Blu-ray needs the cartridge because it uses a much thinner plastic layer, making it more vulnerable to scratches, grease, etcetera. I'd expect DVD Forum blue laser to have durability not much worse than DVD. In my experience, I have plenty of 8 year old discs that play just fine, as long as the kids haven't been at them.
Oh, heck, you don't even need the bits. Typical HD DVD-ROM WMV9 content is anamorphic 1440x wide, plus whatever height the source media used (they don't encode the letterboxing into the video stream, unlike DVD). So, a typical 2.35:1 Hollywood movie would 1440x816 @ 24 fps. You could do 120 minutes of that in 4.3 GB in a more than watchable format, especially with the Advanced Profile features. I'd prefer to have the bits of a DVD-9 available, of course. I've been able to do decent 1280x720 24p WMV9 @ 3 Mbps, and that was before Advanced Profile.
MPEG-2 has historically cost about $2/player, down from $5/player not so long ago. I think VC-1 is something like $0.25/player for the decoder component. Quite a bit cheaper.
There is the in development Ogg Theora, but it's not done, and not competitive against the VC-1 feature set. I hope it'll be interesting later.
As for DRM, that happens above the codec level, on the stream or file level. You can apply DRM to a Theora file, and can use a DRM free WMV file. The codec itself doesn't have any "native" DRM. And even if Blu-ray uses DRM, there's no reason to think it'll be Microsoft's.
Prediction: You won't be able to buy a non-HD set of more than 23" within a few years. It's not that much more expensive to build an HD set these days. And it's a feature people are willing to pay for.
See many new black and white sets anymore? This changeover will be quite a lot faster.
They're pretty comparable in many cases, although I think that today's VC-1 implementations do a better job of maintaining small details like film grain than today's H.264 implementations.
Of course, all we can do is compare implementations, not bitstreams, so it's hard to say what the differences between formats are going to be in practice once stuff is actually shipping.
Everyone seems to be thinking that somehow VC-1 is more patent or license protected than MPEG-2. This simply isn't true. Now, maybe real-world enforcement of the MPEG-2 patents aren't particularly aggressive for OSS software decoders, but every DVD player, and EVERY DVD DISC STAMPED requires a payment to MPEG-LA. And VC-1 license payments will also be handled by MPEG-LA. MPEG-2 or VC-1, there still will be payments, and the checks go to the same company. The interesting differences here are technical, not licensing.
This announcement is only about the codec. DRM will happen on a different layer of the standard, and won't be codec specific. The same DRM scheme will be used for both MPEG-2 and VC-1 content in the standard, I'm sure.
MS DRM might be used, or might not be. Personally, I think the MS DRM is technically very good. It provides a very rich syntax of business rules for the content owners. Certainly, there are lots of overly punitive DRM business rule implementations out there (think the T2 HD DVD-ROM). But that's really the content creator's fault. The WM DRM definitely allows for very user-friendly copy protection, if that's what content creators want to use.
xvid is based on MPEG-4 part 2, which is roughly as patent encumbered and has roughly similar license fees and terms as Microsoft's VC-1. If xvid is good enough from a licensing perspective for you, so will VC-1 be.
Now, if what you want is an open-source VC-1 encoder, I'm sure it'll happen once the standard is fully finalized, ala LAME and Xvid. The same kind of open-source but unlicensed codec implementation should be perfectly applicable there.
Er, you don't know what you're talking about. A high end software MPEG-2 encoder like Canopus ProCoder is absolutely as good as the high end hardware encoders, and can be faster than real time on a modern machine from a file-based source (which is the typical workflow these days anyway for anything other than news and other live broadcasts).
As for hardware VC-1 implementations, there are a good half dozen companies working on them, and some where being quietly demonstrated in the back suites of NAB back in April. Rest assured, this is happening.
Well, there area already H.264 codecs in development ala LAME and Xvid. I expect the same to happen for VC-1 - after all the specification will be published.
The issue is license free codecs, of which there aren't any competitive examples. There are things in development, but they simply aren't far enough along to compete against H.264 and VC-1 in the current HD standard processes.
Yes, I know about Theora. No, it isn't far enough along to be competitive.
Why do you assume MPEG-2 will be dominant? the VC-1 codec (aka VC-9, aka WMV9 Advanced Profile) can provide similar quality to MPEG-2 at half the bitrate. This means that content providers could do a project with a cheaper, single layer disc instead of a dual layer disc for longer projects.
I believe the big driver behind this is the competition form the DVD Forum's own blue laser format. DVD Forum already has tentative support for VC-1 and H.264. Even though the DVD Forum has lower digital capacity, the support for better codecs meant that DVD Forum could actually get more hours of good quality content on the disc. So equalizing the codecs means that Blu-Ray's capacity advantage can shine.
