Tuesday, October 30, 2007

Photo permanence and durability

A "house altar" depicting Akhenaten, Nefertiti and three of their DaughtersExperience your memory fading away.

Historically, we humans have been willing to spend any amount of money on communication. We have been willing to spend even more money on communicating with our descendants after our passing — sky is the limit.

In the past you had to be a head of state to afford having your life cast in stone for posterity, or you had to be so good and work so hard that you would leave behind a historical legacy through the history books.

Today, everyone in our society can afford to leave behind their legacy in the form of digital items, such as scanned or digital photos, HD videos, AAC files, etc. Although such an archive is compact and convenient, it is still subject to bit rot. Actually, with the acceleration of technological progress, the span of time until bits have rotted is getting shorter.

With this in mind, there is still a strong argument for printing your memories and keep them around as atoms instead of bits. There is also a remarkable convenience to hard copies, especially when they are in the form of photo books.

Atoms can also rot, and in particular the dyes and pigments in inks fade. Therefore, it is useful to know how long your prints last. More specifically, you want to know how long prints on a specific media printed with a specific ink last.

In a concrete example, imagine you took a photograph during your honeymoon trip and you would like to be able to still enjoy at at you golden anniversary. Can you use refilled cartridges on generic paper or should you shell out money for fancy HP Vivera inks and special photo paper?

My colleague Ingeborg Tastl wanted to find a good answer just to this kind of problem — how will the photo with my memory fade away during the years? Since Ingeborg is our ICC profiling specialist, she built an interactive tool that allows you to simulate the fading of your memory when printed on two different ink and paper combinations. This is how the tool looks in on my PC:

Yoko in Luzern, 1. August 1991. Häppy 700th börsdey Switzerland!

Ingeborg being a very nice person, she lets you have her tool. You can download it from here and look at your own memory fading away. Of course, the lawyers had to have their say to keep us out of Court, so the selections are somewhat restricted, but it is still an interesting tool. For now the simulator is Windows only, but you can run in on an a virtual PC in an Unix system like a Mac.

Monday, October 29, 2007

An On-Line Color Thesaurus

Color names are a powerful means of selecting and communicating colors. There are a variety of color vocabularies and dictionaries available but there has been less work in capturing the similiarities and differences in color naming. This post is a tool post in that the online color thesarus is embedded directly in the post.

Getting Started
To use the online color thesaurus, click on the screen-shot below to get to the thesaurus page. There simply type the color name in the text field. Once you have typed in your color name click on "submit" and then the results will be displayed. You can use the "clear" button to clear the text field.

The results that are returned are a large color square with a rendering of the color if it was found. If the name was not found, for example "greeb" was entered, then the next nearest color name in terms of edit distance, in this case "green", will be returned. So no you won't have to remember how to spell fuchsia.

In addition to the colored square are the corresponding RGB and hexadecimal values. Finally there is a note about how common the color name is. Below this are the color synonyms and antonyms. Each column has smaller color squares rendering the color names and the names with links so that you can easily click-through to these names. The results are based on analysis of 20,000+ color name database in English collected from a 20+ language ongoing online color naming experiment.

Wednesday, October 24, 2007

MPEG-21 blog

If you are interested in Multimedia Communication for Universal Media Access (UMA) and MPEG-21, you may want to keep an eye on Christian Timmerer's blog on Multimedia Communication. Christian has been involved in MPEG-21 from the beginning and is a very active expert in the committee.

If you do not know anything about MPEG-21, you may be interested in the notes of the course I gave at VCIP 2003. They are a bit dated, but after you go though them you can quickly get up to speed with the current status on Christian's blog.

See also my previous post on MPEG-A.

Tuesday, October 23, 2007

Software patents

Earlier today RocketRoo left a comment on my earlier post on A color scientist's role, but it really is a new thread because my post had nothing to do with patents, so I am answering with a new post. Here is the comment:

Regarding the role of patents, one of the 2007 Nobel laureates in economics, Eric Maskin (http://nobelprize.org/nobel_prizes/economics/laureates/2007/), also did research at MIT on the value of software patents. He determined that software was a market where innovations tended to be sequential, in that they were built closely on the work of predecessors, and innovators could take many different paths to the same goal. In such markets, he concluded, patents might serve as an innovation inhibitor rather than an innovation incubator. See http://www.researchoninnovation.org/patent.pdf

Personally, I have mixed feelings about patents. The original idea was for society to reward inventors for their contribution by allowing them the right to exclusively reap the commercial benefits of their invention. However, now the system is broken.

