Articles by Steve Bain

You are currently browsing Steve Bain’s articles.

PCB featues to suppport a Wi-Fi radioThe Kindle DX includes an area on its main printed circuit board (PCB) that almost certainly implements electrical connections for a Wi-Fi radio. The area includes sites for an integrated circuit package or radio module, various discrete components, and a ceramic antenna. These PCB features are exposed in a photo from a recent tear down of the Kindle DX at the Rapid Repair website.

None of these Wi-Fi related PCB positions are populated with components in the currently shipping Kindle DX, but if they were populated, the Wi-Fi antenna would reside near the upper left-hand corner of a front-facing Kindle DX.

The PCB features suggest that Amazon is in a technical and manufacturing position to introduce a Wi-Fi radio into its Kindle line of e-readers at any time. A decision to introduce a Kindle with Wi-Fi may depend on the results of Amazon’s negotiations with wireless carriers around the world. Wi-Fi would be a reasonable option for wireless connectivity in any locale where cellular service is unavailable or too expensive. Given the intense global competitive environment in e-readers and e-publishing, and the rapid pace that Amazon has been keeping in Kindle developments, announcements relating to international availability for Kindle products should be expected soon.

For more discussion about the potential for use of Wi-Fi radios in future Kindle products, see this previous post at Kindle Zen. For photos from the surgery on a Kindle DX, see this web page at Rapid Repair.

Update 10 July 2009: We have tentatively identified the RF module that mates to these unpopulated PCB features in the Kindle DX. It is the Murata SyChip model LBWA18HEPZ and it implements a complete IEEE 802.11b/g Wi-Fi radio in a tightly integrated package. This power efficient Wi-Fi module incorporates the well regarded Marvell 88w8686 Wi-Fi chip and the module is manufactured by Murata in large volumes. The Marvell chip also appears in the Wi-Fi radio for the Palm Pre and the chip is supported by Linux device drivers.

Wi-Fi module in Embest DevKit270 This Murata module seems to be quite an appropriate choice for a Kindle Wi-Fi radio. The module is mature, cost-effective, and appropriately low power. This choice can be viewed as another in a series of astute, conservative design decisions by the Kindle DX hardware team. To see the PCB implementation of a Wi-Fi radio using this Murata module in a different application, check out  the DevKit270 from Embest.

Of course, these PCB features in Kindle DX do not necessarily portend a business decision by Amazon to bring Wi-Fi to Kindle. The PCB features could just be a left-over from Wi-Fi prototyping or included simply for use as a lever in negotiations with wireless carriers. It would not be entirely surprising if a different motherboard and different radio solutions appear in future DX-style Kindles for international markets.

I’ve been experimenting with the Kindle DX for several days and my overall impression is very positive, but I believe that the support for PDF documents is an embarrassment for Amazon. In this post, I will offer a critique for the PDF user interface in the Kindle DX.

First some background. PDF is an open but Adobe-specified digital document format that seeks to replicate the geometrical richness of a printed page. It is a complicated format that has evolved in nine succeeding releases over fifteen years, and full implementations of the specification are difficult to develop. There are apparently several classes of PDF rendering engines that are licensable from Adobe for embedded applications and some good third party options are also available.

PDF rendering in the Kindle DX is generally impressive. The PDF engine for the DX is apparently based on a mobile device offering from Adobe. This was a conservative choice by Amazon and was likely the least expensive option for PDF rendering that could be licensed from Adobe. However, this PDF implementation is actually a poor value for Kindle DX customers because it does not support linkage. Linkage from PDFs to external web documents might be of questionable value on the Kindle because of the clumsy web browser. However, linkage within PDF documents is an absolutely critical feature for efficient navigation, and it is hard to understand why Amazon would deliver PDF support without this key feature. Linkage enables an active Table of Contents that can greatly ease browsing in a long technical document. Linkage also enables practical access to cross-references within documents such as to footnotes, figures, and tables.

Since there is no support for linkage, how does one navigate? Well, the Kindle DX supports these methods for navigation in a PDF document:

  1. Next Page and Previous Page buttons
  2. Bookmarks
  3. Search
  4. Go to Beginning (of document)
  5. Go to Page

Bookmarks are not an efficient means for navigation, because they must first be set by the user. Further, the Kindle DX provides no way to name PDF bookmarks nor does it provide any context for PDF bookmarks. Rather, the user interface is simply a selectable list of the page numbers that have been bookmarked.