That said, I'm still betting on DVD Forum. 30 GB will mean more hours of HD content that DVD can do of SD. Also, DVD Forum discs are MUCH easier to convert an existing DVD plant to, and likely will be more durable in day to day use.
Blu-ray seems more likely to win inside cameras and that kind of thing, where capacity is a bigger deal. Think VHS v. Beta, where Beta turned into the Betacam format, giving Sony a 15-year dominance in professional video formats.
Well, even if this was released, it's a real stretch saying that it's "only a matter of time" before it'll be widely used by US media companies. A codec is really only a small part of a digital media architecture. Some of the competitive factors that go into these choices include:
Compression efficiency Cost of implementing decode in consumer electronics (read, what's the cheapest chip that can decode it) Support for existing transport mechanisms (like MPEG-2 transport streams) Existence of industrial grade encoders (like massive statically multiplexed encoder arrays) License fee
The license fees do matter a little, but that's really a secondary issue. A more efficient but more expensive codec can actually be cheaper to implement, because the content provider can use less bandwidth per channel, enabling them to sell more channels over fixed bandwidth.
Today, the battle for the next generation "TV" codec is between Microsoft's VC-9 and MPEG-4 H.264. And that battle is already well underway. The BBC codec isn't far enough along to compete for the current generation of standardization efforts for technologies like HD DVD and digital cinema.
The wavelet codec was Ogg Tarkin. Ogg Theora is a more traditional codec, based on On2's open sourced VP3 codec from a few years back.
Ogg Theora is lurching towards an actual release, and is supported in a few tools like VLC, while Ogg Tarkin never really got very far along in implementation. Theora was meant to be the quick interim release while Tarkin was developed, although the schedule has slipped quite a bit since.
I've gone to Vegas every March/April for around 10 years now, for the National Association of Broadcasters convention. It isn't THAT hot that time of the year, and it's quite pleasant at night. I used to walk all the time, since the cab lines during a major show can be incredibly long - a 20 minute walk is often much faster, and a lot less boring.
It's a whole different view of Vegas at street level. It really shows the scale of the Casinos, when you realize you've been walking past Caesar's for the last ten minutes:). Heck, these days I stay in the Hilton, which is right next door to the convention center, and it takes literally 20 minutes to briskly walk from my hotel room to a booth in the South Hall.
Oh, I'm not offering an opinion as to the legality of their claim. However, I'm sure someone will try an equivalent effort with VC-9, making the same claim.
There's a difference between MPlayer and Xvid, though, in the difference between distributing source code and a binary. Xvid and LAME make a plausible case they don't count as an implementation since they're just source code. Hard to say if it's true without a judge's opinion, but no one has cared enough to sue them at least.
TI already has a single chip DSP that can decode all the proposed HD DVD codecs at full HD frame sizes.
FYI, if you're having trouble with PC playback of HD content, check out the WMP 10 beta. It enables hardware acceleration with the latest from Nvidia, and ATI all the way back to 9500. My Athlon 2800+ went from playing 1080 24p content at 2-3 fps to a rock-solid 24p.
These new standards will work with absolutely no existing players, either in a device or in a computer.
We're talking a new physical form factor here in any case. While today's computers high end computers likely will be software upgradable for playback, but a new drive will be needed.
For playback devices, we're need all new systems, with new lasers, decoder chips, etcetera. This is a leap akin to that from VCD to DVD.
Moreso, I'd expect there to be distributed-as-source implementations of VC-9 before long, like LAME for MP3 and Xvid for MPEG-4. MP3 and MPEG-4, like VC-9, are open standards with royalties.
You do know that this has never been a problem with Real any more than other platforms.
It's just that Real was trying to do real-time streaming back in the modem era. None of the modern formats* should have significant numbers of buffering errors with well-encoded content between current versions of the server and player.
*I'm not counting QuickTime here, since it doesn't have a functional scalability system.
Oh, that's just ridiculous. It's not like we have digital ears. You're basically saying that no analog speakers have audiophile quality, which is patently incorrect.
Saying analog is always bad is like saying digital is always good. 12 feet of 14-gauge copper wire isn't going to cause an audible difference.
Slashdot Mirror
Blu-ray needs the cartridge because it uses a much thinner plastic layer, making it more vulnerable to scratches, grease, etcetera. I'd expect DVD Forum blue laser to have durability not much worse than DVD. In my experience, I have plenty of 8 year old discs that play just fine, as long as the kids haven't been at them.