If there is a date to be set for when system broke, it is probably the day the patent for the intermittent wind-shield wipers was enforced. But in reality, the system got broken by the submarine patents.

Some submarine patents came into being innocently. As I wrote in the post on A scientist's role, discovery is in the air and the skill is in being the first to grasp them. An inventor's antenna can pick it up and he or she can intuitively reduce it practice and file for a patent before the rest of the pack does. But when this happens too far ahead of the bleeding edge, the discovery is not yet well defined in the ether and intuition plays a larger role. Because of this, it is very difficult for a disconnected outsider to appreciate the invention, especially when we no longer show up at the Patent Office with out physical prototype..

Patent examiners are in such a difficult position and it can take a lot of back and forth until the examiner is satisfied that all that implicit knowledge has been made explicit, and can grant the patent.

But then there are the slacker or parasites, known more scientifically as defectors, who when they detect a discovery in the air submit a vague patent application based on a hunch, without understanding the issue nor attempting to render it to practice. An outsider, such as an examiner, has no way to tell a defector from a cooperator, so they have to give the benefit of the doubt while at the same time pursuing due diligence.

In Japan, applications are laid open after six months and anybody can comment on the applications. The system is more fair, but comes at a high cost for the engineers who are forced to work through reams of applications every day for a few hours.

Here in the US a similar system is being studied and HP is one of the companies behind this effort. If you are very experienced, I may suggest you join this collaborative effort as a reviewer. Just go to the Peer-to-Patent site, enroll, and review those patent applications that are in your field of expertise.

So far for the ethical issues. Your comment was about software patents. One problem is that software patent applications have been allowed only since about 1989 (AT&T Bell Labs traveling salesman patent). By that time computer technology was as advanced in several research labs, as it is now in the commercial world. However, in these 30 years computer development has changed so much with the use of wizards and frameworks, that today's programmers have no knowledge of what was standard practice 30 years ago. Hence, the wheel keeps being reinvented.

So, shall we get rid of software patents? It depends. We had this discussion about five years ago in one of the Swiss National Science Foundation Review Panels. We came to the conclusion that an invention should be patented only if it is necessary to protect a new business venture. When this protection is not necessary, then the funding should be invested in new engineering, not in patenting, because they cost more or less the same and in the long term engineering is better for society.

This also leads to agile companies that must innovate faster than the competition can catch up with copying. The Swiss have recognized this as a competitive advantage of their industry.

Monday, October 22, 2007

Color space dimensionality

Today RocketRoo posted a comment to my short August post on a paper on Multiscale contrast enhancement. Since that is a few months ago, I will reply with a new post. Here is the comment:

Re: Multiscale contrast: achromatic dims

In this recent study, http://compbiol.plosjournals.org/perlserv/?request=get-document&doi=10.1371/journal.pcbi.0030179, Vladusich, Lucassen and Cornelissen provide evidence that brightness and darkness form the dimensions of a two-dimensional achromatic color space. This color space may play a role in the representation of object surfaces viewed against natural backgrounds, which simultaneously induce both brightness and darkness signals. The 2-D model generalizes to the chromatic dimensions of color perception, indicating that redness and greenness (blueness and yellowness) also form perceptual dimensions. Collectively, these findings suggest that human color space is composed of six dimensions, rather than the conventional three.

Posted by RocketRoo on 10/22/2007 11:45 AM

Let me first admit that I just read the first few paragraphs and not the complete article cited.

I am in violent disagreement with the authors. The opponent color model was first proposed by Leonardo da Vinci, (see chapters CLX and CLXII of his Trattato della Pittura, Langlois, Paris, 2nd edition, 1701) then discussed by Wolfgang Goethe (see also here) in his virtual diatribe with Isaac Newton. The first modern theory of color opponency was proposed in 1872 by Ewald Hering and was very hotly debated until G.E. Müller and Erwin Schrödinger reconciled Helmholtz's and Hering's theories in the zone theory of color vision, which they based on the 1904 law of coefficients proposed by Johannes A. von Kries.

