Snowflakes have become an obsession of mine as an extreme macro photographer, but I never thought I would be able to take it this far: 2,500 hours of work across 5 years, all presented as a single composite photo titled “The Snowflake”:

Depicted in the composite image above are over 400 unique snowflakes, all accurately measured and scaled to that they are all in relative size to one another.

On average, 40 separate images are combined for each individual snowflake. This is required to get the crystal in focus from tip to tip with a process called focus-stacking. Due to the nature of the subject and the hand-held approach to photographing each snowflake, 4-5 hours are spent on each image in post-processing. Why so long?

Each snowflake is photographed on an angle, which allows for reflected light to hit the surface of the crystal and bounce back into the camera lens brightly. If the snowflake was photographed straight-on, the angle would require the light source to be coming from inside the lens itself.

This angled shooting results in only a tiny slice of the subject being in focus, and the hand-held nature of shooting these sky crystals means vertical/horizontal/rotational shifts need to be corrected for. Photoshop can re-align the layers, but it can’t correct for perspective and it has trouble around certain edges when blending the layers; these problems need to be corrected manually.

Measuring snowflakes is a time-consuming task, and the right equipment is needed to get accurate results. Thankfully, a hidden piece of metadata recorded by the Canon MP-E 65mm F/2.8 1-5x Macro lens makes this possible: the magnification factor. Combined with the physical size of the sensor and the total number of pixels across the sensor, an algebraic equation allows us to calculate the number of pixels per millimeter and measure the crystals.

The process of measuring snowflakes is made more difficult by the removal of certain metadata when editing. The “magnification factor” value is stored in a special area of image metadata called “makernotes”, and can be extracted by tools such as ExifTool. This special section of metadata is removed from the file when processed through any Adobe software (and I’m sure others), requiring me to revisit the original raw files for each snowflake to obtain the proper value.

The largest snowflakes measure just over 11mm in diameter, and the smallest are 0.2mm across. Different storms create different kinds of crystals, some symmetrical but always unique. No two snowflakes falling from the sky will ever be identical. This poster shows the beauty in their variety.

This image not only shows the beauty of winter, but it’s an eye-opening experience when you dive into the details.

The working file on my computer is 12 gigapixels in size, so I could theoretically make 60×90-inch print at 480dpi if I wanted, losing no detail in the process. When working on the file, the computer utilizes over 100GB of RAM, so creating something like this required computer hardware up to the task.

“The Snowflake” is available as a poster print with artist’s proofs available as well.

About the author: Don Komarechka is a nature photographer specializing in macro photography. His work with the subject has been highlighted in international publications as well as being featured in an episode of The Nature of Things with David Suzuki. Don is an author, educator, and adventurer with a passion for revealing “the unseen world”. A gallery of his individual snowflakes images can be seen here and his book on the subject here.

‘If savvy consumers can help predict a product’s success, might there not be consumers whose clammy embrace spells its death?’

Spare a thought for the poor darlings who run your typical transnational, fast-moving, consumer goods company. They invest millions launching an exciting new product (Colgate ready meals, say, or Cosmopolitan brand yoghurt. Or Crystal Pepsi: it sounded like a Class A drug, it looked like water and it tasted pretty much like any other kind of Pepsi. How could it fail?) They give the product to focus groups, who like it. They trial it in a few select stores, and it sells well. Their retail partners are convinced. Then the product is launched to global fanfare, and the ungrateful customers refuse to buy it.

It is no secret that many new products fail. Naturally, companies are on the lookout for ways to identify failures earlier in the process, before they have sunk too much time and money into a product that will eventually collapse. If you’re going to flop, it’s better if you can do it quietly in the changing rooms rather than from the highest diving board with camera bulbs flashing.

Back in the mid-1980s, Eric von Hippel, a management professor, suggested working with what he called “lead” customers — people with more advanced or specialised needs whose demands might predict where the market could be heading. Today’s military technology is tomorrow’s household appliance; today’s professional imaging software is tomorrow’s smartphone app. In the innovative industries that interested von Hippel, paying attention to lead customers would produce great ideas for the mass market.