Search can be a useful navigation method, and especially when one is recently and deeply familiar with a document. However, search is no substitute for direct navigation because words are seldom unique within a large document.

This leaves "Go to Page" as the most broadly useful of the navigation methods that Amazon has provided for PDFs.  From a document’s table of contents, you presumably estimate the page number for the start of a section of interest and then you do the following:

  • Press the Menu button
  • Manipulate the 5-way controller to highlight "Go to Page…" in the pop-up menu
  • Select this function by pressing the 5-way controller
  • Enter a page number using chorded keystrokes, Alt-1, Alt-5, Alt-3, etc.
  • Press enter to complete the navigation

This process could be simplified with a better UI design. First, "Go to Page…" could be the default option on the pop-up menu whenever a PDF document is displayed. Then, when a page number is being entered from the keyboard for navigation, the Kindle DX could interpret presses on the top row of keys as numbers so that there is no need for chorded entry on the cramped keypad. This would offer no disadvantage, because the page number entry field does not accept letters.

Clumsy navigation is not the only problem for PDF documents in the DX. The Kindle DX is missing a high-contrast option that would ease reading for PDF documents that were designed for color rendering. With such color PDF documents on the DX, text is displayed in shades of gray that can be difficult to read. Also missing from Amazon’s PDF implementation are support for highlighting, annotation, the dictionary, and text-to-speech. Why should any customer be satisfied that such features are missing for PDF documents?

There are literally millions of existing PDF documents that Kindle users might want to access, but a vast majority are formatted for letter-sized or A4-sized pages, and virtually all are formatted for page sizes larger than the Kindle DX display. This means that PDF pages are usually rendered in portrait mode on the DX at a size much smaller than the document developers originally intended. Even though the Kindle’s resolution is excellent, reading documents with tiny letters can become frustrating in a hurry. The solution on Kindle DX  is to rotate the device into landscape mode, which usually results in roughly a 1.5X zoom and half as much page area displayed. This might be a reasonable compromise except for one major problem: It places the critical Next Page and Previous Page buttons at either the top or bottom of the display where they are inconvenient to access. I find it hard to understand why the DX doesn’t repurpose keyboard keys as alternate paging keys when the Kindle is in landscape mode.

When the DX is in landscape mode, the keyboard space bar could be enabled as an alternate Next Page button. This simple change would result no important collision with the text entry features, and would greatly enhance the usability of PDF documents in landscape mode for both right and left handed users. Another button on the bottom row of the keyboard, perhaps the Shift key, could be enabled as an alternate Previous Page button. This might result in a minor collision with existing uses, but one that is far less important than having convenient, hand-in-place access to the Previous Page function.

It is a major embarrassment for Amazon that the Kindle DX was released to customers with such a poor user interface for PDF documents. The current PDF document support should be viewed as an "experimental feature" at best, and the deficit leaves a hole in Amazon’s product line that could be a nice opportunity for competitors like Sony and Plastic Logic. The hope for Kindle DX customers is that Amazon will rapidly improve PDF usability through a major firmware update. Some improvements, like better menu organization and sane use of the keyboard, are pretty easy stuff. Others, like improved contrast options for color documents and support for intra-document linkage, might require a license for different software from Adobe, or perhaps the switch to a third party PDF rendering engine.

It has been rumored that release of the Kindle 2 was delayed past the Christmas season because Jeff Bezos was dissatisfied with early releases for the user interface firmware. One wonders who within Amazon approved of the PDF document features for the Kindle DX.

Amazon-optimized-for-DX-badge Karen at the excellent Books on the Knob blog notes in a post today that some books in Amazon’s Kindle Store are now marked as "Optimized for Kindle DX":

Will those who use a KindleDX get the same book and just have more screen real-estate for viewing pictures and charts… Or will those with a KindleDX get an entirely different book, with pictures, tables and other layout features omitted from the basic reader device versions?

That is a very good question. We presume that the answer will be clear by late next week as customers begin experimenting with the DX.

Amazon’s offers a short explanation for the "Optimized for Kindle DX" logo in an adjacent pop-up window:

This title has complex layouts and has been optimized for reading on Kindle DX’s larger screen, but can still be viewed on other devices.

