God's Goiter is Michelangelo's Brain

Artists like Leonardo da Vinci and Michelangelo were distinguished by their detailed understanding of anatomy; knowledge declared heretical by the Roman Catholic Church and therefore something that necessarily had to be acquired as a clandestine activity. Michelangelo even destroyed most of his anatomical sketches and notes. But not entirely. It seems that he hid some of them in plain sight of the Pope—on the ceiling of the Sistine Chapel.

Looking up and under at God's goiterous throat
Art historians have long been puzzled by certain subpar anatomical irregularities in Michelangelo’s work, e.g., the panel depicting The Separation of Light from Darkness shows a goiterous formation in God’s neck. Twenty years ago, it was suggested that the region belonging to God Creating Adam, in a central panel of the ceiling, was actually the human brain in cross section. Now, a paper in the May issue of Neurosurgery, identifies the "irregularity" in the Separation panel as an anatomically accurate depiction of the human brain, viz., spinal cord, brain stem, eyes and optic nerves.

Accounting for R&D

People who choose a career in science and technology — where work is hard and poorly paid, the nights are long and lonely — tend to be on the lower side of social skills. Therefore, research institutions have always had a certain level of dysfunction.

Wake Up and Smell the UV

From the This-Actually-Got-Funded department comes, fruit-fly larvae that have been genetically engineered to smell light.

"...given their successful mapping of these larvae olfactory neurons, the researchers next hope to make adult fruit flies go bananas."

top 10 colors for women's fashion for fall 2010

In living color

Light produced using the chemical process of bioluminescence spans the entire range of the visible spectrum. Bioluminescence has evolved independently several times in the tree of life. However, the majority of bioluminescent organisms reside in the open ocean, where their bioluminescence helps species in over 700 genera evade predators, attract mates, and find food. In Science Volume 328, Number 5979, Issue of 07 May 2010 (p. 704–708), Dr. Edie Widder reviews recent advances in understanding the evolution and distribution of bioluminescence in marine systems.

Mrs. Nonaka's garden

Mrs. Nonaka's garden — which she prefers to call her backyard — is a quiet shady oasis in Silicon Valley. Her book club enjoys it for her food and the Chardonnay … and to discuss books; while entrepreneurs like to congregate there for the nonchalant atmosphere and the fusion cuisine served from the grill. Under the trees, the garden is surrounded by rhododendron in various degrees of insolation, so there almost always are some in bloom.

Unfortunately, not all is well with Mrs. Nonaka's rhododendron. When it wakes up from winter, the leaves start becoming increasingly whitish.

The explanation is under the leaves:

There, white critters suck out the sap from the leaves, leaving brownish spots. So, every year she straps the sprayer on her husband's back, who washes down the rhododendron with insecticidal soap.

In the pictures above, the critters have already been washed away, but they are yellow-whitish. They do stand out against the green leaves, but either have no natural predators in Mrs. Nonaka's garden, or they are camouflaged in their predator's visual response, or their color warn (aposematism) chasing them away, as we saw in the post from Sidney.

A propos critters sucking the sap out of leaves, consider the pea aphid Acyrthosiphon pisum shown in the picture below. It comes in a brownish to reddish variant attacked preferentially by lady beetles and a greenish or yellowish variant into which parasitoid wasps deposit eggs.

According to the classical view on the evolutionary ecology of animal color polymorphism, it can be hypothesized that these opposite predation and parasitism pressures maintain the color variation in natural aphid populations.

But where does the red color come from? As the flamingo painter in the Color Selection video in the 1990 ACM SIGGRAPH Video Review, Mrs. Nonaka knows very well that although chlorophyll is related to carotenoids, the aphids would have to feed on something with red carotenoids, such as torulene. This is because animals cannot make their own carotenoids, forcing color scientists to eat lots of carrots to keep their eyes in top performance.

The riddle has a very non-obvious explanation, that is revealed in a recent Science paper by University of Arizona, Tucson's Nancy Moran and Tyler Jarvik: Lateral Transfer of Genes from Fungi Underlies Carotenoid Production in Aphids.

The recently published draft genome sequence of A. pisum provided a crucial clue to the origin of the aphid carotenoids. Phylogenetic analysis revealed that, strikingly, all the aphid carotenoid synthetic genes clustered with fungal genes with high statistical confidence. Genomic features of these genes strongly suggested that they were transferred from a fungus to an aphid ancestor, and subsequently subjected to duplications in the aphid genome.