In more everyday sectors, lead customers might instead be fashionistas whose choices were copied by others. Or lead customers might simply be cutting-edge consumers of music, or coffee, or gluten-free cakes, who are always one step ahead of where the herd is already going. In any case, the strategy for a business is clear: identify these lead customers if you can, and pay attention to what they do and say.

Von Hippel and co-authors wrote in the Harvard Business Review in 1999 that “all processes designed to generate ideas for products begin with information collected from users. What separates companies is the kind of information they collect and from whom they collect it”.

What they meant was that companies should consider the ideas of lead customers, rather than gathering a group of John and Jane Does and showing them a product prototype. But lead customers aren’t the only unusual people who might be worth paying attention to.

If savvy influential consumers can help predict a product’s success, might it not be that there are consumers whose clammy embrace spells death for a product? It’s a counter-intuitive idea at first but, on further reflection, there’s a touch of genius about it.

Let’s say that some chap — let’s call him “Herb Inger” — simply adored Clairol’s Touch of Yogurt shampoo. He couldn’t get enough of Frito-Lay’s lemonade (nothing says “thirst-quenching” like salty potato chips, after all). He snapped up Bic’s range of disposable underpants. Knowing this, you get hold of Herb and you let him try out your new product, a zesty Cayenne Pepper eyewash. He loves it. Now you know all you need to know. The product is doomed, and you can quietly kill it while it is still small enough to drown in your bathtub.

A cute idea in theory — does it work in practice? Apparently so. Management professors Eric Anderson, Song Lin, Duncan Simester and Catherine Tucker have studied people, such as Herb, whom they call “Harbingers of Failure”. (Their paper by that name is forthcoming in the Journal of Marketing Research.) They used a data set from a chain of more than 100 convenience stores. The data covered more than 100,000 customers with loyalty cards, more than 10 million transactions and nearly 10,000 new products. Forty per cent of those products were no longer stocked after three years, and were defined as “flops”.

. . .

The harbinger customers are those who buy lots of flops and, in particular, those who buy flops and then go back for more, repeatedly buying the same unpopular product. It turns out that having identified these flop-loving customers, you can get a good idea of future failures by watching whatever they buy next. This is interesting: in principle there’s no reason why a customer who loves an unpopular flavour of soft drink would be more likely to also love an unpopular brand of shampoo.

In practice, however, the data show a class of people with an eclectic taste in products that others dislike.

Anderson and his colleagues reckon that their results are robust within the particular context of the convenience store data. Whether the technique could also be used for films and books, or computers and tablets, remains to be seen. But it’s already a good example of the kind of patterns that emerge from much larger and more detailed data sets than ones traditionally available to social scientists.

A final question is whether you can spot Harbingers of Failure without access to their shopping habits, purely from demographic information. The answer seems to be no. Harbingers of Failure are much like the rest of us. The only difference is that they love the products that we hate.

Written for and first published at ft.com.

Romanian photographer Radu Zaciu has been experimenting recently with photographing fruits and vegetables using internal lighting. He places bulbs illuminates the produce from the inside and uses that as the main only light source for his series of glowing images. The project is called “The Light Inside.”

He starts by drilling or carving holes into the fruits and veggies. The holes need to be big enough to fit a light bulb, but not so big that it completely removes the flesh of the “subject” (which helps give a glow color that unique to the fruit).

“It all started with a game of words,” Zaciu tells us. “The German word for light bulb is ‘gluehbirne’, what literally means ‘glowing pear’.” His first attempt was literally with a pear, which he photographed as a light bulb:

After seeing how beautiful the internal lighting made the fruit, he decided to try the same thing with other fruits and veggies. Zaciu used small light bulbs from flashlights for smaller fruits and 220V 40W light bulbs for larger ones.

The amount of light shining through is quite weak, so Zaciu shoots in complete darkness and uses long exposure times of more than 10 seconds to properly expose the subjects. He also throws in a tiny bit of light painting for things like leaves in order to light them for the shots.