The matter was finally settled in 1956 when Gunnar Svaetichin was able to record from horizontal fish retinas and show opponent response in red-green and yellow-blue potentials. At the time, he showed that each horizontal cell is presumed to inhibit either its bipolar cells or the receptors, with further processing occurring in the amacrine cells and in the retinal ganglion cells.

You an always postulate a mathematical model approximating some phenomenon, but at the end what counts is physics and you cannot contradict the physiology on which color vision is based.

What may be confusing the authors is that much of color science is for aperture color. If the appearance mode is different, then part of ordinary colorimetry falls apart. This is why there are color appearance models and the authors should have based their work on these.

Thursday, October 18, 2007

Hue Angles blog

The ISCC (Inter-Society Color Council) now has a blog, which allows color scientists to interact on articles published in the Hue Angles column of the ISCC Newlsletter. Here is how its editor Dr. Michael H. Brill of Dataclor descibes this new blog, whose name is Hue Angles:

In fall 2006, Hue Angles began as a column for the ISCC News, devoted to tidbits of interesting lore shared by ISCC members in short-essay form. In its first year, the topics spanned color in wetland preservation, spinning disks under colored lights, personal recollections of selling color-matching systems, green in the fashion industry, how to measure color using a beer cooler, and color contextual effects. Almost any color-related topic is fair game. As of fall 2007, Hue Angles is also being posted here to facilitate lively discussion. As always, you can submit ideas or contributions for the column itself to Michael H. Brill, mbrill@datacolor.com.

Prof. Osvaldo da Pos, University of Padua, Italy contributed the first post, which is on colors and contectual effects.

Wednesday, October 17, 2007

A color scientist's role

Here are some thoughts about a color scientist's role in society.

When people ask me what I do, I answer "color scientist," with scientist being the subject and color being the object. One reason is that when I just say "color" people ask me fashion or design questions, which I cannot answer because I am not a color consultant. My field is called "color science" and that makes me a "color scientist." But is it not arrogant calling oneself a scientist? After all, I am not wearing a lab coat…

"Scientist" is not a bragging word. It is a qualification that brings with it also social responsibilities. Bertolt Brecht has collected a lot of material on this subject collected in Werner Hecht's Materialien zu Brechts »Leben des Galilei«, so I'll just mention a short conversation I had last night.

Yesterday evening I attended as a guest the Computer History Museum Fellows Awards Dinner and Ceremony. I was cruising the room in which the cocktails were hosted to greet old buddies, when in one group somebody noted how many former or current PARC scientists were in the room, commenting on the huge impact they had in the valley.

Nobody attends the event for the food. There are many restaurants where you get incomparably better food for $500 per person. People attend such event for the air — or better, for what is in the air. Maybe "ether" is a more appropriate word than "air."

When I was working at Canon, I had the problem that my boss kept telling me that at my level I was not allowed to do technical work, that my role was to inspire people. This was a problem for me, because I am not an evangelist, quite the opposite. In fact, at my previous job at PARC, where we tended to work in a team of a talker and a doer, I was much more of a doer than a talker.

With the benefit of hindsight, I now know that my boss at Canon was right — at least in part. When you research as a scientist, you do not sit on a chair and squeeze your brains until a world-changing idea pops out. You can see this best in pharmaceutical research.

StaufenThe research for new drugs is very expensive, very difficult, and takes a long time for a team of people. With this you would suppose that a successful team then invents the miraculous drug that conquers yet another great disease, while everybody else is surprised and stands in awe. In reality, it is not like this. When you look at the patent awards, you will find out that there is always a small number of different companies that files for the same discovery a few weeks apart.

This is not what you would expect given the duration of the research and the secrecy in which the companies operate.

The explanation is that discoveries are in the air or ether. Discoveries happen when the time is ripe for them, and at that time many people will have the same insight with a time interval of a few weeks or months. Research is very expensive, it is a high risk investment. Timing is everything, otherwise you lose your investment.