Karen’s post also suggests a question. If an Amazon customer has previously purchased a Kindle book that is now available in updated format as "Optimized for Kindle DX", will the customer have to buy the book again to read the DX-optimized format on a brand new Kindle DX?

kindle_netbook_480

Pixel Qi is a display company that was started about a year ago by Mary Lou Jepson, the former chief technology officer of the One Laptop Per Child project. The company is focused on adapting liquid crystal display technologies into new configurations and applications. The first product is a ten inch dual-mode display that operates in a full color mode with backlight, or in a passive monochrome mode with full readability in sunlight. Some recent press has framed Pixel Qi’s display technology as a possible threat to Amazon’s Kindle business, but the technology is presumably just as available to Amazon as to any other manufacturer or vendor. 

Pixel Qi has recently been working around the clock to produce full-functioning prototypes of this new style of multi-mode liquid crystal display in time for demonstrations this week at Society for Information Display conference in Austin and the Computex trade show now opening in Taipei. First photographs of prototype displays appeared late last week.

At this time, we have little information about the full-color mode of the Pixel Qi display, but color performance will be a key issue for many potential applications. Of particular interest will be parameters like spurious reflectivity, color saturation, and backlight power requirements for color operation.

This post will instead focus on the display’s monochrome reflective mode that Pixel Qi is promoting for use in e-readers. We will examine photographs that have been posted on Ms. Jepson’s blog and on Time.com’s NerdWorld blog. These photos are of the Pixel Qi display prototypes operating in monochrome mode without backlight.

The first is a high resolution photo posted on Ms. Jepson’s blog that shows the Pixel Qi display side-by-side with an Amazon Kindle 1. At a first glance, the photo seems to illustrate higher contrast and substantially clearer text on the Pixel Qi display than on the Kindle. However, some major footnotes are less apparent:

  • The camera’s line-of-sight is not perpendicular to the Pixel Qi display (in a vertical plane) even though the display is hinged and could be easily adjusted to perpendicular. In fact, the display’s front surface is set to specularly reflect ambient light from above the camera, which for all we know could be a black screen or a non-illuminated area of the room. The tilt of the Kindle is a bit less clear.
  • Judging by specular reflections from wrinkles in a plastic liner on the bottom bezel, the Pixel Qi display appears to be illuminated from the left side of the camera. This could  be a bit disadvantageous for the comparative background tint of the Kindle’s display which is positioned to the camera’s right, further away from the light source.
  • The camera is sharply focused on a right side section of the Pixel Qi display (which is in the center of the field-of-view) but the Kindle’s display is positioned well to the side and is not well focused.
  • The Pixel Qi display has a black bezel, whereas the Kindle’s is white. This provides an optical illusion that makes it difficult to eyeball the comparative brightness of backgrounds for the two displays. In specific, this illusion makes the Kindle’s display background appear darker than the Pixel Qi’s. Actual pixel gray values, however indicate that the Kindle’s background is on average slightly brighter.
  • The specific Kindle in the photo might not be generally representative of the performance of current Kindle displays. For example, in testing at Origin Instruments, two later model Kindle 2 units exhibit a noticeably whiter background than a Kindle 2 that was purchased at that product’s introduction.

Thus, this photograph does not indicate that the Pixel Qi display will exhibit superior contrast to current Kindle displays. However, the apparent performance of the Pixel Qi prototype is very impressive. It seems to exhibit excellent monochrome contrast when viewed from a roughly perpendicular perspective.

Another photo, however, is a bit more problematic for the Pixel Qi technology. This photo is provided at lower resolution, but it shows the Pixel Qi display at an oblique angle to the camera and, in the background, a Kindle 1 at an even more oblique angle. The structure of ambient illumination is not apparent, but the photo allows us to make two rough judgments:

  • The Kindle’s display offers substantially higher contrast than the Pixel Qi display at oblique angles.
  • Contrast for the Pixel Qi display varies strongly with angle at oblique angles, as can be seen from the obvious right-to-left variation over the display surface in the photo. Yet for the Kindle, display contrast is largely unaffected by viewing angle.

We conclude that the Pixel Qi display could be an especially interesting option for use in netbooks and notebooks, where a friction-hinged display can be adjusted and the keyboard base can be rotated for the best viewing angle in any given ambient light environment. It is not surprising that netbook displays are the first market that Pixel Qi is addressing.

