Amazon and Apple's dramatically different strategies

Today's announcement of three new black & white Kindle models, plus the Kindle Fire tablet, have shaken up both Amazon's competitors and a number of industry observers. The Kindle 4 models (the Touch, Touch 3G and just plain Kindle) each set new low price points for eReaders with their functionality. The Kindle, at $79, will be a stocking-stuffer for the coming Holiday season. The Kindle Touch Wi-Fi, at $99, provides comparable functionality to Barnes & Noble's Nook for $40 less, and the Kindle Touch 3G adds always-on wireless connectivity. for $149. (All these prices are for Kindles with Special Offers, or in other words, advertising. Add $30 to the Kindle and $40 to the Kindle Touch models if you don't want ads.)

The Kindle Fire tablet, at $199, is cheaper than any brand name Android tablet except Lenovo's forthcoming A1, which comes with front and back cameras and GPS, all of which Amazon leaves off. However, Amazon has built its own user interface on top of Android that's designed to make the Kindle Fire much easier to use than Google's standard Android. Amazon has also built its own browser, called Silk, that uses Amazon's cloud services to pre-render web content for better performance.

Everything about the Kindle Fire demonstrates how radically Amazon's and Apple's strategies differ. Apple uses software to sell hardware; the iTunes Store and App Store are there to increase demand for Apple's hardware. Apple makes money from software and content, but it's a drop in the bucket compared to its hardware revenues. Amazon, on the other hand, uses hardware in order to sell its goods and services. It's likely that the new Kindles, including the Kindle Fire, are all being sold at close to break-even or possibly even at a small loss, in order to generate more sales of other goods and services. Amazon sees the Kindle Fire as potentially being as stimulative for music, movie and television show sales, as well as Amazon Prime subscriptions, as the black & white Kindles have been for eBook sales.

There's not a lot of daylight between Amazon and Apple for competitors to exploit. Amazon is setting price expectations at the low end and is trying to dominate the content market; Apple dominates the developer community, and the iPad will continue to have a much bigger selection of apps than any of its competitors. Can competitors sell more expensive tablets than Amazon with a mediocre selection of content, or less expensive tablets than Apple with a mediocre selection of apps? I doubt it.

This would have been a great opportunity for a radically different tablet, like Microsoft's Courier, that could compete in a completely different market segment, but Microsoft killed the Courier, and there's nothing on the horizon from first-tier competitors that can forge its own path. In hindsight, HP was probably right to kill the TouchPad, and I wouldn't be at all surprised to see other companies rethink their entire approach to the tablet market. Cloning Apple or Amazon won't work; competitors need radically different products.

Who buys Hulu? Most likely, no one

Silicon Alley Insider is reporting that, now that Hulu's auction is completed, the company's owners have some hard decisions to make. Comcast, News Corporation, Disney and Providence Equity Partners were looking for well north of $2 billion for Hulu, but they didn't get it. More accurately, they got it, but not in the way they wanted.

(Update, October 13, 2011: AllThingsD has reported that Hulu's owners have called off the company's sale and will continue to manage it themselves.)

The top bidder for Hulu was Dish Network, which bid around $1.9 billion dollars, more than either Yahoo or Amazon. Google apparently offered far more--around $4 billion--but the company wanted guaranteed access to Hulu's owners' content for much longer than the two to three years that had been offered. Without that kind of concession, the Hulu deal is really a two to three-year non-exclusive license to its content, not an "acquisition" in any real sense.

That's why Dish, Yahoo and Amazon weren't willing to spend even $2 billion for the company. Hulu's partners could more than double Dish's bid overnight by accepting Google's terms, but I don't think they will. They believe that their content, and the investment they've made in the Hulu platform, is worth more than $1.9 billion, and they're not willing to extend longer terms, given the rate of change in the online content market. Therefore, it's most likely that they'll cancel the auction and keep Hulu themselves.

Poof! Your set-top box has turned into a brick

Yesterday, Sezmi, a Silicon Valley-based video service provider, announced that it is discontinuing its services in the U.S. in order to focus on selling its platform to international service providers. Sezmi used a combination of an Internet connection and digital broadcast signals to provide a subscription package of broadcast, cable and Internet-only channels. While the service was eventually offered in 36 U.S. markets, the cable channels were never made available outside Los Angeles (even though customers in other markets were told they would get them), and the cable channels in Los Angeles were discontinued last December. Subscribers purchased a $150 package including a set-top box with a 1TB hard drive, and a digital antenna for picking up over-the-air broadcasts; the monthly price was $5 for basic service, and another $20 for the cable channels.