Timing means that you need to be be at the right place at the right moment. This is why we are in an expensive location like Palo Alto, just a couple of freeway exits from the Computer History Museum. And this is why we spend $500 for a plate of ravioli — which allows us to get the buzz from the ether, emanating from all those reunited luminaries, before the guys working for the competition get it.

The social responsibility of scientists is put out their antennas and transcieve. You cannot do this kind of visceral networking with LinkedIn. You have to be there. There are no shortcuts, no miracles.

Scientists are like bees. A bee can be a busy bee, a worker bee, etc., but by itself it is not worth much. Wham!!! … and you can wack it out with a newspaper. Try that with a bee hive. The art of managing research is like the art of a bee keeper who has learned to create and groom a bee hive.

You are reading my contribution to society, emanating through the ether from my antenna. And this is where my manager at Canon had it wrong — you need to get your hands dirty and do real work, otherwise there is nothing to transmit and you do not know on which channel to tune in. This is why in the wardrobe separating their offices, Bill Hewlett and Dave Packard kept a cart with an oscilloscope, a soldering iron, and small tools.

These days the difficulty is to survive without having your neck broken 24 years later when the job is done, as Dr. Faustus would have told you if his brother in law Mefistopheles had would not have grabbed him first that fatal day in 1539 and illustrated above.

Monday, October 15, 2007

Blog action day: the environment

Today is blog action day and this year's issue is the environment. This blog is on color perception, so I should write about the visual perception of the environment. However, I am not working on complex color and have nothing new and original to write on this. I could brag about all the things HP does for the environment, but you can read that on our Global Citizenship Report site. Instead, I will do something completely different…

Bloggers Unite - Blog Action Day

Dave Packard and Bill Hewlett were lifelong environmentalists, who bought quite a bit of land for conservation. They even instituted a large well equiped park in the Santa Cruz mountains so employees and their families can enjoy nature. And they really enjoyed inviting all employees to BBQs in their parks.

In particular, Bill Hewlett has a lifelong interest in nature. He has photographed and cataloged hundreds of flowers over the years. A decade ago I asked him to send me a few of his favorites. What I got are photos of some of the most beautiful wildflowers of the Western United States.

My contribution to blog action day is to share these photos and let you reflect on nature's beauty.

Rosa Californica

California Wild Rose Rosa Californica

Wild rose is one of less than a dozen species of Rose native to California where it occurs in moist sites below 1800 meters mostly west of the Sierra Nevada. The flowers of this species have been used for perfume, jelly, candy, and tea. The hip, or mature fruit rivals oranges for its vitamin C content. Upon removal of the seeds, the small apple-like hips can also be used for making tea or jelly.

Mentzelia Lindleyi

Blazing Star Mentzelia Lindleyi

As might be inferred by the common name, this plant produces flowers of a rich golden color. The silky textured petals expand to expose the many stamens that stand upright to form a large tuft in the center of the flower that brushes insect visitors with a generous supply of pollen. Plants of Blazing Star are covered with barbed hairs that cause them to cling to whatever they come in contact with. These plants grow on rocky slopes, coastal scrub, and oak/pine woodlands in California typically at elevations below 800 meters.

Epipactis Gigantea

Stream Orchid Epipactis Gigantea

Because of its wide distribution in California and western North America generally and its ability to tolerate a wide range of habitats from near sea level to 2600 meters in the mountains, the stream orchid has avoided the threats that so many of its relatives are up against worldwide. This orchid attracts pollinators by mimicking their food choices without providing a true reward. It is pollinated by syrphid flies that are attracted by a floral odor that mimics the "honeydew" fragrance given off by aphids, but the aphids are nowhere to be found in the flowers of this orchid.

Achillea Millefolium

Yarrow Achillea Millefolium

Yarrow is widely distributed in most countries of the northern hemisphere. Its finely divided fernlike leaves and flat-topped or umbrella-like clusters of flowers make it one of the easiest members of the sunflower family to identify. Its dried leaves which are occasionally used in tea have a mint-like flavor. This plant is probably best known for its medicinal properties. Achilles, for whom the genus is named, evidently used extracts from this species to treat the wounds of his soldiers in the battle of Troy. It avoids the deserts of California but is otherwise common in many habitats below 3500 meters.

