Post a Comment
News and New Products
Voices: Microsoft's Amir Majidimehr: a "window" to the world of digital media
By Brian Dipert, Senior Technical Editor -- EDN, 11/23/2006
A Microsoft vice president weighs in on the future of consumer electronics, telecommunications, and more. EDN interviewed Amir Majidimehr, corporate vice president of the Consumer Media Technology Group within the Mobile and Embedded Devices Division at Microsoft. His team delivers Microsoft digital-media technology to mobile operators and consumer-electronics manufacturers and owns Microsoft's digital-rights-management and advanced media strategies, including HD DVD.
CODEC BASICS
Microsoft unveiled its "Corona" audio and video codecs, later renamed the Windows Media 9 series, in late 2001. They delivered significant feature and quality advancements over their predecessors, but five years later, version 9 is still the latest and greatest Windows Media variant. Granted, there have been some minor enhancements to the codec suite: a bit-stream-incompatible audio algorithm that Verizon's V CAST service uses, for example, along with a still-image codec, a broadcast-tailored tweak of WMV9 (Windows Media Video 9), and other video enhancements incorporated as part of the VC-1 (Video Codec 1) repackaging for SMPTE (Society of Motion Picture and Television Engineers). But what's behind the version slowdown as compared with past, faster version rates, when should we expect to see the next major Windows Media revision, what will it accomplish, and will the new version be bit-stream-compatible with prior-version audio and video decoders?
In the early days of Windows Media development, [Microsoft] was mainly focused on Web streaming on PCs. We had the luxury of power and convenience of the Internet to make the decoder download to PCs very easy. Thus, we didn't have to be concerned about the decoder compatibility too much, and thus could come up with new codec formats every year.
But we soon realized that the great strength, quality, and flexibility of Windows Media should not be limited to PCs only. We wanted to broaden our scope to the consumer-electronics industry to ensure content-flow compatibility and benefit more users. In the CE [consumer-electronics] space, decoder compatibility is the basic requirement for any codec. The MPEG-2 decoder for DVD player has been the same for the past 20 years. But the encoding and authoring companies were able to continue to improve the encoding quality during this period given the decoder constraint.
Because of this, when we designed the Windows Media Video 9 bit-stream syntax and decoding process, we spent extra time making sure the bit-stream design was flexible and extensible before we shipped it as part of Windows Media Video 9 Series so that we can continue to make encoding improvements. We also came up with the Advanced Profile of Windows Media Video 9 to improve the interlaced mode and made the bit stream even more flexible, so that the codec elementary stream could be mapped to any file container such as MPEG-2 Program Stream and Transport Stream. We then submitted the entire spec to SMPTE for standardization, and it became VC-1. The effort of standardization is enormous. We believe making this available through standardization so that the users can have yet another advanced codec as a choice will benefit the consumers a lot.
Since then, we have made a lot of quality and speed improvements on the encoder side for various applications. We submitted WMV9/VC-1 to the DVD Forum for consideration as a mandatory codec for the [next-generation] high-def DVD format. We won both DVD Forum tests conducted in late 2002 and early 2003. This led to VC-1 being chosen as one of the three mandatory codecs for both HD DVD [high-definition DVD] and BD [Blu-ray Disc]. From that point on, we have also worked very hard with Studios and post houses to continue improving the VC-1 encoder for HD encoding. We have also developed very powerful VC-1 encoding and analysis tools for HD production. This VC-1 encoder and tool set has become THE tool that HD DVD and BD post houses are using today for HD production. Consumers are greatly enjoying the excellent picture quality of HD today. It is all a result of the work we put up on this.
We have also continued to ship the encoder improvement we did as part of Format SDK/Windows Media Encoder. For example, we shipped two new encoder releases, one as part of Format SDK 9.5 in 2004, and one in Format SDK 11 in 2006. We have received great feedback from users that the latest encoder in Format SDK 11 has much better quality and encoding speed compared to the previous versions. We have also enabled quad-thread encoding so it can saturate a fast dual-core, dual-socket workstation.
On the audio side, we froze the syntax of Windows Media Audio in 1998. After eight years of development and encoding only work, we were able to make another 10% improvement in WMA [Windows Media Audio] for WMP 11 [Windows Media Player 11] over the last version. But we also needed to continue to look at other scenarios and see if we need to come up with a more advanced codec. As you mention, Verizon is using the Windows Media Audio 10 Professional codec's new low-bit-rate modes. These provide around a 2× improvement in compression efficiency over WMA for the 96 and below kbps range. WMA Pro at 96 kbps is equivalent to a WMA Std 160-kbps version… It's that good. We've got an independent study showing it's better than HE AAC [high-efficiency advanced audio codec], as well. And I'm really excited we were able to implement it in a backward-compatible way. The core of the WMA 10 Pro LBR [low-bit-rate] stream is straight-up WMP9 compatible WMA 9 Pro, plus some extra frequency-synthesis data. The content will still be playable on older players, just not with the full frequency range.
