Skip to main content

BDJ vs HDMV Mode

10 replies [Last post]
bddeveloper
Offline
Joined: 2008-01-14
Points: 0

Hello,

We have clients wondering about BD-Live based on what they read about CES 2008.
I have some questions that I can not seem to find anywhere so I am posting them here.

1. Where is the official BD Spec versions for each player, in other words when I'm looking at a player how can I tell which spec it supports?
2. Is it required that each CE manufacture develop/co-develop their own VM or is Sun providing a VM for the devices?
3. If there is a VM that sun is providing is it possible to obtain this VM as the one to test with? You might ask why what does it matter but if you've developed a Java Application for enterprise systems you'll know that using the IBM VM makes a world of difference when it comes to testing your application for performance.
4. Has anyone (Hollywood Studio) created a BD-J Disc, as in the ENTIRE disc is all done in BD-J expect for maybe the First Play title which could be in HDMV?
5. Do you currently recommend creating a disc with only BD-J? If so are there any performance, and compatibility issues that should be considered?
6. Is there compiled a list of players that handle BD-J well and those that do not?

These questions will help me direct our clients interested in "BD-LIVE" brilliant marketing by the way giving the network connectivity to a box a name. Got my clients attention!

Thank you in advanced.

Scott

Reply viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.
Bill Foote

bd-j-dev@mobileandembedded.org wrote:
> Hello Joe,
>
> <...>
> I hope Sun can provide some answers on the VM issue.

OK. I agree with Joe :-)

What Joe said is exactly right - the VM performance is only one
part of the overall performance question. It really is a question
of the entire stack, and the kind of hardware it runs on.

This is true for every consumer device I know of - legacy DVD, other
optical formats, mobile phones (with or without Java), etc.

> It is too bad that we have to test on all the players for performance issues.

It's one of those "it depends" things. You probably don't have to test
on *all* players - with experience, I'm sure folks will develop good
information about a representative sample of different players, with
different performance profiles.

But yes, if you want to push the performance, then you do need to
do research on what different hardware models are capable of.

> I know about the stitching "trick" do you for see this going away?

I agree with Joe here too. It's really a function of how image
decoding works. The other thing is, once you have your frameworks
in place, doing image mosaics isn't very hard at all. GRIN provides
a bunch of support for it, for example, and one of our soon-to-be-completed
RFE's will make it easier to take just the mosaic part out of GRIN, if that's
what you want.

In an interesting way, image mosaics show what's really /right/ about
Java. If a declarative format were baked into the hardware, you'd
pretty much be forced to just do images the slow way, unless they changed
the spec and updated all of the players in the field. With Java, you can
build a library to do things the best way, and you can learn from experience.

Cheers,

Bill

> As in I can load in my images one by one?
>
> Thanks again.
>
> Scott
> [Message sent by forum member 'bddeveloper' (bddeveloper)]
>
> http://forums.java.net/jive/thread.jspa?messageID=253837
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: bd-j-dev-unsubscribe@hdcookbook.dev.java.net
> For additional commands, e-mail: bd-j-dev-help@hdcookbook.dev.java.net
>

---------------------------------------------------------------------
To unsubscribe, e-mail: bd-j-dev-unsubscribe@hdcookbook.dev.java.net
For additional commands, e-mail: bd-j-dev-help@hdcookbook.dev.java.net

bddeveloper
Offline
Joined: 2008-01-14
Points: 0

Hi Bill,

As a Java Developer since 1999 I understand the hardware issues when running a VM. The better the hardware the better the performance. However at the same time using the classic example of the IBM VM verse Sun VM, the IBM VM performed much better on any windows based machine you placed it. Therefore the reason I ask about this is that I'm working on a list of "unknowns" that could cause more development work. If I do not know the VM this is an unknown, for example if Panasonic releases a new player and it has better hardware but they switched VM's then I have two new Unknowns. Verse if they had one VM and I knew the hardware for the new machine was faster then the old then the chances of the new player performing better are high, still an unknown risk but there is a good calculated assumption I can add to that unknown. Does this make sense?

