio or video streams that you might construct from data passed over HDMI. Plus I assume that the audio data is buffered since it generally takes longer to process and scale the video than the audio.

--
Steve

OK, someone please ask an HD DVD/BD question before we all get in trouble :D....

I bolded that in Amirs' response. What he is referring to is called Time of Day. That is not only does your TX/RX have to get where they are supposed to be at the right time, but the when is also very important too.

The solution is a Stratum 2 clock, but then the prices of the players would shoot, "Straight to the Moon, Alice:! :D

You don't like this guy do you?

Nice explanation Amir re high-quality digital audio downloads, thanks.

It'd still want something I can hold in my hand rather than go the streaming route. That's why I like BD and HD-DVD (I can hold it, store it and be proud of my collection).

I'll keep an eye on what actually transipers. The content providers need to be carefull, place too many hurdles in front of legitimate consumers risks losing those consumers who end up say "this is not worth the hassle". A fine line indeed balancing content owners vs consumers rights and protections.

Thanks.

Dennis.

I love Darin and respect him a lot. Folks working for me vouch for him to high heaven on top of that. It is just that he is asking a complex question that created many debates last year, and when I answered a similarly complex one in another topic, he got quite upset at me. So I have not yet figured out how to reply to him.

Please, please, let's not continue with these side discussions. I am in San Diego, going for a nice dinner anyway. When I come back, I expect to see nice, simple questions. :) Someone must want to know something. Like who is producing the first AVC title in US in HD DVD!

OK, maybe the third time will be the charm :p

How about this one?

ok i have an HD DVD question. If anyone knows, why has WB been silent, i was expecting they would release a few titles next Tuesday but their normal Thursday announcement time has come and gone. :(

Good question, even their on-line store is behind the times and doesn't list their latest title ATL. http://whv.warnerbros.com/WHVPORTAL/...functionmode=6

One which I will not accept thank you.

This may be one post too many on this topic but I think there is room for slightly more clarification. There are two sources of timing errors in a received digital audio stream: One, the timing errors of the clock in the source device; these errors are common to both the clock and data bits. Two, jitter in the received clock due to errors in the sensing of the received clock transitions caused by such items as ground noise and cross talk contaminating the clock signal. Reducing the second type of clock error can be done with an ordinary PLL with a tight loop filter or at more expense with a crystal controlled PLL and an even tighter loop filter. Since the second type of timing error doesn't affect the clocking of data bits at the source, reducing it does not require any extra data buffering beyong the register already built into the receiver. Eliminating the first source of timing errors does require buffering though and extremely tight PLL loop filters intended to eliminate jitter could also uncover some of the first type of timing errors thus necessitating some buffering. The use of compression of course has no effect on whether or not the clock the receiver recovers from the source contains timing errors.

I don't know if you can answer any of these due to agreements, but it's worth a try:

-Apparantly the iHD advocates at Disney couldn't overcome their BD-J supporters. Do you know if there are still people at Disney that care for this technology, or have they moved on/no longer work there?

-Once a studio encodes/authors a title in one format (let's say WB), is it basically a no-cost situation in doing a port to another (HD DVD VC-1 encode being used for Blu-ray for example)? This is supposing it would fit on whatever size discs were being used (let's say a 10GB movie for arugment's sake).

-On a scale of 1-10, how would you rate the relationship in HD DVD movie collaboaration with the studios for VC-1 support and disc checking/review? My guess would be a 9 or so for Warner and Universal, 8 for Studio Canal, and 5 for Paramount (this based of what Microsoft has made known publicy).

-What are the chances we see Weinstein titles before the end of the year?

-Do you expect the same effort from Studio Canal that Warner and Universal gave with their titles (visuals + audio + extras)?

-If you were a betting man, what would you rate the chances of another studio announcing HD DVD support before Jan. 1st here in the US (any of the 5 "major" not yet supporting HD DVD) - less than 50%, or more than 50%?




p.s. thanks for all your answers so far

Amir:

That is great news on the Studio Canal releases being compatible with the Toshiba HD DVD player in the U.S. Do you have any news on whether the extras from these non-U.S. releases will be viewable as well? Will they be in PAL format and therefore not compatible with NTSC sets here?