Tell EDN's readers a bit about your efforts at getting digital cameras to support the Windows Media still image codec in addition to, or instead of, more traditional formats, such as JPEG and proprietary RAW schemes (along with Adobe's standardized Camera Raw format). Where do you currently stand on obtaining this support? Why should EDN's readers consider Windows Media Still Image? And what about Windows Media as a video capture format for these cameras, versus today's more prevalent MPEG-4 Simple Profile/Quicktime, along with other common formats and their wrappers?
Windows Media Photo is a new file format for end-to-end digital photography that provides many capabilities not provided by other file formats. The format is being introduced with the launch of Windows Vista, and also provided for Windows XP and Windows Server 2003 via .NET Frameworks 3.0. We also offer a Device Porting Kit for hardware and alternate platform partners. This DPK [Device Porting Kit] provides complete documentation of the file format and reference ANSI-C source code that is compatible with both little-endian and big-endian architectures.
To date, we have only provided information for developers to allow the earliest possible support for this new format from third parties. When Windows Vista is released for consumers early next year, I'm sure many partners will have information to announce regarding Windows Media Photo. Until then, we're under non-disclosure and cannot speak on their behalf.
Windows Vista introduces a new graphics architecture—WIC (Windows Imaging Components)—that replaces the imaging functionality of GDI+. WIC is the underpinnings of WPF (Windows Presentation Foundation) in Windows Vista and .NET Frameworks 3.0, and (among many other advance features) it enables extensible still image codecs. Microsoft or third parties can add new still image codecs to Windows Vista at any time; the operating system (and any application that takes advantage of WIC or WPF) will recognize and support these new formats. One of the most exciting applications for this feature is the support for camera RAW codecs. We expect the leading digital camera manufacturers to provide RAW codecs that will add native support for their proprietary file formats to Windows Vista and any WIC or WPF enabled Windows application.
With WIC, Windows Vista and .NET Frameworks easily support the still image codecs required for any customer scenario. Windows Media Photo complements existing codecs and provides new capabilities including dramatically improved compression, high-dynamic-range wide gamut support, compressed domain operations and advanced decoding modes.
Windows Media Photo provides an advanced high-performance compression technology producing photos of higher quality than JPEG at half the file size. By using Windows Media Photo, cameras and other photo-capture or -viewing devices will use less battery power and be able to store many more photos that will save, display, and transfer faster.
Windows Media Photo is an integral component of the new printing architecture in Windows Vista, and enables end-to-end high dynamic range wide gamut printing. For the first time, a user can print a photo, maintaining the full tonality and color gamut from native color space of the camera all the way to the native color space of the printer. Unlike most current workflows, the photo's dynamic range or gamut is never reduced to an intermediate working color space (such as sRGB.) An image may either originate in the Windows Media Photo format directly from the device, or use the device's native high dynamic range wide gamut color space via an installable codec. The image will be converted to Windows Media Photo for inclusion in the printer spool file, using XPS (XML Paper Specification), the new XML-based page description language introduced by Microsoft and shipping with Windows Vista. An XPS compatible printer or printer driver uses the services provided in Windows Vista to convert the high-dynamic-range wide-gamut photo directly to their native printer color space with no loss of fidelity. At WinHEC, we clearly demonstrated how this new XPS printing architecture visibly improved the quality of photographic prints while simultaneously freeing the user from many of the arcane details of color management and image workflow.
Windows Media Photo is also a key enabler for a new generation of Web-based imaging solutions. At SIGGraph this year, Microsoft debuted a prototype of a new community photo service called PhotoSynth. This powerful photo viewing solution is enabled in part by the use of Windows Media Photo to store the photo library. The advanced decoding capabilities of Windows Media Photo allows the server to efficiently produced reduced resolution previews or smaller rectangular regions of the individual high resolution photos without ever decoding the compressed image data. These compressed image elements are delivered to the client with the smallest possible amount of data, where they are decoded for reassembly and display. The result is a high-performance, seamless viewing experience for hundreds or thousands of related photos. The advanced decoding capabilities of Windows Media Photo will enable an entirely new class of Web-based imaging solutions; PhotoSynth provides a first glimpse at the type of services that will be possible.