Oh, heck, you don't even need the bits. Typical HD DVD-ROM WMV9 content is anamorphic 1440x wide, plus whatever height the source media used (they don't encode the letterboxing into the video stream, unlike DVD). So, a typical 2.35:1 Hollywood movie would 1440x816 @ 24 fps. You could do 120 minutes of that in 4.3 GB in a more than watchable format, especially with the Advanced Profile features. I'd prefer to have the bits of a DVD-9 available, of course. I've been able to do decent 1280x720 24p WMV9 @ 3 Mbps, and that was before Advanced Profile.
MPEG-2 has historically cost about $2/player, down from $5/player not so long ago. I think VC-1 is something like $0.25/player for the decoder component. Quite a bit cheaper.
There is the in development Ogg Theora, but it's not done, and not competitive against the VC-1 feature set. I hope it'll be interesting later.
As for DRM, that happens above the codec level, on the stream or file level. You can apply DRM to a Theora file, and can use a DRM free WMV file. The codec itself doesn't have any "native" DRM. And even if Blu-ray uses DRM, there's no reason to think it'll be Microsoft's.
Prediction: You won't be able to buy a non-HD set of more than 23" within a few years. It's not that much more expensive to build an HD set these days. And it's a feature people are willing to pay for.
See many new black and white sets anymore? This changeover will be quite a lot faster.
Care to say why?
They're pretty comparable in many cases, although I think that today's VC-1 implementations do a better job of maintaining small details like film grain than today's H.264 implementations.
Of course, all we can do is compare implementations, not bitstreams, so it's hard to say what the differences between formats are going to be in practice once stuff is actually shipping.
Folks,
Everyone seems to be thinking that somehow VC-1 is more patent or license protected than MPEG-2. This simply isn't true. Now, maybe real-world enforcement of the MPEG-2 patents aren't particularly aggressive for OSS software decoders, but every DVD player, and EVERY DVD DISC STAMPED requires a payment to MPEG-LA. And VC-1 license payments will also be handled by MPEG-LA. MPEG-2 or VC-1, there still will be payments, and the checks go to the same company. The interesting differences here are technical, not licensing.
This announcement is only about the codec. DRM will happen on a different layer of the standard, and won't be codec specific. The same DRM scheme will be used for both MPEG-2 and VC-1 content in the standard, I'm sure.
MS DRM might be used, or might not be. Personally, I think the MS DRM is technically very good. It provides a very rich syntax of business rules for the content owners. Certainly, there are lots of overly punitive DRM business rule implementations out there (think the T2 HD DVD-ROM). But that's really the content creator's fault. The WM DRM definitely allows for very user-friendly copy protection, if that's what content creators want to use.
xvid is based on MPEG-4 part 2, which is roughly as patent encumbered and has roughly similar license fees and terms as Microsoft's VC-1. If xvid is good enough from a licensing perspective for you, so will VC-1 be.
Now, if what you want is an open-source VC-1 encoder, I'm sure it'll happen once the standard is fully finalized, ala LAME and Xvid. The same kind of open-source but unlicensed codec implementation should be perfectly applicable there.
Er, you don't know what you're talking about. A high end software MPEG-2 encoder like Canopus ProCoder is absolutely as good as the high end hardware encoders, and can be faster than real time on a modern machine from a file-based source (which is the typical workflow these days anyway for anything other than news and other live broadcasts).
As for hardware VC-1 implementations, there are a good half dozen companies working on them, and some where being quietly demonstrated in the back suites of NAB back in April. Rest assured, this is happening.
Well, there area already H.264 codecs in development ala LAME and Xvid. I expect the same to happen for VC-1 - after all the specification will be published.
The issue is license free codecs, of which there aren't any competitive examples. There are things in development, but they simply aren't far enough along to compete against H.264 and VC-1 in the current HD standard processes.
Yes, I know about Theora. No, it isn't far enough along to be competitive.
Why do you assume MPEG-2 will be dominant? the VC-1 codec (aka VC-9, aka WMV9 Advanced Profile) can provide similar quality to MPEG-2 at half the bitrate. This means that content providers could do a project with a cheaper, single layer disc instead of a dual layer disc for longer projects.
I believe the big driver behind this is the competition form the DVD Forum's own blue laser format. DVD Forum already has tentative support for VC-1 and H.264. Even though the DVD Forum has lower digital capacity, the support for better codecs meant that DVD Forum could actually get more hours of good quality content on the disc. So equalizing the codecs means that Blu-Ray's capacity advantage can shine.
That said, I'm still betting on DVD Forum. 30 GB will mean more hours of HD content that DVD can do of SD. Also, DVD Forum discs are MUCH easier to convert an existing DVD plant to, and likely will be more durable in day to day use.