Here are some of the photographs in the project so far:

Cauliflower

Cabbage

Strawberry

Potato

Kiwi

Fennel

Cauliflower

Endive

Christmas Melon

Celery Root

Cabbage

Pineapple

You can follow along with this project and find these images in higher resolutions in this Flickr set.

Image credits: Photographs by Radu Zaciu and used with permission

These have to qualify as some of the coolest car photographs we’ve ever seen. Yes, we’ve seen some amazing composites and some real live racing photos that are quite epic in their own right, but these images by photographers Dmitry Chistoprudov and Nikolay Rykov have the entire Internet wide-eyed.

First important note is that the photos aren’t faked, and we have the behind the scenes photos to prove it. Christoprudov and Rykov actually wanted to illuminate a car from below 1 meter (3.28 feet) of solid lake ice — you know… weekend project — and so they set about doing just that.

In the end they succeeded, as the photographs above and below go to show, but not before encountering all sorts of challenges.

The first problem they had was getting a waterproof light source that would be strong enough. By Rykov’s estimation, they needed something that could shine around 1000Lm, and they needed it on a budget. In the end, a desperate Google search yielded the MagicShine MJ-810B.

Then they had to actually drill the hole in the ice, which they did with an ice drill borrowed from a nearby village. Only one problem with that: it wasn’t wide enough. “The diameter of the hole was too small to let our light come through,” explains Rykov on his LiveJournal (translated by English Russia). “So we had to stop fishermen on the roadway and offer them 500 roubles for widening our hole with their special tool.”

That led to an hour worth of cleaning up scratches left by the fishermen’s spiked boots and clearing away the ice that had been dug up. Finally, after all of this, they were ready to lower the light and shoot. Here are some of the behind the scenes photos:

All of the work was well worth it. The images are going viral and the story is a testament to ingenuity, creativity and not giving up when your ice drill is too narrow (a problem I’m sure we’ll all face at some point).

You can find out more about the project by either Google Translating the duo’s Live Journal post or checking out the small bits that have been translated more accurately by English Russia here.

(via Jalopnik)

Image credits: Photographs by Dmitry Chistoprudov and Nikolay Rykov and used with permission

Last week, WikiLeaks made public a portion of a treaty that the White House has been secretly negotiating with other nations and 600 special interest lobbyists. The draft of the Trans-Pacific Partnership Treaty, which is on intellectual property, shows that HealthCare.gov isn’t the only tech topic on which the Obama administration has some serious explaining to do.

The White House claims that it is supports copyright reform. It should be in favor of remaking the framework, because today’s copyright system is a mess: It grants protection that is too long (70 years or more), fair use is notoriously unclear and vague, and statutory damage laws create a massive deterrent to lawful creation. Economists and scholars argue that modern copyright, as opposed to constitutional copyright, greatly impedes innovation and content creation. But the TPP, which is being negotiated by 11 countries, would be a step in the completely wrong direction.

In its present state, treaty would expand copyright and effectively make real reform impossible. Worse, it would essentially disregard constitutional limitations on copyright and reject pillars like fair use, the first-sale doctrine, and having copyright be for “limited times.” The worst part: While the White House was publicly proclaiming its support of cellphone unlocking, it was secretly negotiating a treaty that would ban it.

Cellphone unlocking is the ability to take a phone and alter its settings so that it can be used on other carriers. Essentially this technology allows a consumer to bring her phone from one carrier to another when her contract expires (if technologies are compatible). In January, following appeals by AT&T/Verizon’s main trade association, the Librarian of Congress issued a ruling making unlocking a felony punishable by five years in prison and a $500,000 fine. This was a terrible idea: Economists and market participants have explained that this ruling would result in reduced competition in the industry, a decimated resale market, and restricted consumer rights. And indeed the impact has been devastating.