|
One fundamental design decision that EDN's readers grapple with, in incorporating multimedia encoding, decoding and transcoding functions into their systems, is the degree of hardware-versus-software emphasis that they should implement. On one end of the video spectrum, for example, is a hardwired single- to few-codec chip, such as from Ambarella. On the other is a fully software-based processor, such as Texas Instruments' DaVincDSPs. And between these two extremes are FPGAs, along with a plethora of chips containing generic processor cores tightly coupled to codec-specific hardware-acceleration blocks. What guidance can you provide as to which points along the spectrum are appropriate for various application criteria such as performance, power consumption, cost, and current and future codec flexibility?
That's a very broad question! The choice of processors and chips depend a lot on your applications and scenarios. The good news is that we don't have to make a bet on that. We provide porting kits including an ANSI C version and several optimized ports which a number of different companies license and implement VC-1 or the whole Windows Media stack on top of. These range from multiple FPGA and DSP solutions like Tandberg, to simpler appliances like VBrick's, and a forthcoming hybrid hardware/CPU like you describe.
On the decoder side, since the bit stream and decoder have been fixed for many years, most of the companies use the hardwired solution to reduce the cost and power consumption. Companies like Broadcom, Sigma Design, and ST Micro have done the work and shipped their chip solutions.
On the encoder side, we are seeing more varieties. DSPs win for cost, obviously. FPGA's offer a good mix of performance, price, and upgradability. And a straight up x86 CPU remains compelling for many tasks, especially as we enter the multi-core era. We've got customers implementing on all of the above.
I guess my answer is that I'm as excited as you to find out the answer to that question.
OPTICAL STORAGE, AND DOWNLOADABLE/STREAMABLE MULTIMEDIA
High-resolution MPEG-2, MPEG-4 Part 10 (aka MPEG-4 AVC, and H.264) and WMV9/VC-1 have risen to the top of the heap, at least for high-definition video, and at the expense of other codec contenders such as On2 VP7, RealVideo, and DivX. Will it continue to be a three-horse competition going forward, and how do you see the race evolving as time goes on?
VC-1, MPEG-2 and H.264 have all been standardized and are open. Because of this, companies are more willing to deploy solutions using them. They have also undergone rigorous testing over the past few years as was the case when codecs were being evaluated for inclusion in the new optical disc specifications. MPEG-2 is mainly there for backward compatibility whereas the advanced codecs (VC-1 and H.264) are the future. We will continue to see a great deal of new product offerings utilizing these codecs now and in the future.
On the other hand, the other three that you mentioned are proprietary and have not proven themselves in the HD arena. They can certainly handle the resolutions that HD requires but they aren't as efficient in doing so. As these codecs improve over time, they will likely find their niche like Web downloading/streaming at some point but VC-1 and H.264 will be improving as well. So we should expect that both VC-1 and H.264 will continue to be the mainstream codec choice for HD content. And from the early signs of HD DVD and BD production and quality comparison, VC-1 is likely to be the primary force. I don't see VP7, RealVideo or DivX taking a meaningful role here any time soon.
Assessing the quality difference between MPEG-2, H.264 and WMV9, and between the Blu-ray and HD DVD optical formats, has so far proven difficult to impossible, even on identical-title material. There are just so many variables—different hardware platforms from different manufacturers with different video processing settings, different encoders, different bit rates, different native resolutions and progressive-versus-interleaved options, different player-to-display interfaces, and others. Given that both blue-laser formats require playback support for all three codecs, what criteria should guide EDN's readers' video codec selection for encoding and other related codec-flexible applications, such as high-resolution camcorders?
If you look at the major studio releases today they are predominantly encoded with VC-1. In fact 99% of all HD DVD releases are VC-1 titles. Blu-ray's use of MPEG-2 for their releases was far from impressive as evidenced by feedback from their users. So HD DVD has received overwhelmingly better feedback on picture quality than Blu-ray (BD). All of the studios have a choice and most are choosing VC-1. Their preference should serve as an indication as to which one provides the best quality.
Secondly, one must consider the encoding tools available and whether they are cost-effective. There are currently only a handful of options here and more will be available over time. These tools will be all over the spectrum as far as features and flexibility is concerned. Some will allow you to quickly encode media at a limited number of settings and be done with it. This will result in media that doesn't look as good as it could. Others will take their time to do it right and provide a plethora of options for conducting third passes to address difficult scenes. The feedback we've been getting from the studios is that the VC-1 encoder we've been providing them is by far the most flexible and powerful encoding solution they've ever seen, and we're aggressively enhancing it on a frequent release schedule, providing better default initial encoding, and more and more tools to tune the encoding.