However, the display’s monochrome mode could be a bit less attractive in handheld applications like media tablets and e-readers, because these devices are sometimes used at a wide variety of viewing angles. A very typical use case for e-readers involves an oblique viewing angle, with the display lying flat on a table or desk for reading during a meal or while in use as a reference for activities that are progressing on another screen.

For applications in e-readers, key disadvantages of the Pixel Qi display as compared to the Kindle’s current E-Ink display seem to be:

  • Somewhat darker background and lower contrast at best viewing angles
  • Significantly less display contrast at oblique viewing angles
  • Order-of-magnitude higher power consumption, even in the lowest power monochrome mode

The Pixel Qi display’s key advantages seem to be:

  • Higher pixel resolution for a given panel size
  • Much faster display response that could enable greatly improved UI’s for organizing, selecting, and navigating e-books
  • An integrated full color mode, albeit one that operates at reduced resolution and likely much increased power consumption

From what we know today, the Pixel Qi display appears to be a possible but not a clear win as compared to the current E-Ink displays for use in devices like the Kindle that are optimized for long-form reading.

We look forward to definitive information about power consumption for the Pixel Qi displays. However, we should note that Ms. Jepson’s enthusiasm has sometimes shaded the truth about pesky issues like power consumption, so it may be prudent to discount performance estimates from Pixel Qi until independent test results are available.

A useful 13 minute video demo of the Pixel Qi prototype is now available on YouTube:

Abhi has an excellent post up today, Kindle Store Needs Discoverability, that addresses some of the problems associated with browsing Amazon’s Kindle web store in search of books and blogs of specific interest:

Basically, Kindle Store is only useful if you already know what you want – its really difficult to find new content that you would like.

Discoverability is indeed a major weakness in the current Kindle store, and one that could be ripe for targeting by Amazon’s competitors. Read the whole thing.

Amazon has opened the Kindle blog delivery service to all blog publishers. If you want to submit your blog to the service, be sure to read and understand Amazon’s "Digital Publication and Distribution Agreement – Blogs and Periodicals." The agreement is presented at a late step in the sign-up process. This agreement grants Amazon quite a lot of control over the distribution of your content, including the authorization to set any pricing that it chooses, to offer your content in any venue, and to aggregate your content at will. The agreement specifically prohibits the use of advertising in your publication. The publisher’s cut is 30% of Amazon’s revenue, and the proceeds are prorated among multiple publishers if the content is aggregated. The publisher can, however, exit the agreement with a 30 day notice to Amazon.

This service appears to be exclusively intended for the publishing of content that exists in typical blog format. However, the Kindle is more suited to the display of narrative-style blogs than to link-centric blogs because the Kindle’s web browser is basic and often problematic.

It appears that Amazon’s terms and service settings would allow a publisher to experiment with the development of a unique-to-Kindle periodical that is delivered as the superset feed of a blog. Such a publication could offer content that is only available in full through a Kindle subscription. However, a big footnote is that Amazon will set the price for the subscription, so this avenue is not yet suitable as an outlet for existing specialty periodicals like newsletters with paid subscribers.

It is not yet clear what controls are in place to prevent "publishers" from gaming the system with hard-to-police pirated content, prevent "publications" that merely aggregate the content of others without permission, or prevent "blogs" that package book-length content in each post. Such potential problems may be among the reasons why Amazon refers to the service as a beta release.

For most blogs, this new Kindle publishing service should be viewed as a way to modestly improve exposure while perhaps covering the cost of an occasional cup of coffee at McDonalds. One publishing strategy that small blogs should consider is to aggregate content with other publishers using a tool like Sam Ruby’s Venus so as to create new publications that are jointly owned and promoted, and that can offer comprehensive coverage of specialty topics at great value for Kindle readers.

As the Kindle publishing service evolves, publishers should encourage Amazon to offer less onerous terms, increase publishers’ share of the revenue, and allow more flexibility in formats and pricing. We suspect that a more Apple-store-like revenue model (perhaps 70% to publisher, 30% to Amazon, with usage-based data delivery costs explicitly funded from the publisher’s share) would invigorate the market and thereby encourage the creation of premium content. That would be good for Amazon, for publishers, and for readers. The service as currently implemented seems unlikely to produce much additional value for anyone.