Triteleia Laxa

Ithuriel's Spear Triteleia Laxa

The blue to blue-purple flowers of Ithuriel's spear can add dazzling color to the California landscape in years with good winter rainfall. The corms which can be eaten raw or cooked were a favorite food of early California Indians. Ithuriel was an angel in Milton's Paradise Lost who found Satan squat like a toad, close at the ear of Eve, and transformed him by a touch of his spear to his proper form.

Papaver Nudicaule

Iceland Poppy Papaver Nudicaule

Iceland Poppy, originally described from Siberia is a widespread species of arctic regions of North America and Eurasia where it is one of the commonest yet most colorful wildflowers. The silky petals range in color from yellow, white, pinkish-coral, and orange. It is best known in California because it is a favorite garden plant in the cool coastal climate of the Pacific states. Each flower which measures 10-12 cm (4-5 inches) across is borne on wiry stems. They make superb cut flowers lasting up to a week if the flowers are cut in bud and the stalk tip scalded in boiling water before being placed in a vase.

Tragopogon Porrifolius

Oyster Plant Tragopogon Porrifolius

Oyster Plant, a close relative of Chicory, is distinctive because of its narrow grass-like leaves, dull lilac or purple flower heads, and milky sap. In Mediterranean Europe where this plant is native, the young green shoots are added to salads. It is also cultivated for the swollen fleshy rootstock that is cooked and said to have the flavor of oysters. In California, where this plant is introduced, it is a widespread weed of waste places largely unappreciated for its culinary virtues.

Bill's Blooming Hobby

Visitors to a select private Northern California campground have a unique tool for identifying the trees and flowers they see — an album of photographs and copies of identifying leaves assembled by Bill Hewlett. For nearly 50 years, Bill has been studying the plants and trees in all the places where he has spent time. An avid outdoorsman all his life, Bill's career as a part-time naturalist was sparked when the Army stationed Bill and his late wife, Flora, in Washington D.C. during World War II. On one of their frequent visits to Rock Creek Park, he realized that he didn't recognize any of the trees in the area. And when he returned to California, he realized he didn't know much about the trees and flowers here, either.

After reading to acquire a background in botany, he was soon photographing and identifying the trees and wildflowers he saw on camping, hiking, mountain climbing, and fishing trips. Over the years, his collection of photographs has grown to more than 400 different trees and flowers, from areas as diverse as the Santa Cruz and Sierra mountains of California, the American Great Plains, and the mountains of Europe.

Among his favorites from all the beautiful flowers he has photographed are those with the common name Mariposa, including the White Mariposa (Calochortus venustus). The name ties these flowers to the butterflies and Sequoia groves in the foothills and mountains of Mariposa County in eastern California.

The dream of every naturalist, amateur or professional, is to discover an as yet unnamed flower or plant and bring it to the attention of the scientific community. While this has not happened in Bill's years as a naturalist, he still enjoys the challenge of making a difficult identification.

"It is not too hard to make an educated guess as to the genus," he said. "It is the species that is difficult, but the average person is not interested in whether it is an 'Iris douglandiana' or an 'Iris macrosiphon.' Except for the expert, it is sufficient to know that it is an 'Iris.' But there is a challenge to try and find out the species. It is the difference between a job well done and a job half done."

And, as he notes happily, "there will always be new plants to identify."

Thursday, October 11, 2007

More on How Canon got its flash back

As reader juadlam suggests in his or her comments to my previous post on the book about Fujio Mitarai, the comments and questions raised require a new post. First, here is the comment:

So did the book include much about digital photography? The title seems spot on for a good bit of discussion about how their digital cameras came to be so strong in market. I'd be curious if their analysis covers how they seem to have made the transition to digital so well. Also creating a new division seems like quite an undertaking for a research lab. This almost sounds like another post. I expect that this is especially challenging so if the new division has any overlap with the existing divisions. It's probably even equally challenging if there is zero overlap with the existing divisions.