In comparing quality, the key is to make it as apples-to-apples as possible, We've certainly seen a lot of MPEG-2 BD titles that some reviewers claimed looked great that show painful and obvious artifacts on a good 1080p display via a digital connection. So a good full-resolution 1080p display is the first step. Beyond that, the player differences have been overstated. Early on, some were blaming the poor quality of BD titles on a noise reduction circuit in the Samsung player. But that same player, without modification, produces identical quality to a HD DVD player when playing one of the Warner titles that use the same VC-1 encode for HD DVD and BD. So far, we've found that the better the playback environment, the more palpable the advantage in quality in VC-1 titles over MPEG-2 titles has been.
Now, there hasn't been much in the way of AVC titles so far, so we don't have a lot of real-world evidence to cite. But looking at the basic structure of the codecs, AVC has some intrinsic design in its codec that would lead to softening the image. This effect would remove/reduce the film grains and the detailed texture in the picture.
The marketplace standoff between Blu-ray and HD DVD shows no signs of ending any time soon. Do you agree with this pessimistic view of current events? How long do you think it'll last, with what end result, and why? And what role do you think high resolution video (using a modern codec such as H.264 or VC-1) on red laser media might play as a 'bridge' technology until the blue laser picture is clearer?
We're very happy with how HD DVD is doing so far. From reviews to blogs to the AVS Forum, we're seeing on par a much better response to HD DVD. The players are better, there's a larger selection of titles and there's a much higher consistent quality of the titles. I'm also seeing positive response to our decision to offerHD DVD playback as an accessory for the Xbox 360 instead of making every customer pay for it as part of the core console. We're also seeing consumers buying far more HD DVD players and titles than Blu-ray—the latest Nielsen and Video Scan numbers are showing an amazing attach rate for HD DVD titles.
So, I'm very pleased and optimistic about where things stand now.
I can't predict the future, but so far, HD DVD has delivered extremely well, and much better than the competition. We're going to continue to help the studios make great looking titles with great interactivity. And the success of HD DVD certainly hasn't been lost on content owners—we're seeing a huge amount of support for the format, from studios down to smaller independent producers.
We aren't seeing any sign of red laser media being considered for shipping HD titles by either HD DVD or BD studios.
CinemaNow's "Burn to DVD" doesn't seem to support Windows Media codecs, which is surprising given that the "Rent" and "Buy" options from both CinemaNow and Movielink do, along with implementing Microsoft's DRM [see "Cinema...Naw," from the Brian's Brain blog, Sept. 26, 2006]. Do you know what factors drove this change?
Microsoft is working closely with the content service providers and the DVD CCA [Copy Control Association] to enable consumers to download movies protected by WMDRM [Windows Media Digital Rights Management] and then burn them on a Windows PC to a protected DVD that's compatible with the millions of CE-based DVD players in the market today.
At the Audio Engineering Society Convention in early October, two Microsoft engineers presented a visionary argument as to why, as fiber-to-the-home and other high-bandwidth broadband connections become more pervasive and as hard-drive storage becomes ever-cheaper, downloadable standard- and high-definition audio and video media will increasingly dominate at the expense of traditional packaged media. Do you agree with this assessment? And if so, how will it transform not only the means by which we obtain and consume media, but also the kinds of media we obtain and consume?
Well, first off, there's definitely still a place for new packaged media. We've been supporting the HD DVD format for example. With discs at 30 Gbytes and beyond, it can be inconvenient for some to download. Another factor is that online delivery today usually means more compression which means lower quality. Those looking for the optimal experience will continue to look toward shiny discs until the pipes are bigger.
There are a lot of great new models enabled by the ever-improving combination of faster connections and more efficient video and audio compression. In the digital media world, audio is always about five years ahead of video, since the data rate and decode performance requirements are so much less. So, I think we'll see electronic delivery of video evolve the way we saw it for audio. I know it's an overused phrase these days, but it's also a great long-tail play, since the publisher doesn't need to make a bet on the title for replication and distribution. They can just make the master, and if it doesn't sell, they're just out the cost of preparing the assets.