As far as the animation engine is concerned I understand the advantages to a language such as Java or any other good OO language. However the one thing that I think Sun has done very well is to implement specifications that can be used or not used. Such example include a Business Rules Specification, an EJB spec, XML Spec, and servlet engine. Some of these specs are implemented in certain SDK's or they are implemented by a vendor, such as BEA. This model seems to work for the Java Development industry and companies. Each of these specs and implementations are optional. For example if I want to write my own XML parser I can, I can also read the spec and implement my own or go to Jakarta and download an implementation that i think performs well and is in my budget. The reason I bring this is up is when a feature becomes so important such as stitching together images and writing a custom animation engine it makes sense to develop a standard that the industry uses. This helps raise the level of performance for all titles that are released.

Since I've been writting hava work for so long I guess I am just getting an understanding how how this one off of the entire Java world works. Maybe it doesn't work like the rest of the industry and that is the way it is.

Scott

Bill Foote

bd-j-dev@mobileandembedded.org wrote:
> Hi Bill,
>
> As a Java Developer since 1999 I understand the hardware issues when running a VM. The better the hardware the better the performance. However at the same time using the classic example of the IBM VM verse Sun VM, the IBM VM performed much better on any windows based machine you placed it. Therefore the reason I ask about this is that I'm working on a list of "unknowns" that could cause more development work. If I do not know the VM this is an unknown, for example if Panasonic releases a new player and it has better hardware but they switched VM's then I have two new Unknowns. Verse if they had one VM and I knew the hardware for the new machine was faster then the old then the chances of the new player performing better are high, still an unknown risk but there is a good calculated assumption I can add to that unknown. Does this make sense?

Sure, understood.

> As far as the animation engine is concerned I understand the advantages to a language such as Java or any other good OO language. However the one thing that I think Sun has done very well is to implement specifications that can be used or not used. Such example include a Business Rules Specification, an EJB spec, XML Spec, and servlet engine. Some of these specs are implemented in certain SDK's or they are implemented by a vendor, such as BEA. This model seems to work for the Java Development industry and companies. Each of these specs and implementations are optional. For example if I want to write my own XML parser I can, I can also read the spec and implement my own or go to Jakarta and download an implementation that i think performs well and is in my budget. The reason I bring this is up is when a feature becomes so important such as stitching together images and writing a custom animation engine it makes sense to develop a standard that the industry uses. This
helps raise the level of performance for all titles that are released.

Well, right, and that's what we've started at https://hdcookbook.dev.java.net.
Again, participation is welcome!

> Since I've been writting hava work for so long I guess I am just getting an understanding how how this one off of the entire Java world works. Maybe it doesn't work like the rest of the industry and that is the way it is.

The way the rest of /which/ industry works?

This is a really important thing to understand. The media industry,
and especially the optical disc/Hollywood/movie industry, is not the
same thing as the IT industry. Indeed, in many ways, they're light-years
apart.

In terms of business culture, I personally think that the two will meet
somewhere in the middle. We're already a whole bunch more open with
BD-J than was ever the case for legacy DVD, for example, but folks
coming from the IT or other more computer-science-y pursuits will
find some culture shocks along the way, too.

I do expect that as time progresses, we'll see things become
increasingly open. Rome wasn't built in a day, and all that.
Also, I'm really looking forward to the interesting, creative
directions this platform and format will take -- and it ain't
all gonna be movies, either.

Welcome to a grand adventure!

Cheers,

Bill

> Scott
> [Message sent by forum member 'bddeveloper' (bddeveloper)]
>
> http://forums.java.net/jive/thread.jspa?messageID=253871
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: bd-j-dev-unsubscribe@hdcookbook.dev.java.net
> For additional commands, e-mail: bd-j-dev-help@hdcookbook.dev.java.net
>

---------------------------------------------------------------------
To unsubscribe, e-mail: bd-j-dev-unsubscribe@hdcookbook.dev.java.net
For additional commands, e-mail: bd-j-dev-help@hdcookbook.dev.java.net

Joe Rice

I guess I should clarify that there are a number of issues affecting
performance. Image mosaics are mainly used to improve image decode
(and to a lesser extent file i/o) performance. Animation display once
an image is decoded is another piece of the puzzle.