Blu-ray seems more likely to win inside cameras and that kind of thing, where capacity is a bigger deal. Think VHS v. Beta, where Beta turned into the Betacam format, giving Sony a 15-year dominance in professional video formats.
A 50% reduction over MPEG-2 isn't anything special these days. Codecs like WMV9/VC-9 and AVC/H.264 can do at least that well today.
Still, "seems to give a two-fold reduction" is a pretty squishy measure. It'll take a double-blind test before we can make a meaningful comparison.
But twice as good as MPEG-2 is a good minimum benchmark for a competitive codec these days.
Well, even if this was released, it's a real stretch saying that it's "only a matter of time" before it'll be widely used by US media companies. A codec is really only a small part of a digital media architecture. Some of the competitive factors that go into these choices include:
Compression efficiency
Cost of implementing decode in consumer electronics (read, what's the cheapest chip that can decode it)
Support for existing transport mechanisms (like MPEG-2 transport streams)
Existence of industrial grade encoders (like massive statically multiplexed encoder arrays)
License fee
The license fees do matter a little, but that's really a secondary issue. A more efficient but more expensive codec can actually be cheaper to implement, because the content provider can use less bandwidth per channel, enabling them to sell more channels over fixed bandwidth.
Today, the battle for the next generation "TV" codec is between Microsoft's VC-9 and MPEG-4 H.264. And that battle is already well underway. The BBC codec isn't far enough along to compete for the current generation of standardization efforts for technologies like HD DVD and digital cinema.
The wavelet codec was Ogg Tarkin. Ogg Theora is a more traditional codec, based on On2's open sourced VP3 codec from a few years back.
Ogg Theora is lurching towards an actual release, and is supported in a few tools like VLC, while Ogg Tarkin never really got very far along in implementation. Theora was meant to be the quick interim release while Tarkin was developed, although the schedule has slipped quite a bit since.
I've gone to Vegas every March/April for around 10 years now, for the National Association of Broadcasters convention. It isn't THAT hot that time of the year, and it's quite pleasant at night. I used to walk all the time, since the cab lines during a major show can be incredibly long - a 20 minute walk is often much faster, and a lot less boring.
:). Heck, these days I stay in the Hilton, which is right next door to the convention center, and it takes literally 20 minutes to briskly walk from my hotel room to a booth in the South Hall.
It's a whole different view of Vegas at street level. It really shows the scale of the Casinos, when you realize you've been walking past Caesar's for the last ten minutes
Oh, I'm not offering an opinion as to the legality of their claim. However, I'm sure someone will try an equivalent effort with VC-9, making the same claim.
There's a difference between MPlayer and Xvid, though, in the difference between distributing source code and a binary. Xvid and LAME make a plausible case they don't count as an implementation since they're just source code. Hard to say if it's true without a judge's opinion, but no one has cared enough to sue them at least.
TI already has a single chip DSP that can decode all the proposed HD DVD codecs at full HD frame sizes.
FYI, if you're having trouble with PC playback of HD content, check out the WMP 10 beta. It enables hardware acceleration with the latest from Nvidia, and ATI all the way back to 9500. My Athlon 2800+ went from playing 1080 24p content at 2-3 fps to a rock-solid 24p.
These new standards will work with absolutely no existing players, either in a device or in a computer.
We're talking a new physical form factor here in any case. While today's computers high end computers likely will be software upgradable for playback, but a new drive will be needed.
For playback devices, we're need all new systems, with new lasers, decoder chips, etcetera. This is a leap akin to that from VCD to DVD.
But the same thing was said of CDs and DVDs, but .VOB files are showing up on P2P now. Consumer bandwidth as been going up up up.
That said, yes, for someone with a reasonably adult ratio of time to money, it's way better to buy a DVD than to try and download one.
Moreso, I'd expect there to be distributed-as-source implementations of VC-9 before long, like LAME for MP3 and Xvid for MPEG-4. MP3 and MPEG-4, like VC-9, are open standards with royalties.
You do know that this has never been a problem with Real any more than other platforms.
It's just that Real was trying to do real-time streaming back in the modem era. None of the modern formats* should have significant numbers of buffering errors with well-encoded content between current versions of the server and player.
*I'm not counting QuickTime here, since it doesn't have a functional scalability system.
Oh, that's just ridiculous. It's not like we have digital ears. You're basically saying that no analog speakers have audiophile quality, which is patently incorrect.
Saying analog is always bad is like saying digital is always good. 12 feet of 14-gauge copper wire isn't going to cause an audible difference.
Well, you know we just tell them that the best place in Portland to live is Troutdale!