Posted by juadlam on 10/9/2007 3:58 PM

The book is on Fujio Mitarai and not on Canon's technology, but let me try to answer your questions anyway. The question on transitioning from analog to digital has to do with the culture of a company's head honcho, as we affectionately call presidents here in the Silicon Valley. When companies have a lock on a market, their financial success can be increased more easily by investing in a big sales force than investing in technologists. As a corollary, when a leader advances through the ranks to become the president, this leader is likely to come from sales not technology.

In sales, the formula for success is to not kill the goose that lays the golden eggs, and a president with a sales background will conservatively tend to muzzle anybody trying to rock the boat. In contrast, a president that has risen through the ranks as a technologist, will declare that you're not paranoid if they really are out to get you, and the competitors will indeed be out to top your technology.

You can find many case studies on this in business books. Burroughs was a classical example of a company of the first kind. More recent examples are Xerox, where in the late 70s Gary Starkweather (who later coined the cliché "how Xerox stumbled the future") had built the Lilac color laser printer/copier and explained that you can do color xerography only digitally, while his corporate management wanted to hold on to light lens processing. And of course Kodak, who early on invented many digital color technologies only to have corporate management stuck on AgX and photochemistry.

For the second kind of companies, the most vocal one is perhaps Intel with their motto that only the paranoid survive. In HP, Dave Packard had the business rule that at least 80% of the product catalog had to be in the catalog for 18 months or less, and Bill Hewlett's mantra was that HP had to create new divisions killing the old divisions before the competition did it.

Canon is such a technology company. While Xerox was busy fighting with digital vs. light lens, Canon was busy developing the digital color laser copier CLC-1, which was an immediate smash hit. Behind the scenes, Susumu Sugiura (a.k.a. Sid Sugiura in Australia), had built a large team with deep knowledge in digital color imaging. At the Canon developer conferences in 1991 and 1992 they held workshops on color appearance modeling, demonstrating they were ahead of the bleeding edge.

In 1993 the Imaging Research Center in Shimomaruko started the Digital Eye project with an initial staff of 100 R&D personnel. At the 1996 EI conference, the discussion of Yoshiro Udagawa's paper Color image processing in Canon's digital camera demonstrated a very deep understanding of the image processing for digital cameras and especially of how to make trade-offs between the various parameters.

Kumada and YamadaIn 2000, Canon started a company-wide movement to establish a unified standard for high image quality in all of their products, from input to output, which they called the "concept of Canon's unified high-quality color system." The technical wizards behind this effort were Shuichi Kumada and Osamu Yamada — portrayed at right — and the result was the Kyuanos color management system.

Essentially Kumada and Yamada tossed the sRGB color model operator and the ICC profiles with all their limitations out of the window and build a new system from first principles, based on color appearance modeling. Kyuanos is implemented in all Canon products and in the case of the digital cameras you are asking about, it is implemented in hardware as part of the DIGIC chip, which is at the core of all of Canon's cameras.

In essence, Canon's image processing is so good because they have been at it consistently for more than 25 years. The people behind it have become so good at what they are doing, that part of Kyuanos was even adopted by Microsoft for their Windows Vista operating system.

As I mentioned in my previous post, grooming people to excel as leaders is a difficult task but it is a crucial task for technology companies. In Canon's case, in phase III of their Excellent Global Corporation Plan, one of the key strategies is Nurture truly autonomous individuals to promote everlasting corporate innovation, which they express as follows:

For Canon to become a world-class company, our employees must strive for excellence. From a human-resource development standpoint, we will further enhance our education and training programs to cultivate capable employees who are trusted by society, and encourage employees to put into practice Canon's "Three Selfs" guiding principle. At the same time, we will step up efforts to develop insightful global leaders and business managers who actively contribute to not only progress at Canon, but also to the business world and society as a whole.

In the case of science and technology, this results in the Canon Academy of Technology with the theme Specialists Cultivating Technology.

Tuesday, October 9, 2007

The arcane art of leadership gestation

Today I will lift a bit the kimono to give you a glimpse on this aspect of governance. I barely have enough time to stay alive, so I apologize to use a compact European writing style instead of the more eloquent American style I am supposed to use in this blog. The occasion is today's Nobel Prize announcement.

When I used to have work assignments in corporate governance, the only business book that really helped me was Gordon Bell's book High-Tech Companies. Of course the most important lesson was on how to organically grow a balanced company, but there was also the lesson on the pygmy principle and how to build the company's leadership team.