If you nonetheless want to publish your blog for delivery from the Kindle store, visit Amazon’s Kindle Publishing for Blogs Beta.

The Vulture at the New York magazine website has a humorous post up today that lampoons the New York Times for some of its recent Kindle coverage. The spy photo of a Kindle-waving internet pirate is not to be missed.

StudentWithBooks In much recent press coverage, Amazon’s introduction of Kindle DX is seen as marking the beginning of a transition to e-textbooks. That may well be true, but with special emphasis on “beginning”. Regardless of how impressive the DX turns out to be, we expect that publisher caution and market inertia will combine with the DX’s near-$500 price to limit Amazon’s near-term business opportunity in e-textbooks.

US textbook publishing is dominated by a small number of very profitable businesses, and one may expect them to be cautious about initiatives that could impact a lucrative business model. From a publisher’s perspective, a Kindle channel for e-textbooks will reduce costs for printing, shipping, and returns, and perhaps more importantly will support no market for used textbooks and thus could increase unit sales while reducing costs associated with the current practice of rapidly updating textbook editions so as to discount the relevance of used previous editions. However, Amazon’s position as the exclusive sales channel of DRM’ed content for Kindle is problematic for publishers and they may be understandably reluctant to empower Amazon as a critical link in their sales channel.

A recent article in the Wall Street Journal about the announcement of the Kindle DX included this passage about Amazon’s e-textbook initiative:

Amazon also announced partnerships with textbook publishers including Cengage Learning, Pearson PLC and John Wiley & Sons Inc., to sell their textbooks on the Kindle DX.

A spokeswoman for Pearson said its education division, which publishes textbooks for all grade levels, hasn’t yet decided how many titles will be available or how much the Kindle versions will cost.

The emphasis is ours. Largely skeptical views of the near-term prospects for e-textbooks have also recently appeared at Inside Higher Ed and in a posting at ZDNet’s education blog.

Given the surging competition, we expect that the Kindle DX will be followed by the next generation of large-format Kindle in 2010. During its market life, the DX may see more educational use as a reader for scanned handouts, open textbooks and informally pirated textbooks than for published e-textbooks. A new Apple media pad of Kindle-like size may appear very soon and it will likely be popular with college students. Delivery of e-textbooks for use on such non-Kindle devices may emerge as an important element of Amazon’s longer-term strategy.

Our skepticism about near-term volumes for e-textbooks is not intended to discount the brilliance of Amazon’s business plan nor the attractiveness of the Kindle DX. In fact the DX appears to be a well engineered product and we expect that it will open a variety of new opportunities for Amazon and for you, our readers.

Kindle and Wi-Fi

WhereIsMyWiFi We are surprised that specifications for the Kindle DX include no mention of a Wi-Fi radio. We were not anticipating that a Wi-Fi radio would replace the cellular radio. Rather we were expecting that the two radios might coexist side-by-side.

A recent article at SFGate headlines: Why not a cheaper Wi-Fi-only Kindle? The article’s theme is that the Kindle’s price could likely be reduced by by eliminating the cellular radio and its associated service cost. The answer to the headline’s question is likely that Amazon is seeking a user experience that could not currently be provided by a Wi-Fi radio. Recent informal demographic estimates, developed at the Kindle Culture blog and elsewhere, suggest that the existing Kindle user base is largely older and majority female. This sounds to us like a demographic that may be especially appreciative of configuration-free operation. The Kindle’s existing wireless is fully operational out of the box with no configuration. We geeky types sometimes forget that subscribing to internet service, configuring a Wi-Fi router, enabling WPA security, and entering passwords are hardly inexpensive and intuitive tasks.

For current Kindles, Amazon’s Whispernet service is provided through the Sprint US cellular network. Sprint’s geographic coverage, both CDMA and EVDO, is actually quite impressive. However, Sprint’s coverage is far from universal and is especially spotty in parts of the US northwest. In addition, Sprint’s network has no international coverage. Thus a Wi-Fi radio could be of great benefit to Kindle users in many rural areas and to users who travel internationally. Such a Wi-Fi radio could be disabled by default to retain the existing Kindle’s user experience and battery life.