For consumers, it lets them get immediate access to a much bigger library of content. And it enables a lot of new models other than buying or renting physical media. Subscriptions get a lot more interesting, for example. Advertising gets a lot more interesting, since you can have dynamic advertising. And as we move towards a connected household for the entertainment experience, you don't need to walk to disc between rooms—you just have it on the hard drive in your Media Center, and the kids can watch it from their Xbox 360's in their bedrooms. And you can copy it to a portable media player to watch in the car to the cabin, and the plug into a TV at the cabin.
We're in the nice position of having our technologies work well in a wide variety of uses, and letting content and device companies do a lot of great innovation around business models, using our technology as the base.
Back in the "old days" (a year or two ago) the streaming multimedia market was generally concentrated on three codec/wrapper combinations: Apple's Quicktime (generally in conjunction with Sorenson's codec), Microsoft's Windows Media, and RealNetworks' RealMedia. Now, the picture has shifted significantly; Adobe's Flash Video (employing an On2-developed codec) has risen to prominence, and vendors such as Google and recently-Google-acquired YouTube are using proprietary schemes. What's behind the shift, and what does the future hold here?
Yes, the choice of Web video formats has really gone from one between Windows Media, QuickTime, and RealMedia to between Windows Media and Flash. The biggest factor to that shift is installed market share—Flash is second only to Windows Media player in installed base for web video playback. Each technology has different strengths. Windows Media today has a growing lead in codecs (audio and video), quality and pricing for servers and hosting services, and ecosystem of content creation tools. Flash has some advantages in rich media integration on non-Windows platforms today, but we're working hard on providing an even better solution there.
WINDOWS VISTA AND OTHER OSs
What "under-the-hood" improvements with respect to multimedia (of all sorts) processing (encode, decode, surround sound virtualization, virtual bass, and noise suppression) have been made to Windows Vista compared with today's dominant Windows XP? And how do you balance any additional built-in processing against your software partners' desires to continue to bring added value to the platform via their multimedia processing products?
Vista is major step forward for media consumption and interaction. We added a number of DSP components to the operating system that make for a more engaging multimedia experience. For example, we added room correction technology that quickly assesses the acoustics of the playback environment and adjusts the characteristics of the audio pipeline to deliver a balanced experience as if you were in a specialized room. Prior to Vista, you would have needed to spend $15,000 on a dedicated hardware device to perform the same function. This is but one of many examples.
The audio architecture is based on what we call "local effects" and "system effects." Third parties can replace these effects to provide benefits to only their software or all applications on the system. Feedback from software and hardware vendors show they are excited about the possibilities this new architecture provides.
Where are you with Windows Media support in Linux, both generically and specific to DRM-inclusive media, and how do you see that situation evolving over time? More generally, how can EDN's readers designing systems based on non-Windows (including non-Windows CE) operating systems work with Microsoft with respect to format and DRM support?
Companies can license Windows Media codec and DRM technology today if they are looking to build portable or network-connected devices that can play back content protected using Windows Media DRM. The licensing terms for Windows Media support on non-Windows devices are published on the Web—and dozens of companies are shipping devices today that support Windows Media DRM.
In terms of implementing Windows Media DRM on a general-purpose PC operating system (such as Linux)—that is more technically challenging, because users can easily install applications and software (potentially including programs designed to circumvent DRM) on such a machine. In order to be a viable solution, the Windows Media DRM implementation must satisfy the requirements of content providers in three specific areas: robustness, compliance, and renewability. These issues would need to [be] fully addressed—which isn't the case at present—before we could consider developing a licensing program for Windows Media DRM for general-purpose operating systems.
When, if ever, will we MPEG-4 support be built into Windows Media Player? And how do EDN's readers obtain MPEG-4 (Simple Profile, Advanced Simple Profile, Part 10, etc) support in the meantime?
I can't comment on future product plans. But MPEG-4 is a huge spec, and to my understanding, no one company has implemented the complete spec/codec of it. And we know Advanced Simple Profile is rarely used in any application/scenario. Even QuickTime, the most installed MPEG-4 playback platform doesn't support all modes/profiles yet like Advanced Simple and H.264 High Profile.
But we will always carefully examine the standards and popular scenarios in the industry and properly support them in our products. We want to ensure users have the best experience using our products like Windows Media Player. For example, we have been supporting the playback of MPEG-4 Simple Profile in ASF file for many years, since there are cameras producing this format. We will support MPEG-4 Part 10 decoder in the scope of HD DVD playback in the upcoming Xbox 360 update for HD DVD player this winter.
There are plenty of third party codecs that can install MPEG-4 playback for Windows Media Player, but they all have their own limitations. I'd hesitate to recommend any in particular.