Cheers,
Joe

On Jan 14, 2008, at 12:53 PM, bd-j-dev@mobileandembedded.org wrote:

> Hello Joe,
>
> Thank you for the quick response. For the titles that are all
> 100% BDJ mode would it be best to look at the index.bdmv file to
> verify each title on the disc is run in BDJ mode? We will pick up
> these titles tonight for some R&D.
>
> I hope Sun can provide some answers on the VM issue. It is too bad
> that we have to test on all the players for performance issues. I
> know about the stitching "trick" do you for see this going away? As
> in I can load in my images one by one?
>
> Thanks again.
>
> Scott
> [Message sent by forum member 'bddeveloper' (bddeveloper)]
>
> http://forums.java.net/jive/thread.jspa?messageID=253837

---------------------------------------------------------------------
To unsubscribe, e-mail: bd-j-dev-unsubscribe@hdcookbook.dev.java.net
For additional commands, e-mail: bd-j-dev-help@hdcookbook.dev.java.net

bddeveloper
Offline
Joined: 2008-01-14
Points: 0

Hello Joe,

Is there any reason why an animation engine that performs well, isn't part of the specification or maybe have the spec outline what the performance requirements are so that each VM must ad-hear to them. I remember doing servlet development before there was a Sun Servlet engine (actually we had to use J++) and the performance was all over the place.

Just so I understand how this BD-J eco system works, how much say does the BDA have in such things as a BD-J animation entire and requirements and how much does Sun have in the say. For example since Sun worked with IBM to develop the servlet engine 1.0 in 2000 would that be similar to the relationship Sun has with the BDA, and therefore would be able to develop a spec for an animation engine?

This is just thinking out loud here but do you for see the BD-J UI spec as a solid spec or do you see it like AWT, which was a good spec but hard to develop in and later SWING was released, and now the next version of JSF stuff.

Bill Foote

bd-j-dev@mobileandembedded.org wrote:
> Hello Joe,
>
> Is there any reason why an animation engine that performs well, isn't part of the specification

I can answer that - because it's not necessarily one-size-fits-all.
Java lets people write libraries, and this means that we can enjoy
better and better libraries as the format matures, rather than something
written by committee.

Check out the animation engine at https://hdcookbook.dev.java.net,
and the related GRIN scene graph that uses it.

> or maybe have the spec outline what the performance requirements are

The spec, along with the guidelines document, does just that.
This does help, but realize that specifications and guidelines
can't capture everything.

> so that each VM must ad-hear to them. I remember doing servlet development before there was a Sun Servlet engine (actually we had to use J++) and the performance was all over the place.

I'm hopeful that our freeware animation engine will help to have a similar
effect here, but a bunch of other smart people are out there writing libraries
too. I'm sure a couple will shake out as the format develops.

> Just so I understand how this BD-J eco system works, how much say does the BDA have in such things as a BD-J animation entire and requirements and how much does Sun have in the say.

Well, I go to a lot of the BDA meetings for Sun (though I'm in the process
of cutting back), so I can tell you that the formal answer is very close
indeed to the de-facto pragmatic answer. That is: The BDA has 18 board
members, and Sun is one of them.

Now, I think it'd be wonderful if a consortium of folks could get together
to work on a standardized framework. I would absolutely welcome that; indeed,
I've put some feelers out there and looked for ways to structure this. And,
I should say, we are getting good cooperation and help from a lot of the other
folks in the BDA on what we're doing with the hdcookbook repository, and other
projects.

I'd love to see this develop further, and anything you might be able to do to
help would be most welcome.

Cheers,

Bill

> For example since Sun worked with IBM to develop the servlet engine 1.0 in 2000 would that be similar to the relationship Sun has with the BDA, and therefore would be able to develop a spec for an animation engine?
>
> This is just thinking out loud here but do you for see the BD-J UI spec as a solid spec or do you see it like AWT, which was a good spec but hard to develop in and later SWING was released, and now the next version of JSF stuff.
> [Message sent by forum member 'bddeveloper' (bddeveloper)]
>
> http://forums.java.net/jive/thread.jspa?messageID=253848
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: bd-j-dev-unsubscribe@hdcookbook.dev.java.net
> For additional commands, e-mail: bd-j-dev-help@hdcookbook.dev.java.net
>

---------------------------------------------------------------------
To unsubscribe, e-mail: bd-j-dev-unsubscribe@hdcookbook.dev.java.net
For additional commands, e-mail: bd-j-dev-help@hdcookbook.dev.java.net

bddeveloper
Offline
Joined: 2008-01-14
Points: 0

Hello Joe,

Thank you for the quick response. For the titles that are all 100% BDJ mode would it be best to look at the index.bdmv file to verify each title on the disc is run in BDJ mode? We will pick up these titles tonight for some R&D.

I hope Sun can provide some answers on the VM issue. It is too bad that we have to test on all the players for performance issues. I know about the stitching "trick" do you for see this going away? As in I can load in my images one by one?