Sezmi's unique selling proposition was that it was a low-cost replacement for cable service, but once the cable channels went away, Sezmi lost its primary selling point. In addition, if subscribers lived too far away from local television transmitters to get a good digital picture, they couldn't use Sezmi, either. Sophisticated industry observers could see that Sezmi was doomed, but consumers didn't necessarily have that insight.

Consumers who bought into Sezmi are stuck with bricks that will be useless for anything except watching YouTube by November 1st. Sezmi isn't offering any refunds for hardware, no matter when it was purchased. Subscription video services fail all the time--for example, Cablevision's VOOM was the first HD satellite TV service, and lasted for approximately two years. The VOOM set-top boxes could also be used as ATSC digital receivers, so a few of them are probably still in use, but most of them ended up in the scrap pile.

The lesson is that consumers should beware when being asked to purchase a set-top box, especially an expensive one. If the company offering the service goes out of business, changes strategic direction or is acquired, its set-top box is likely to become a paperweight. Apple and Roku have the right idea--none of their set-top boxes are priced over $100. It's clear that no one feels any particular responsibility to their customers if they discontinue their services, so caveat emptor.

Adobe pins Flash's future on 3D

Adobe Flash's clock is ticking, and not in a good way. Apple wants it dead, Microsoft is about to make life for it much less comfortable with Windows 8, and Google supports it, but only to get under Apple's skin. The better that browsers and authoring tools that support HTML5 get, the smaller the remaining market space for Flash becomes.

For Flash to remain viable, Adobe has to position it for applications that can't be done well, or at all, with HTML5. It appears that Adobe has decided to focus on 3D applications with Version 11 of Adobe's Flash Player and Version 3 of Adobe AIR, both of which are scheduled for release in early October. Flash 11 has a new GPU-accelerated 3D API called Stage 3D, which should dramatically improve 2D and 3D rendering rates on devices with compatible hardware. Adobe claims that, compared with the current Flash Player's capability to render thousands of non z-buffered triangles at 30 Hz, Flash 11 will be able to render hundreds of thousands of z-buffered triangles at 60 Hz.

Flash is currently very popular for browser-based casual games, but Adobe wants Flash 11 to be used for far more graphically-intensive games, and has signed up EA Interactive, Ubisoft and Zygna to support the new capabilities in their games. The problem, as Wired's Webmonkey site points out, is that WebGL can provide at least the same level of 3D performance in browsers without plug-ins, and WebGL is an industry standard. However, Internet Explorer currently doesn't support WebGL at all, and other browsers have widely varying WebGL performance.

So, Adobe may be able to carve out a niche with Flash for 3D applications, but it's hard to see it as anything more than temporary. Browser developers will be improving their WebGL performance in parallel with increasing their HTML5 compliance. The real test will be how many developers successfully market games written with Stage 3D.

Nikon unveils Nikon 1 EVIL cameras

Nikon finally entered the small-form-factor interchangeable lens camera (sometimes called EVIL) market with a new system, called Nikon 1. The first two members of the Nikon 1 family are the J1 and V1. Both models use a new 1" 10MP CX sensor developed by Nikon (hence the "Nikon 1" name). Both the size and the resolution of the sensor are substantially smaller than the sensors used by Panasonic and Olympus in their Micro Four-Thirds cameras, or by Sony in its NEX-series cameras. The smaller sensor gives the Nikon cameras a larger crop factor (2.7x) than its competition (for example, a 100mm lens on a Nikon 1 camera would be comparable to 270mm.)

The J1 is the entry-level model in the Nikon 1 line. It has an electronic shutter that limits sync speed to 1/60th of a second, but that operates at up to 1/16,000th of a second. It's got a 73-point autofocus system that switches between phase- and contrast-detection to get the best focus under current lighting conditions, and a 3", 460,000 pixel resolution LCD. Its burst rate with adaptive autofocus is 10 fps, and a blazing 60 fps when focus is locked. In video mode, it can record at 1080i60 and 1080p30. It's got lower-resolution slo-mo modes at 400 and 1,200 fps, as well as a somewhat confusing Motion Snapshot mode that records brief clips at 1080p60 but plays them back at 1080p24 for a slow-motion effect. The J1 will ship in the U.S. on October 20th with a 10-30mm lens for $649.95.