So far, this is the first animal of which we know it is able to make its own carotenoids.

By the way, phylogenetic analysis is a tool also used by co-blogger Nathan for his research in color science.

Mrs. Stine's garden

Some 25 years ago, when I was exploring my new homeland, I came across Mrs. Isabel Stine's garden on highway 9. Mrs. Stine — often referred to as Mrs. Oliver Stine — is better known as the patron behind Gaetano Merola, the founder of the San Francisco Opera. In fact, Puccini's Madama Butterfly (the most-performed opera in the U.S.) had its West Coast premiere in Mrs. Stine's garden. Oliver and Isabel Stine also financed the construction of the Palace of Fine Arts for the 1915 Panama-Pacific International Exposition.

It was at this exposition, that Mrs. Stine fell in love with the Japanese tea house. The same year she bought the 18 acre property on highway 9, in Saratoga.

During my first visit, I was not impressed by the garden, with its buildings infested by termites and mother Nature clearly having the overhand on the gardeners, with the formal Japanese garden almost disappeared in the wild Santa Cruz Mountain vegetation. Therefore, I never returned, until recently.

I was surprised by the restoration. The garden was taken over by a foundation, which expanded on Mrs. Stine's construction achieving an almost complete formal Japanese garden. The foundation built a new Cultural Exchange Center (CEC), which is rented out for executive retreats.

Although the garden is strictly Japanese, its is used for various ethnic events, showcasing the multi-cultural amalgam of the Silicon Valley. For example, this year it hosted a Chinese New Year festival, while next year it is planning a Nowruz festival for the Persian New Year.

I do not know why it has been named Hakone Garden by Mrs. Stine, because it does not evoke Hakone (elevation 410 m) to me. In fact, to me the view of Silicon Valley from the CEC terrace (elevation 130 m) is more evoking the view of Kyoto (elevation 40 m) from a terrace of a restaurant in a formal Japanese garden along the Philosopher's Walk.

This is a view from the CEC:

This is a view from a restaurant in Kyoto (at right is Mrs. Nonaka, see next post):

But then, everybody has their own experiences. Evoking is also the source at the restaurant in Kyoto

and that at the upper tea house in Saratoga's Hakone garden (note the effect of climate on vegetation):

which is used

for tea ceremonies (茶道 , sadô)

Speaking of tea, Hakone Garden has also a larger tea room:

In fact, the entrance to this tea room

reminds me to a similar view on the Philosopher's Walk:

Some more images from Hakone Gardens in Saratoga:

(to be continued)

A neat example of color break up (blue-yellow)

Color break up happens for example when color components of an image to be displayed are presented in a time sequential manner. How would you demonstrate 'color break up' in a coltrolled way? Micheal Bach published a visual illusion in which the observer is asked to focus on a particular point in the image and moves the test object. In addition, once in a while the test disk changes color for an instant and you can watch what happens (just click on this link: http://www.michaelbach.de/ot/mot_flashlag1/index.html):

Saving an investment

Almost every time I step out and encounter a former colleague in my age group, I hear the story of the wasted brains of Silicon Valley. Although, the official
unemployment rate (seasonally adjusted) for California is 12.6%, the news media has variously estimated the number of older unemployed or underemployed R&D people here to be rather slightly above 50%.

Ball Lightning Hallucination—Balls!

This illustration depicts the rare phenomenon of ball lightning. It appears on p. 55 of Wonders of Science Simplified (Metro Publications, New York, 1949) and shows a lightning ball erratically ricocheting around and, having descended the chimney, traverses the interior of a farmhouse and finally explodes.

When I was eight years old, my much older neighbor gave me his copy of this book—presumably because he recognized my interest in science was more intense than his own. I still have it and although I remember this picture well, I never thought I would see such a thing. I was wrong.

Years later, as a young teenager, my father and I were looking out of my bedroom window at the "fireworks" on display during a severe electrical storm. Like the illustration, my bedroom was on the second story of our house so, we had quite a panoramic view.

Why <img>? How <img>?

The teaser posting of a Deep Dive into HTML5 includes an interesting section on the question: Why <img> for HTML?

The <img> tag is of course the HTML tag for an image.

Which is such a simple question given the possible alternatives at the time, such as <icon> and <include>. It's a good read. As is the observation that "shipping code wins".