The marginal cost of the needed electronic components for a low power Wi-Fi (IEEE 802.11-g) radio manufactured at the Kindle’s volumes is currently very roughly at around $5.00 per unit, meaning that such a capability would likely add less than $25 to a Kindle’s sales price. Based on teardowns of the Kindle 2, we expect that Amazon has the technical ability to add Wi-Fi at any time to updated versions of either the Kindle 2 or Kindle DX, possibly by changing an integrated wireless daughter card and an internal antenna at the time of manufacture.

We are guessing – and hoping – that the next major Kindle news from Amazon will involve international availability. This will almost certainly require new Kindle models configured with TDMA/GPRS cellular radios, and it is possible that some international markets could be best addressed with Wi-Fi radios exclusively.

The addition of Wi-Fi radios to Kindles for the US market is purely speculative on our part, but Amazon may soon be under competitive pressure to offer such a feature. Thus far, Amazon seems to be doing a good job of positioning and executing to stay a step or two ahead of its e-reader competition.

Gigaohm has a nice post up today about Amazon’s focus on operating efficiency, including this Jeff Bezos quote in a footnote to Amazons’ 2008 letter to shareholders:

At a fulfillment center recently, one of our Kaizen experts asked me, “I’m in favor of a clean fulfillment center, but why are you cleaning? Why don’t you eliminate the source of dirt?” I felt like the Karate Kid.

Read the whole thing. Amazon is one of the more interesting businesses on the planet.

DX and K2 So you want a Kindle, but now you have the new problem of deciding which one.

The most important thing to understand is that the Kindle 2 and Kindle DX are very similar products that were undoubtedly developed in concert. They share almost all of the same advantages and disadvantages.

Kindle 2 weighs just about a pound including its leather cover. It is more convenient to carry than almost any book in your library, and it fits easily in a purse. Operation in a single hand is possible and even pleasant. Kindle 2 does not natively display PDF files, but Amazon’s free conversion service performs well in our experience for most narrative documents in PDF format, and the service couldn’t be more convenient to use (assuming you are willing to pay pocket change for wireless delivery).

However, the Kindle 2 and its available data formats are not suited to display of documents that require the geometrical richness of a printed page or that need to reference the original page numbers in printed documents. If research papers or technical manuals are an important part of your needs for Kindle, you would be well advised to wait for the DX.

The key distinctive features of the DX are its much larger display and its associated support for a much wider range of PDF documents, including bit-mapped PDFs. If you need to capture paper documents for e-reading, an inexpensive scanner will likely suffice without any requirement for optical character recognition and editing. Thus the DX could be especially useful in a range of applications that depend on scanning paper documents. It is in precisely such applications for complex and bit-mapped documents that the DX’s larger memory may be handy.

Kindle DX has almost twice the weight and surface area of the Kindle 2, so it will be a little less convenient to carry. In fact, the DX is large enough and heavy enough to be considered an exclusively two-handed device.

When both are packaged with the near-essential cover, Kindle DX is priced $150.00 higher than the Kindle 2. To paraphrase one of our team members, $150.00 is a lot of fajitas.

Our take: The Kindle 2 is for pleasure and the Kindle DX is for research, education, and business, but with lots of crossover for both devices at the boundaries. The Kindle DX earns extra credit for users with low vision, since the larger display is sure to be convenient for reading books at large font sizes.

In our experience, the Kindle is one of those rare products that grows on you as you use it. All of the minor issues that may be worrisome as you first experiment with the device rapidly fade into the background of reading. Then many features creep up on you as being much more useful than you originally anticipated. These include the paper-like display, the week-long battery life that largely eliminates any worry about power,  books that never loose your place, an always-available queue of reading material to suit your mood, and the convenience of simply emailing books and other documents for automatic conversion and wireless delivery to the device.

We are very pleased with the Kindle 2, and we look forward to testing and development for the Kindle DX. If you enjoy reading, we expect that you will be pleased with either product.

Hello, World

Welcome to the Kindle Zen blog. We will provide linkage and commentary for Kindle, e-reader, and e-publishing news from around the web, and offer a stream of original research with a special focus on Kindle technology and accessibility.

We begin this new internet publication about Amazon’s Kindle on the very day of the Kindle DX product introduction. However, our timing is the result of serendipity rather than planning.  We hope you’ll turn to us here at Kindle Zen, our Atom feed, or our twitter stream @KindleToday for the latest news about the Kindle and its universe of applications.