The V1 is the "enthusiast" Nikon 1 model. It's positioned by Nikon as the world's smallest camera with both an .47', 1.44MP resolution electronic viewfinder and interchangeable lenses. It has both an electronic and mechanical shutter; the electronic shutter is the same as in the J1, and the mechanical shutter supports sync speeds of up to 1/250th of a second and can operate up to 1/4000th of a second. The V1 also has connections for stereo microphones and a "multi-accessory" port for attaching a speedlight or GPS module. Other than these features, the V1 is functionally identical to the J1. The V1 will also ship in the U.S. on October 20th with a 10-30mm lens for $899.95.

Nikon also announced a range of lenses for the Nikon 1 cameras: The 10-30mm zoom included with the two new cameras, a 30-110mm f3.5/5.6 zoom lens, a 10-100mm f4.5/5.6 power zoom for movies, and a 10mm f2.8 pancake lens. Nikon will also offer the FT-1 F-mount adapter for using conventional Nikkor lenses with Nikon 1 cameras. Prices and availability have not yet been announced.

I have to admit that I'm underwhelmed by the Nikon 1 cameras and lenses, based at least on their specifications. The 1" CX sensor is both substantially smaller and lower-resolution than either the Micro Four-Thirds or APS-C sensors. Nikon compensates to a degree with the camera's hybrid autofocus system and very fast burst rate (up to 60 fps with focus locked), but I doubt that's going to be enough to save these cameras. Nikon says that it's pursuing consumers who want better cameras than point & shoot models, and who want interchangeable lenses without the complexity of DSLRs. The problem is that the new Nikon models don't appear to be significantly smaller, cheaper or easier to use than competitive cameras that take better pictures and are more flexible (although the new hybrid autofocus system is a significant advance). In addition, the big plus of the Nikon 1s, their fast burst rate, is of most use in applications such as sports photography, where image quality is a paramount consideration.

Nikon's new lenses are also fairly slow, especially the power zoom, which negates some of the cameras' depth-of-field advantages. On paper, at least, there's not much about the Nikons that would cause anyone other than a devoted Nikon customer to buy them, rather than buying a Sony, Panasonic or Olympus model.

Windows 8: Destination unknown

Earlier this week, Microsoft formally introduced Windows 8 to its developer community at its Build conference. The new operating system was well received--so much so that Microsoft CEO Steve Ballmer had to caution people who downloaded the developer preview version that it's very early software and should in no way be used for production environments. Windows 8 is, in a way, Microsoft's answer to both iOS and OSX from Apple: It will run on PCs, notebooks and netbooks, like Windows 7, but it will also run on tablets, like iOS. It'll run on Intel processors like Windows 7, but it will also run on ARM processors like iOS. It runs all the Win32 desktop-style applications that Windows 7 does, but it also has a touch-oriented user interface and supports touch-based applications like iOS.

Windows 8's "secret sauce" is a new touch-oriented user interface called Metro. Metro is based on the Windows Mobile 7 operating system's user interface for smartphones, which is in turn based on the now-defunct Zune user interface. It does the sliding, zooming and selecting kinds of things that you'd expect from iOS and Android, but it does it on top of Windows. That represents both an opportunity and a problem.

The opportunity is that one operating system, Windows 8, could service both the legacy PC market and the new tablet market. Most observers believe that Apple will eventually unify iOS and OSX; Microsoft has done it with Windows 8. It's much easier for developers to build and support applications for one operating system instead of two.

Windows 8 supports two different kinds of applications: Win32 and Metro-style. Win32 applications are the programs and user interface elements that we've seen on Windows since Windows 3.1. They've been refined and updated over the years, but a Windows XP user shouldn't have any difficulty using a Win32-style program on Windows 8. Win32 apps are designed for a keyboard and mouse, and will work poorly, or not at all, with a touch interface. Metro-style applications, on the other hand, are all full-screen, use a radically different user interface, and are optimized for touch. You can use a mouse and a keyboard with Metro-style apps, but you probably won't want to do so.

Windows 8 systems will open into the Metro user interface, using a touchscreen. If you want to run a Win32 app, you'll switch back to the old user interface, using a keyboard and mouse. You'll then go back and forth between user interfaces as you switch from program to program. You can configure Windows 8 to always use a Win32-style GUI, but then you'll lose access to the Metro-style apps. That's the problem with running both Win32 and Metro-style applications on the same device.