Thanks again.

Scott

Joe Rice

> I hope Sun can provide some answers on the VM issue. It is too bad
> that we have to test on all the players for performance issues. I
> know about the stitching "trick" do you for see this going away? As
> in I can load in my images one by one?

I think image mosaics are probably with us to stay. You may find with
more elaborate animations that you'll need to have an animation engine
which supports altering frame rate / skipping frames on slower players.

Cheers,
Joe

---------------------------------------------------------------------
To unsubscribe, e-mail: bd-j-dev-unsubscribe@hdcookbook.dev.java.net
For additional commands, e-mail: bd-j-dev-help@hdcookbook.dev.java.net

Joe Rice

On Jan 14, 2008, at 11:52 AM, bd-j-dev@mobileandembedded.org wrote:

> Hello,
>
> We have clients wondering about BD-Live based on what they read
> about CES 2008.
> I have some questions that I can not seem to find anywhere so I am
> posting them here.
>
> 1. Where is the official BD Spec versions for each player, in other
> words when I'm looking at a player how can I tell which spec it
> supports?

You'll have to check with the manufacturer of each player. Profile 1,
Version 1.1 is also known as Final Standard Profile (or FSP), and as
"Bonus View" in marketing parlance. Profile 2.0 is known as "BD-Live."

Any new players being introduced now have to be FSP. Players on the
market now that support FSP are The Panasonic DMP-BD30 and the PS3 (w/
firmware 2.1 or higher). PowerDVD and WinDVD will shortly release FSP
versions if they haven't already.

I don't believe any BD-Live players have been released.

> 2. Is it required that each CE manufacture develop/co-develop their
> own VM or is Sun providing a VM for the devices?

There are various VMs used in various players. Sun might have one
available, but the manufacturer can choose which they use.

> 3. If there is a VM that sun is providing is it possible to obtain
> this VM as the one to test with? You might ask why what does it
> matter but if you've developed a Java Application for enterprise
> systems you'll know that using the IBM VM makes a world of
> difference when it comes to testing your application for performance.

Somebody from Sun could address this better than I can.

> 4. Has anyone (Hollywood Studio) created a BD-J Disc, as in the
> ENTIRE disc is all done in BD-J expect for maybe the First Play
> title which could be in HDMV?

Yes, there are dozens of such discs. I'd say most A-list titles are BD-
J at this point. (Some do use HDMV for first play). You can look at
Disney's The Guardian, Chicken Little, Pirates BDs, Cars, Ratatouille,
or Fox's Legion of Extraordinary Gentleman, or Sony's Spiderman 3, to
name only a few.

> 5. Do you currently recommend creating a disc with only BD-J? If
> so are there any performance, and compatibility issues that should
> be considered?

There have been compatibility issues, but the player manufacturers are
working to resolve these quickly. Performance is variable between
players, and it's wise to code in a way that checks performance and
accounts for it.

> 6. Is there compiled a list of players that handle BD-J well and
> those that do not? \

All BD players must support BD-J.

Cheers,
Joe

---------------------------------------------------------------------
To unsubscribe, e-mail: bd-j-dev-unsubscribe@hdcookbook.dev.java.net
For additional commands, e-mail: bd-j-dev-help@hdcookbook.dev.java.net

billshepp
Offline
Joined: 2004-05-20
Points: 0

> On Jan 14, 2008, at 11:52 AM,
> bd-j-dev@mobileandembedded.org wrote:

>> 2. Is it required that each CE manufacture develop/co-develop their
>> own VM or is Sun providing a VM for the devices?

> There are various VMs used in various players. Sun might have one
> available, but the manufacturer can choose which they use.

It's more than just the VM, it's the full BD-J stack. A full BD-J stack can be developed in-house or can be sourced from vendors such as Philips, Alticast, Sonic, and Osmosys. Sun provides the VM which is in most of the BD-J stacks.

>> 3. If there is a VM that sun is providing is it possible to obtain
>> this VM as the one to test with?

Compatibility concerns are affected by the full stack more than just the VM, and performance is variable based on both the stack and the hardware it's running on. Currently I believe most developers find the PC with one of the commercial players (WinDVD, PowerDVD, Arcsoft Digital Theatre, etc.) to be the most productive early development environment, then once optimization and testing are underway moving to the PS3 and other popular players (Sony BDP-S300, Samsung BD-P1400, Samsung DMP-BD10/30, etc.).

Bill