Avoiding a re-posting of the seven principles that drove a competing vision for HTML (and a lengthy digression on color standards), the only unique contribution of this post is how to create a Blogger post about a tag.

Do not use the greater than and less than signs directly - instead use the & l t ; and & g t ; (with the spaces removed) otherwise the title field will try to interpret the tag directly. Which means it will disappear. Instead use something like this:



Ironically I used a hand-written <img> to embed the above image into this post. Of course getting the post title right doesn't necessarily mean that search engines will but at least your tagophilic readers might appreciate the effort.

Parallel Processing for Imaging Applications

Parallel Processing for Imaging Applications

Conference EI111

Part of program track on Image Processing

This conference has an open call for papers:

• See submission guidelines for Authors & Presenters
• Submit an abstract

Red is the New Black

For those of you that haven't forwarded it to us already, yes we know that that xkcd has conducted a color naming survey and has posted some analysis of their color survey results. Fun stuff, with interesting similarities (and differences) with some of the work we've done over the years relating to color naming. To maintain parity, we offer the following cartoon response to their survey.



And yes, the comic sans font was used purely for it's inflamtory powers.

SSH-2 bites

We all have to deal with constant massive attacks on our servers, so we have to zip them up very tight and communicate with them using the secure shell protocol SSH-2. The first set of problem arises when we have to propagate our private keys in our zoo of machines. Submit this problem to Google and you will be served a gazillion war stories. The fundamental reason is that SSH-2 does not specify a file format and each utility has its own incompatible format; on top of this, add the different end-of-line conventions used by the various operating systems.

I have done this so many times that I thought I can do this in my sleep, but today I got bitten very badly because a GUI hid from me an error message that was revealed only when I finally used the terminal. As a pro memoria, let me recapitulate the procedure.

28 Colors with 1 Impression

A short video on Pantone, including some footage of the production press used and an iPad-based "swatchbook" (via Tim via Sharon).

The beamer

Here in the US, 'beamer' is known as Till Tantau's LaTeX class for creating fancy presentations with very little effort. However, on the other side of the big pond, 'beamer' refers to a digital overhead projector, and this is where Till got the name for his class. Similarly, few people know the beamer's historical origins.

The history goes back much further than many believe, to a train ride from Bern to Zurich late 1939. The passenger was ETH professor Fritz Fischer, and he filed his patent application the 8th of November 1939. The fundamental idea was to separate image formation from light source.

More in detail, the idea was to coat a concave mirror with an oil film and to write on it the image with an electron beam. Where electrons hit the oil film, bumps emerged, creating a relief of the image corresponding to the bright and dim image points. The mirror was then illuminated with a very bright xenon gas discharge lamp. Where there was a bump, the light would be reflected through a lens onto the projection screen, while where there was valley, the light would be reflected back into the lamp.

Prof. Fischer called his invention the eidophor, Greek for image carrier. The image below shows the first eidophor used in 1943 to project TV programs in a large auditorium.

This first prototype had the problem that the oil film would decompose very rapidly. The problem was solved in a second prototype, but Prof. Fischer died a few days before the scheduled public unveiling the 31st of December 1947 at age 49.

The research was carried on at the ETH by Prof. Ernst Baumann, who in 1948 demonstrated a new prototype with movie quality (a color version was demonstrated only later, in fall 1951).

At that time, it was believed that the future of television would be deployment in movie theatres. Only few believed TV would instead make its way into living rooms on tiny screens. For this reason, for the commercialization of the eidophor, the scientists sought the collaboration with a company having good connections to the US and its movie industry.

They found their man in a Stanford graduate who had married a local Palo Alto girl and then returned after his Ph.D. to his native Zürich-Altstetten to start his data encryption company, Dr. Edgar Gretener AG.

As a side story for the color scientists among our readers, Dr. Gretener later renamed his company Gretag AG, moved it to Regensdorf, and also produced industrial automated photo printers and instruments for the printing industry. When his engineers developed a new tiny handheld spectrophotometer and Mrs. Gretener saw it, she exclaimed "Oh, it's just like a Topolino" [the nickname of the Fiat 500]. The name stuck and the instrument is called 'Lino' internally and 'Spectrolino' on the market.

Anyway, Dr. Gretener had the required connections and late 1950 a team from 20th Century Fox traveled to Zurich to look at the technology. Although they enthusiastically proclaimed they would order over 1000 units, they never sent a P.O.