In most herd animals, leaders are selected in duels. However, early on humans have developed the art of gestating — or grooming, in Silicon Valley lingo — leaders. It probably started with shamans, but by the time of Egypt's first dynasties it was already a well developed structured and formal process assigned to the monasteries, an institution the Pharaohs most likely invented for this specific purpose.

In Far Eastern cultures the main contributor to this art was Confucius, who coined the term naming names for what here in the Silicon Valley today we call pygmy hiring when it is done poorly (see for example Ryûichi Abé's The Weaving of Mantra — Kûkai and the Construction of Esoteric Buddhist Discourse for a detailed historical analysis of how the specific method used for naming names profoundly influenced the Japanese culture and was the germinating event for the formation of Shingon). The reason I mention this specific book is that it always was — and in good part still is — an art beyond the reach of the general population, i.e. it is esoteric.

If in the past it was esoteric, today it is mostly based on wisdom and implicit knowledge, which allow the leader gestator to extrapolate current trends, assign them as directions to follow, select gifted individuals, nurture them, and finally, when they have achieved, laudate them publicly so society can follow them as examples.

When are the gestators themselves recognized? Quietly, when they have successfully predicted leadership. Today, out of sight, a former HP Labs director and the members of a selection committee in Japan are quietly celebrating their successful early identification of leaders.

Today the event is giant magnetoresistance (GMR). You can read about it and the inventors all over today's press and blogosphere because they just received the Nobel Prize in physics.

Recognition goes to Chuck Moorhouse, who at an early time recognized its merits and had HP pursue research on this theme.

Recognition goes to Koichi Kitazawa, Takehiko Ishiguro, Hidetoshi Fukuyama, Tatsuo Izawa, Tetsuya Osaka, Katsuaki Sato, Junichi Sone, Kohei Tamao for recognizing the importance of this basic research in inspiring innovative devices and giving them the 2007 Japan Prize.

And now let's close the kimono and move over to Albert Fert and Peter Grünberg, and their laudation on Nobelprize.org.

Monday, October 8, 2007

Color stereoscopic images

Researchers in Israel have shown that we perceive 3-D color images even when we are presented with only one color image in a stereoscopic pair, with no depth perception degradation and only limited color degradation.

The latest print issue of SPIE's Optical Engineering dated August 2007 (Volume 46, Issue 8), has an interesting article on page (or I should write Citation Identifier, CID) 087003 with title Color stereoscopic images requiring only one color image. This paper is a beautiful piece of color psychophysics, in which the experiments were conducted both with a 1905 stereoscope and with a state of the art head-mounted display (HMD).

Stereoscopic images yield a much improved depth perception and operator performance. However, the amount of information transmitted is doubled. Obviously the left and right images contain a lot of redundant data, and various methods to compress motion images have been proposed to reduce the data stream, although a considerable computational cost hit must be taken.

The authors asked themselves, if the human visual system's fusion capability can be used to process color only of one eye's image, processing the image for the other eye just in luminance. This would cut down both device cost and data before any compression has been performed.

Indeed, the psychophysics results show that subjects perceived 3-D color images even when they were presented with only one color image in a stereoscopic pair, with no depth perception degradation and only limited color degradation in the form of a loss in vividness.

Monday, October 1, 2007

Mini review. How Canon got its flash back

Published in 2004, this book is not new. However, it was published by John Wiley Asia in Singapore, so if unlike me you do not periodically check out a Kinokunia book store, you probably never came across it.

How Canon got its flash backHow Canon got its flash back was written by the editorial team of NIKKEI, which stands for Nihon Keizai Shimbun, Inc. and is the Japanese equivalent of Dow Jones here in the U.S.

In my opinion, the title promises more than the book holds, because it is not a critical business review of Canon, as we are used to get when we read similar American books about high-tech companies. In fact, the subtitle The innovative turnaround tactics of Fujio Mitarai would have been a much more appropriate title, because the book is a laudation of Fujio Mitarai.

Indeed, we learn about the positive changes Fujio Mitarai has introduced, like a better integration of the hundreds of companies that make up Canon, the introduction of a consolidated balance sheet, accountability, the ability to get an up-to-date status of the company, and the restoration of lifetime employment.