Oh, yes, and who, pray tell, is producing the first AVC title in the US in HD DVD?

SCD

I think this one was missed.

Thanks Rover :). Yes on #1. On #2, yes to component but not being 480p that is mandated for analog out (same as Toshiba/Samsung, etc.). No comment on HDMI one way or the other :).

I'm keen on the answer to this one as posted before:

Thanks, Ian.

Amir,

You talked about noone being able to force you to update your firmware if region codes were ever added to the HD-DVD spec. With the Xbox 360 HD-DVD addon, would HD-DVD software updates be bundled with XBOX Live software updates? What I am really asking is if Microsoft would be able to force you to update your HD-DVD software, if you wanted to continue using Xbox Live.

Cheers.

Amir,

Let me try this another way (I'm not on your ignore list, am I? : :confused:

Here is a post by "dr1394", Ron, who I don't know but have read some of his posts and seems to know his stuff:
Here is a post by me in which I was asking for input on this issue, as it contradicted what I'd previously believed wrt HD DVD and progressive output:
Here is another post by me directed to you, as no one had responded (maybe I'm on everyone's ignore list :p ):

Finally here is my last post on the subject:


The last just links back to the previous. I'd just like some input. It doesn't have to be the last word, can be clarified later, etc; but basically when everyone on here pretty much came to an agreement that "HD DVD is encoded 1080p", despite the flag or header difference to BD; and I had assumed that there would be no issue in getting "pure" progressive material off the disc; now Ron seems to be saying that the only way is to de-interlace ala the Sammy.

Thanks,

Re: 1080P, I would like to suggest that people be a bit more precise about the term "deinterlacing". Removing the 3:2 sequence ("Inverse Telecine") is often a relatively simple-minded process that only takes a state machine and very little compute resources.

True deinterlacing (phase correlation, motion adaptive, temporal upconversion, etc.) can be extremely expensive to implement - it is generally far more complex than simple inverse telecine.

Clearly differentiating between these two concepts may help deconfuse some of the discussions I've seen here (and on other parallel threads).

Just a thought.

Amir...

So I listened to the Xbox podcast about HD DVD.

Kevin Collins said that in the future you will be able to download new HD trailers to your persistent storage on the HD DVD player. This I can buy.

He also said that you might also be able to download a director's cut. How is this possible? Will the HD DVD player read from the HD DVD disc and then jump to the on-board memory to read a new scene, and then jump back to read from the HD DVD disc again?

Is it safe to assume when you speak of the above "True deinterlacing" you are refering to a source that was captured (ie 1080i 30fps video) vs. a "created" 1080i 30fps that was captured as 24fps film ?

b2b

Two examples would be:

1) 1080I30 captured source (sometimes called 1080I60) to 1080P30 or 1080P60

2) 1080P24 captured source carried in a 1080I30 "container" (with 3:2 cadence) temporally upconverted/resampled to 1080P30, 1080P48, 1080P60, or 1080P72. (And I don't just mean simple frame repeats.)

Thanks for the reply. I could see where that could be a nasty bit of work to deal with. Are we speaking of FLI8638 "expensive" or even more than that ? Do any of the Broadcom chipsets seek to address this level of processing ?

b2b

The DVD release of the anime .Hack//Sign (at least the original series, I haven't seen anything else) had in addition to the usual English / Japanese audio tracks, a soundtrack-only track. Yummy :)

Tom -

I think the problem is you can only use the "relatively simple-minded process that only takes a state machine" under those conditions where you are already certain that you have pure telecined 24p material. If the source has been edited as interlaced or processed as interlaced in the player to add PIP, overlays, etc., or if the decoder cannot assure you that it is 24p then you have to assume the worst and feed everything into the sausage maker.

- Tom

edit: While not a hidef DVD insider I have indeed professionally worked on program code to do most of the above on previous projects.

[quote=Tom McMahon][quote=rdjam]Hi Tom - Can the Broadcom chips in the A1 output the 1080p24 video it is decoding on HD DVD?
Quote:
Yes.

Tom, can your decoder output 1080p24sf? When you say that 1080p24 is possible, is it with your current firmware? I know that the manufacturers of the players would have to update their firmware, but I am asking about the present Broadcom firmware.