Apple knew that tablets were going to be used very differently from desktop computers, so it optimized iOS for touch on smartphones and tablets. iOS doesn't try to run OSX-style keyboard-and-mouse applications; it offers similar functionality, but the user interfaces are very different. Windows 8, on the other hand, wants to be equally good at desktop and tablet applications, even though the use cases are very different.

For existing Windows users who want to continue using Win32-style apps, from what Microsoft showed this week, there's not much value in upgrading to Windows 8,  other than cutting down on boot times. Customers who are primarily interested in Metro-style tablet apps will find that app developers are starting from scratch, and it may take years for Microsoft to reach parity with the selection in Apple's iOS App Store. Customers who want to use both types of apps, and switch back and forth, are likely to get frustrated very quickly with dealing with two dramatically different user interfaces.

My suspicion, sitting a year or more away from when Windows 8 formally launches, is that Metro-style apps will be used almost exclusively on tablets, and that desktops, notebooks and netbooks will almost exclusively be used for Win32-style apps, despite Microsoft's goal of getting touch-enabled displays on every kind of computing device. This approach will allow customers to avoid switching between user interfaces, but it won't do much for Microsoft's growth prospects. Windows 8 on conventional PCs and Windows 8 on tablets will represent two different markets, and on the tablet side, Microsoft won't be able to leverage its huge PC installed base. It'll be starting from zero. Given that Apple will probably be on the iPad 4 and Google will be on Android 5 by the time that Windows 8 ships, that's not a good place for Microsoft to be.

Atomos' little solution to a big problem

From the standpoint of getting high-quality video and audio out of a camcorder, either HDMI or HD-SDI will work, but HDMI is a consumer connection--the connector doesn't lock, and it's designed for casual use. SDI, on the other hand, is a professional connection--the connectors lock, and it's designed for much heavier usage than HDMI. There have been converters that input HDMI and output HD-SDI and vice versa, but they've tended to be big and line-powered--not something that you'd want to carry with you wherever you go.

At IBC, Atomos, the Australian video recorder folks, introduced a new product, Connect, that solves the problems with existing converters. The Connect is available in two versions: HDMI to HD-SDI, and HD-SDI to HDMI. Each one is priced at 249 Euros/$349 (U.S.), and is only 42.5mm W x 29 mm H by 72.5mm L. They're powered by a Sony battery and fit on Sony, Canon and Panasonic battery plates, so they can be mounted easily on almost any device. They can also be ganged together, so that multiple converters can share a single external battery.

Atomos' Connect converters can add a HD-SDI interface to any camcorder, make DSLRs work with professional video recorders and switchers, and enable lower-cost HDMI monitors to be used with HD-SDI devices. The company expects to begin shipping the devices worldwide by the end of 2011.

Sony: Getting its act together (at least with cameras and camcorders)

I've pointed out over the last several years the many times that Panasonic has eaten Sony's lunch in the camcorder and camera businesses:

• Panasonic's P2 and SD memory cards were the start of the solid-state, file-based camcorder revolution, while Sony's XDCAM optical disc has largely been a dead-end format.
• Panasonic and Olympus pioneered the big-sensor, small-camera interchangeable lens format with Micro Four-Thirds, while it took Sony several years to get on board with its NEX series of cameras.
• Panasonic and Canon introduced professional-level video to their DSLRs several years ago, while Sony's A77, just released a few weeks ago, is the company's first DSLR with professional-level video.
• It took Sony more than a year to respond to Panasonic's AF100 Micro Four-Thirds-based camcorder, with the F3 and FS100.

Finally, Sony is back in the game in a big way, in three different product lines:

• The new A77 DSLR is a first-rate still camera that can compete with Canon and Panasonic on video as well. It has a 24 megapixel sensor, a 2.4 megapixel OLED electronic viewfinder that's actually much higher resolution than its 920,000 pixel pull-out LCD display, and a 12 fps burst rate. It can record video at 1080p60 (1080p50 in Europe) using AVCHD 2.0 (Panasonic refers to it as AVCHD Progressive), which supports a maximum bit rate of 28mbps, versus 24mbps for previous AVCHD implementations. It also supports live autofocus in video mode. U.S. body-only pruce is $1,399.
• The NEX-7 is Sony's new top-of-the-line EVIL (electronic viewfinder, interchangeable lens) camera that competes with Micro Four-Thirds models. It has the same sensor and OLED electronic viewfinder as the A77, in a pocket-sized package. It also records video at 1080p60 using AVCHD 2.0, with the same live autofocus as the A77. Expected U.S. body-only price will be $1,199.
• The NEX-5n replaces the former top-of-the-line NEX-5 EVIL camera. It uses a 16.1 megapixel sensor and doesn't come with an electronic viewfinder; the optional FDA-EV1S viewfinder, which is connected to the top of the camera, has the same OLED display as the NEX-7 and A77. It records video at 1080p60/50 using AVCHD 2.0, and has live autofocus. U.S. body-only price is $699.
• The LA-EA2 adapter allows Alpha-mount lenses to be used with the NEX-series E mount. The adapter supports autofocus on the Alpha lenses. The LA-EA2 is almost as big as the NEX cameras themselves, and it costs $399, but it could be valuable for those NEX-family camera users who want to use Alpha lenses without losing autofocus capabilities.
• The F3 is Sony's lowest-cost digital cinematography camcorder. It uses a Super 35 Exmor CMOS sensor and PL mount for supporting film lenses. It records in 4:2:2 1080p59.94/50, and optionally in 4:4:4 1080p59.94/50, on SxS media. Its U.S. list price is $16,800 without lens.
• Sony's NEX-FS100 is Sony's answer to Panasonic's AF100, with a unique, squared-off form factor and top-mounted electronic viewfinder. It uses the same E-mount lenses as Sony's NEX-family of still cameras, and comes with an 18-200mm zoom lens. The FS100 uses the same Super 35 Exmor sensor as the F3. Like the new NEX cameras, it records video at 1080p60/50 using AVCHD 2.0. The FS100 records on SDHC/SDXC and Sony Memory stick cards, and the optional HXR-FM128 records onto 128GB of flash memory. U.S. list price with lens is $6,550.

For all of Sony's new competitiveness, the company still sometimes makes questionable decisions about product features. For example, the FS100 has dual XLR audio inputs, but only outputs video via HDMI rather than HD-SDI. HDMI connections, which don't lock, are notorious for becoming disconnected at the worst possible time. By comparison, the less-expensive Panasonic AF100 has both HDMI and HD-SDI outputs. Also, the AF100 has built-in neutral density filters, but the FS100 doesn't.

Also, Sony believes that, as a matter of principle, its products are worth more than the competition, even if they're objectively no better. Going back to the FS100-AF100 comparison, Sony's FS100 is about $1,000 more than Panasonic's AF100, even though the AF100 arguably has more professional features than the FS100.

Even given those caveats, however, Sony has clearly gotten back into the game with both still cameras and camcorders. It's good to see the company competing for business instead of resting on its laurels.

Speculation: The game any number can play

Late last week, TechCrunch's MG Siegler reported that he had not only gotten his eyes on a prototype of a new Amazon tablet, but his hands as well. He reported many details about the hardware and software, the price, when it will be announced, and even what kinds of promotions Amazon plans for it. Siegler's story triggered a flood of speculative articles, all primarily based on his description. For example:

• From Forbes:  Amazon Can Become Tablet King With E-Commerce Underwriting
• From The Wall Street Journal: Amazon.com Says It Is Testing Website Redesign (Changes, Which Will Roll Out Gradually, Were Made With Tablet Computers In Mind)
• From Wireless World: Amazon's $249 tablet could shake up pay TV business

I take Siegler at his word that he saw and used a prototype of an Amazon tablet, and that he spoke with a source with inside information about Amazon's plans. However, we have no independent verification that the tablet he used is representative of the final product, or that the pricing, availability and promotional details that he got are correct and won't change by the time the product finally ships. Nor do we have any independent verification of what he wrote about a second tablet, or about Amazon's plans for other black & white Kindles.

It's easy (and fun for the whole family) to base articles on a single, unconfirmed story, but they take as fact what is only rumor and hearsay. The problem with piling speculation on top of hearsay is obvious: If the hearsay is incorrect, the speculation based on it is even more incorrect, and the whole pile tumbles down like a badly-constucted Jenga tower.

We know what TechCrunch published. Let's get some independent confirmation of the facts before we start drawing conclusions or determining what the impact will be on the industry. If Siegler is correct, we don't have long to wait before we get confirmation from Amazon itself. There's plenty of time to determine the implications of Amazon's actions once we know the "true facts".