Here and there the book relates to the reader Fujio Mitarai's thoughts on various corporate governance topics, like the appointment of external directors, the role of auditors, and how you implement meritocracy in a traditional Japanese company.

Regarding manufacturing—which is a key Canon competency—the book explains in detail how cell method and ma-jime (closing the gap) was introduced and how it paid off (Chapter 2).

What we are never told in this book is what happened before Fujio Mitarai. On page 155 we learn that "The rapid appreciation of the yen in 1986 led to a sharp drop in the company's profitability. When this was then compounded by the deflating of the bubble in the Japanese domestic economy, the period from the mid 1980 to the mid 1990s turned into something of a 'lost decade' for Canon."

This concept of the lost decade comes up several times in the book, but we are never given a satisfactory explanation. In fact, the bubble did burst in 1993 and Canon had very rough time, with layoffs and abysmal employee morale. However, this cannot be the whole story.

Reading the book we are left with the impression that the lost decade was more akin to the Warrying States period in Japan, also known as Sengoku period. The book should have a chapter on this lost decade, which should answer the many questions the book leaves open. Indeed, while the book covers in detail the period of Canon's first president Takeshi Mitarai, it is completely silent about the presidents between the founder and Fujio Mitarai: Takeo Maeda (1974-), Ryuzaburo Kaku (1977-), Keizo Yamaji (1998-), and Hajime Mitarai (1993-).

Did they screw up? Where they unable to control the "war lords"? If so, who were these? During the lost decade, when I was asking Canon Inc. employees why something happened, the standard answer was to watch Ran (Chaos) and then I would understand. I got an idea, but I did not really understand who King Lear was and who Hidetora's sons were.

So the book has these strange voids, such as the Central Research Lab being like a magic castle that suddenly disappeared from Atsugi only to reappear in remote Susono in Shizuoka province, beyond Hakone. Was there a carnage like when in 1571 Oda Nobunaga destroyed the Enryaku-ji monastery.

Why was the Central Research Lab not moved to the Shimomaruko campus, like Yamaji did with the Headquarters? From the book we get that the scientists must have been more unruly than Enryaku-ji's sôhei (warrior monks), because there is a whole section entitled "Discipline paramount." Why did Canon have to implement the rule of the Five Ss: proper arrangement (seiri), cleanliness (seiso), orderliness (seiton), neatliness (seiketsu), and discipline (shitsuke), as well as Communal Possession and Functional Beauty?

When the authors write on page 78 that of these shitsuke is the most important, and on page 81 that a dress code had to be drawn up, which forced researchers to wear a prescribed jacked and forbade the wearing of jeans, one must think that these researcher must have been quite an unruly pack. This is difficult to understand when Canon historically had the tradition of cultivating their staff as heros and still continues to do so as evident from their Web site The Minds Behind Magic Special Interview.

Indeed, historically Canon has excelled in virtue of its principle of strategy being a top down process and tactics being a bottom up process. For Canon science and technology have never been intangible assets, but always brains attached to bodies that are nurtured. Today this is exemplified by their Canon Academy of Technology as depicted in the Web site Specialists Cultivating Technology.

Comparing to HP Labs, where the emphasis is on alignment with the Divisions, in Canon's Central Research Lab the emphasis is on the creation of new Divisions (page 156). Thus, one would expect their researchers to be disruptive revolutionaries or sôhei, not disciplined soldiers. Indeed, it contradicts Canon's Thee Selfs concept (page 110): self-motivation, self-management, and self-awareness.

Finally, there is the mystery of the prologue, which chronicles the exit of the PC business. This is described as the divestiture of FirePower. The FirePower system was not a business or consumer PC, it was a workstation. Its architecture with two PowerPC processors and a signal processor made it one of the best imaging systems available at that time, that would have been the ideal platform for embedded systems for a high-end printer and copier architecture.

Equally mysterious is the complete lack of any reference to Canon's competitors, such as Ricoh, Fuji Xerox, Nikon, Epson etc. Without having an idea of the ecosystem in which Canon operates, it is hard to form an overall appreciation of Fujio Mitarai's merits.