IPTV, DVB AND 3G CELLULAR
SBC, now AT&T, delivered a glitzy demo of its Project Lightspeed IPTV plans at Bill Gates' 2005 Consumer Electronics Show keynote. Yet, nearly two years later, IPTV is still stuck in limited-trial status (at both AT&T and Verizon), at least in the United States. Meanwhile, cable- and satellite-television services continue steamrolling along. Are there other North American deployments of IPTV that EDN's readers should be aware of? What's the situation elsewhere in the world? What's behind the delay, and how do you anticipate the rollout will progress through the remainder of the decade?
We will soon begin to see IPTV deployments scale upwards at the end of this year and into 2007. Results have been positive, and as such, the world's premier telecommunications companies are stepping forward to make IPTV a reality. These companies include AT&T, Bell Canada, BellSouth, BT, DT, Reliance Infocomm, Swisscom, Slovak Telekom, T-Online France, Telecom Italia, Telkom South Africa, Verizon (hybrid), and YA.COM. The companies (as well as some unannounced ones) collectively represent 26% of the world's residential fixed access phone lines, 28% of the world's xDSL subscribers, and 75% of fixed access lines in both Europe and the United States. Based on this, we are comfortable with IPTVs position and are encouraged with the direction things are heading.
As IPTV deployment schedules begin to map out, analysts are beginning to adjust their projections and predict even higher growth than initially expected. Proof points:
- " Worldwide subscriptions to Internet Protocol TV will rise from 3 million in 2005 to almost 49 million in 2010. Revenue will also grow rapidly, from $401 million in 2005 to $13.2 billion by the end of the forecast period. (Elroy Jopling, Gartner, August 2006)
- " In 2009, there will be 36.9 million IPTV subscribers worldwide.(Multimedia Research Group, October 2005)
- " IPTV will have 2.7 million subscribers worldwide by the end of 2005, but this number will have risen to around 25.9 million by the end of 2010, according to the report, IPTV: A Global Analysis, published Thursday. (Informa Telecoms and Media)
Microsoft's secured some trendsetting design wins in the wireless-multimedia-delivery space: Verizon's V Cast, Modeo's DVB-H service, and Sirius satellite radio's fledgling video vision. (Are there any other notable examples?) Yet consumer-survey conclusions are, at best, mixed as to potential users' perceived value of being able to watch television on mobile systems' miniscule screens. How do you think the mobile-TV and -movie revolution will roll out and end up? And how do personal place-shifting products, such as the Windows Media-based Sling Media SlingBox series, capable of streaming content to those same mobile systems, affect service providers' future fortunes?
Everyone in the industry knows how big these things are going to be eventually, so it's easy to fall into the trap of being disappointed by each individual step forward. Think about Web video, portable music players, the PC as a media playback system. All had a few years of growing pains, and all are huge markets today. We're really happy with how this market is evolving. And I think you hit on the reasons why these are good long-term bets—they enable the customer. Getting a great audiovisual experience onto mobile devices, and freed from the space and time constraints of historical broadcast television, both let the consumer make freer choices about when, how, and where they're going to enjoy their content.
XBOX 360
Xbox 360 can stream audio and still-image data from UPnP (universal plug-and-play) systems such as Windows XP PCs and NAS [network-attached-storage] boxes. But currently it's only able to play back video content stored on Windows Media Center Edition PCs. A broad range of codec support already exists in the console (or can be handled via on-the-fly transcoding on the PC the content's stored on). What's behind the Media Center-only restriction for video? And will it always exist?
We have no announcements relative to any planned changes in Media Center Extender related functionality for Xbox 360 but are always seeking ways enhance our offering. We'll keep you posted on any announcements we make in the near future.
[Editor's note: As of the Fall Dashboard Update, streaming of non-DRM'd video content from generic UPnP storage devices has been enabled, and Microsoft is considering adding support for DRM'd video material in the future. For more information, see my blog post "Microsoft's Xbox 360: A Maturing Multimedia Marvel" —Brian Dipert]
Red-laser disc-stored, WMV-encoded video content is available from a few sources including CinemaNow, Google, and Movielink, for example, as well as Broadcast.com, and IMAX (see Microsoft's WMV HD Showcase). And the Xbox 360 even provides an Internet connection that could implement the DRM-validation process. But these discs can only be played back on PCs; the Xbox 360 can't handle them. Why? Will the HD-DVD drive for the Xbox 360 remove this restriction? And how will the Xbox 360's HD-DVD drive handle future media that, courtesy of the not-yet-implemented Image Constraint Token flag, instructs the player to downscale video traveling to a display over nondigital, unprotected links?