Therefore, Gretag marketed the eidophor to conference centers and schools for closed circuit television (CCTV). The first mass-produced model was the 1959 Eidophor ep2, shown in the figure below:

It was quite popular for the projection in large rooms, like for lectures in universities, shareholder meetings, sports events, and contests. The figure below shows an eidophor deployed in the Zurich Congress Centre in 1959 for the live transmission of the Federal government elections in Bern.

When I was a student, the Eidophor was used extensively in experimental physics lectures, because the experiments being performed in the front could be seen in all detail by all 350 students filling the auditorium. All large auditoriums at the ETH were equipped with an eidophor. The image quality was excellent, because the oil film acted like a low-pass filter, giving smooth and rich images without noise. By comparison, 16 mm film had a lot of noise from the film grain.

Eidophor was the dominating large format projection technology until the 1990s. Gretag exited the projection market in 1997 and beamers based on LCD and DLP technology took over the market.

For the detailed story of the eidophor, see Caroline Meyer: Der Eidophor. Ein Grossbildprojektionssystem zwischen Kino und Fernsehen 1939–1999. Chronos Verlag. Zürich 2009.

Images: Library ETH Zürich.

Location, location, location

Today, many conglomerates are not run by visionaries but by gnomes toting spreadsheets. The key factor for success is then the objective function being optimized by the model. More often than not, the objective function is cost minimization.

Because there are still substantial salary differences around the world, but all scientists are equally dumb (otherwise they would all work in the financial industry), it boils down to location. You can convince yourself of this by driving around the research parks of the Silicon Valley: it looks more and more like Bodie, although some optimistic realtors still leave up the "for lease" signs. The research labs moved from Palo Alto to Beijing, and from there to the Amazon. Who knows where the objective function will send them next.

Unless, somebody uses a different objective function. On 30 April 2010, Disney opened a new research lab in the most expensive city in the world: Zürich in Switzerland, joining companies like IBM and Google, who also have research labs there. Obviously, they concluded that the best location is not the one with the lowest wages, but the one with the sharpest brains, all in a fertile eco-system.

The story began four years ago: Disney's new CEO Robert Iger and Edwin Catmull, chief of the Pixar animation studios that belong to Disney since early 2006, decided to perform more research in house. Through Joe Marks, ETH professor Markus Gross prepared the winning proposal for a joint venture between ETH and Disney.

Disney is paying the high Swiss salaries, while the ETH is contributing the infrastructure. This includes the building at Clausiusstrasse 47. The created intellectual property and their licensing fees will be shared between ETH and Disney. At the opening April 30, Disney Research Zurich started with 20 researchers, which will grow to 40 by 2011.

In the portrait above, Prof. Markus Gross is holding cheap 3-D goggles, but the agenda for Disney Research Zurich goes much further than that. For example, part of the eco-system in Zurich is a small company selling a mobile autosteroscopic video player the size of a paperback for just 400 francs. This video player presents 3-D movies without the spectator wearing special glasses.

Not far from Disney Research Zurich, near the city zoo, is FIFA, the Fédération Internationale de Football Association, who is organizing the 2010 FIFA World Cup South Africa. Disney's ESPN sports channel has announced that is will broadcast the Johannesburg games in 3-D. This will bring big financial opportunities for owners of virtual sky boxes around the world.

One especially hard problem in presenting 3-D contents on a 2-D surface is vergence, and this is one of the areas being researched at the new laboratory, for now with Swiss cows:

Funnel vision: Waveguides in your retina

The light-sensitive rods and the color-perceiving cones are at the back (posterior to the pupil) of the 10-layer retina. Why is all that junk in front of the sensors? New research indicates that the intervening layers contain glial cells (called Müller fibers), like the white matter in your brain, which act as waveguides to improve photoreceptor efficiency.

The performance of JPEG implementations

1. DCT ALGORITHMS

JPEG compression is a method consisting of the following steps:

1. Transform the color coordinates to an opponent color system
2. Subsample the chromatic coordinates, for which the human visual system's (HVS) MTF is about half
3. Decorrelation: perform a discrete cosine transformation (DCT) to de-correlate the spatial
   information
4. Quantization: truncate the coordinates to exclude spatial information invisible to the HVS
5. Entropy coding: compress the data using run-length encoding followed by Huffman encoding