WMV HD was designed primarily to work on PCs. We expect that most all WMV HD content will be available in HD DVD. There are many ways to experience HD content on Xbox 360, including our lineup of high-definition games and content available on Xbox Live Marketplace. For those gamers who want to experience HD DVD, we also offer the HD DVD Player separately. HD-DVD playback in Xbox 360 will comply with the requirements of the Image Constraint Token; however, it is our hope that this will not be an issue over the lifetime of this generation of console
Earlier in this interview, you were asked about downloadable media. A few television programs, music and promotional videos, and other video clips are available for download from the Xbox Live Marketplace. But for the Xbox 360 to be a serious stand-alone download-and-playback device, it will require a much larger hard-disk drive. Do you agree? For now, will the Xbox 360's predominant use be as a Windows Media Extender to a Media Center Edition PC? Or are larger hard-disk drives, or perhaps support for USB-tethered stand-alone hard-disk drives, on the console roadmap?
No. Xbox 360 is first and foremost a video-game console and is a pioneer in bringing high-definition content to the gaming community. Choice is at the heart of what we offer gamers this generation. Choice in games. Choice in entertainment. While we don't have any news to announce at this time, we are always looking to enhance the gaming and entertainment experience for gamers and will continue to expand the choices they have to make that experience their own.
We haven't confirmed any additional hard drives for the Xbox 360 beyond what's available today, but with the removable hard drive, we've built Xbox 360 to allow the flexibility for a larger hard drive in the future.
ZUNE
Microsoft's vertically aligned Zune infrastructure (with one company providing the hardware, DRM, and content) is strongly reminiscent of Apple's iPod/iTunes approach and stands in stark contrast to the PlaysForSure ecosystem that Microsoft was previously 100% committed to cultivating. Also, Zune's DRM is incompatible with PlaysForSure; anyone who's bought DRM-inclusive PlaysForSure content will be unable to play it as-is on a Zune device. What's behind the change, or perhaps alternatively stated expansion, in the company's stance? What should EDN's readers conclude about Microsoft's commitment to PlaysForSure going forward? And, more generally, how can EDN readers reconcile partnering with Microsoft on codecs, DRM, and other features while simultaneously competing with Microsoft in the marketplace?
Microsoft continues to be dedicated to supporting and cultivating the PlaysForSure platform and ecosystem. We are proud of our work with partners and the products delivered out of those relationships and we'll continue to work closely with them to make Windows a great platform for any digital media solutions. For example, this past year we released updates to PlaysForSure to incorporate certification for mobile phones and we are working with a wide range of device partners who are bringing exciting new devices to market. Just 18 months ago, we launched with a handful of devices and partners; today, the program has grown to include more than 140 devices from over 20 partners.
In terms of the Zune and PlaysForSure relationship, both platforms deliver entirely different experiences and value. With Zune, our goal is to provide consumers with an easy-to-use, connected digital experience through which they can discover and share new music. To do that effectively, we feel it is important to utilize an integrated and internally supported end-to-end model. Windows is the preferred platform for companies who want to deliver digital media experiences whether through dedicated, end-to-end managed systems like Zune, or devices and/or services, such as Urge or Creative, which favor compatibility with multiple devices and services. Ultimately, these varying approaches to delivering consumer media experiences stimulate innovation and provide customer choice.
By press time, Zune players should be in stores, and the Zune music download service should be up and running. But the Zune movie download service probably will still be under wraps. What's behind the delay? How much of an impediment is Steve Jobs' presence on Disney's board to Microsoft's desire to sell Disney content? And how much of an impediment are the antidownload grumblings of important Zune hardware retailers, which are also large brick-and-mortar DVD retailers, such as Target and Walmart?
Video has always been a key part of the Zune vision. In fact, the product is designed to give consumers a great video experience with a big, bright, horizontal screen. And, Zune does support any video content you have downloaded to your PC, including formats such as .WMV, MPEG-4 and H.264.
While we think video will ultimately be a compelling feature for consumers, and a video store is something we are exploring, the video market is still evolving. When we took a deep look at what consumers really wanted today, we think there is a lot more to we can do with music—in the most immediate future, taking the music experience to the next level with wireless sharing. The great thing about Zune is its built-in wireless technology and powerful software provides a strong foundation to continue to build new shared experiences around music and video.
Downscaling (as well as, possibly, transcoding) high-resolution video content (originally formatted for PC or HDTV viewing) that you purchase online for Zune's smaller screen and lower capacity storage is, I'm guessing, an important feature you'll be rolling out. Correct? Yet you must accomplish this feat while preserving DRM across both source and destination files, which hasn't yet been feasible. You can't, for example, translate a 192-kbps Yahoo Music Unlimited audio file to 96 kbps for playback on a limited-capacity portable player whose headphones, listening environment, and other attributes are amenable to lower quality content. How will you handle this situation in the Zune era?
Currently, Zune supports transcoding of content that is in the clear. Zune video profiles have been carefully defined such that they provide a good experience on the device as well as on the PC and as such no alterations of the media are necessary.
DRM
Microsoft is currently engaged in a cat-and-mouse DRM patch-and-hack tussle with "Viodentia," the developer for FairUse4WM. Viodentia claims that his sleuthing has uncovered a fundamental flaw in the Windows DRM scheme (see this interview on Engadget) that can be solved only by recreating it from scratch in a manner that is not backward-compatible. How do you respond to his claims, technically, legally and otherwise? How concerned are your hardware and content partners about Viodentia's accomplishment? And will we ever get to the point where DRM schemes that are "revokeable and reissuable" are no longer required?
We are responding to this Windows Media DRM circumvention with both technical and nontechnical measures. We have known since we first designed Windows Media DRM that, as with any DRM system, it would likely become the target for hackers at some point. That is why we made sure to design the system in such a way that its protection could be renewed in the event of a circumvention. So our technical response involves using renewability to update the system to address the versions of the tool that Viodentia has released. Once an update has been developed and tested we work closely with our content partners around the world on its deployment.
Additionally, we have filed a copyright infringement lawsuit in federal court, in which we allege that this tool infringes on Microsoft's copyrights by directly copying Microsoft's computer code (consisting of compiled functions protected by Microsoft copyright that shipped as part of the Windows Media Format SDK) without permission. Our goal is to determine who authored and distributed this tool, and, in the interest of our customers and partners, stop its further use and distribution.
Looking to the future, I think it's inevitable that widely deployed DRM systems will continue to come under attack from hackers—that has been our assumption since the first release of Windows Media DRM back in 1999, and I see no evidence that the assumption is incorrect! So we will continue to back up the renewability in Windows Media DRM with the 24/7 breach response team that we have in place, with world-class dedicated program management, developers and testers, so we can move as quickly as possible to renew Windows Media DRM if and when circumventions occur.
Many folks who have tried out subscription music services end up enthusiastic promoters of them, but the lion's share of the music business remains one-time purchase. How do you see this situation evolving over time, both generally and specifically as it relates to the fact that subscription music is something that Microsoft and its partners currently offer, and iTunes does not? And is the subscription music situation analogous to what you expect will occur with video and other types of subscription-amenable content?
That's a question I've asked myself. I think the newness of coupling subscription to music has been the biggest issue. We're definitely providing great value here. MTV's URGE service is $129.95 a year for unlimited music downloads, playable on three different computers and all your PlaysForSure devices. You can pay that much a month for TV content that is much less portable (although we'll be offering Media Center Integration with Comcast and DirectTV with Vista). And that price is very competitive with buying just 12 albums a year via download. One of our employees made an interesting observation discussing this—that it's really a generational thing. Folks who are 25 or older grew up collecting music, where shopping in record stores, getting collectors editions, etcetera, was a big part of the experience of being a music fan. They're used to that sense of hunting and possessing. It's really the younger folks, who are used to music always having been available as bits, that get the subscription model. It's once you've been buying songs one by one, and realize how much you spent of stuff that you don't wind up listening to, that subscription starts making sense. It takes away the risk of exploring a new artist, or a whole album you liked a single off of. It's a buffet instead of a cafeteria, which is what you want you're really hungry.
IN CLOSING
Microsoft is predominantly known as a supplier of operating systems and application software for PCs. Yet, most EDN readers do not design PCs, although they may reference PC hardware in their system designs, their designs may interact with PCs, or both. What opportunities do you see for Microsoft to work, and to work better, with EDN's readers? What will you contribute to the effort, and what would you like to see from them in return?
As you see from the answers and topics in this interview, technologies developed for PCs are finding use well beyond it. This includes our advanced audio/video compression technology, which ships in millions of consumer-electronics products of every sort, and rights-management software, which will power the digital content commerce of the future. There are many opportunities to build electronics that power these technologies in non-PC products such as hardware encoder/decoders for VC-1 video codec for all applications from IPTV to consumer camcorders. And building efficient implementations of these for mass volume products such as cell phones. And as PC products converge more and more with A/V devices, the same components can power both type of products.