At the time, I spearheaded an unpaid national campaign to legalize unlocking, which included a White House “We the People” petition (I wrote a bit about our campaign here). Our petition reached 114,000 signatures, and the White House responded in favor of cellphone unlocking:

“The White House agrees with the 114,000+ of you who believe that consumers should be able to unlock their cell phones. ... It's common sense, crucial for protecting consumer choice, and important for ensuring we continue to have the vibrant, competitive wireless market that delivers innovative products and solid service to meet consumers' needs.”

The FCC came out in favor of our petition, as did numerous outside groups such as Freedomworks, Public Knowledge, R Street and the editorial boards of the New York Times and the Washington Examiner. We were unable to find a single group, or Member of Congress, that was in favor of unlocking being a felony. But somehow, while a number of bills were introduced, none passed, and the one that had widespread support, H.R. 1892, never received a hearing or was brought up for a vote.

The leaked treaty draft shows that while the White House was championing restoring free market principles to phones, the U.S. proposed that the TPP lock in the process that allowed the Librarian of Congress to rule this technology as illegal through international law. This would make potential reforms like H.R. 1892 impossible.* make unlocking even more illegal—potentially even making it a more serious crime. It should be noted that Canada did submit an amendment proposal that could allow unlocking, but neither the United States nor any other country supported it.

But the TPP draft doesn’t stop there. It would ban numerous other technologies that have beneficial uses. In particular, the legislation would ensure that jailbreaking—which is installing a different operating system on your phone, tablet, or e-reader—is illegal. It’s already on precarious ground in the United States, but under TPP it would be illegal in all circumstances. What type of nation would arrest 23 million people for installing a different operating system on their own device?

This treaty is still being negotiated, so all of these issues could be addressed in the final text, but so far what has been made public demonstrates a massive and nearly unprecedented power grab by special interests rather than sound public policy considerations.

This treaty has long been shrouded in unprecedented secrecy. Congressional staff, press and general public weren’t allowed to read it; in many cases, even members of Congress were kept in the dark. Meanwhile, special interests were given full access. Now we know why: The White House didn’t want the public to know what was being negotiated in their name.

Correction, Nov 18, 2013: This blog post originally misstated effect of the U.S. proposal to TPP.

Words. They have their uses. When you're trying to communicate something, but for some reason you don't have the relevant photos or videos handy, words can be just the ticket. Certainly they represented a marked improvement over grunts and roars when they were first introduced. They've had a good run.

But let's face it: Words are old hat. They're suboptimal from a UX perspective, because they're aesthetically monotonous, and their meaning is not always easily grasped at first glance. They demand of the user a form of engagement—"reading"—that can be mentally daunting. Some people can't do it at all. And even those who can are likely to find their attention straying if confronted with too many words at once. After a while, if they're not careful, the words all sort of run together and the eye begins to yearn year for something more instantly gratifying upon which to alight.

Be honest: You've already skipped on down to the pictures below these words, haven't you? And if you haven't, you're sorely tempted to do so at this very moment.

Instagram gets it. Pinterest gets it. Snapchat and Tumblr get it. Even Facebook sort of gets it.

And, as of today, Twitter gets it too—finally. The last of the text-centric social networks, Twitter until approximately 4 p.m. eastern time on Tuesday, Monday, Oct. 29, 2013, stood as a vestigial bastion of logophilia on the social Web, clinging to 140-character text strings as its base unit of sharing. You could link to pictures and videos in tweets, but they remained subordinate to the words, concealed behind a link or a "view photo" button. No longer. In an update rolled out quietly, almost apologetically, Twitter this afternoon began displaying pictures and Vine videos directly in tweets in users' timelines. Now the words are subordinate: The images are far larger than text-based tweets,

I could use more words to describe this change, but why waste your time when I can show you instead?

The innovation has of course dismayed a smattering of loyal Twitter power-users. That's to be expected. But as Twitter reviews its balance sheet in preparation for going public, it surely can't help but notice that it possesses a surplus of loyal power-users and a deficit of casual visitors. That deficit is what has prevented the site from approaching Facebook-level ubiquity, not to mention Facebook-level riches. And it's all words' fault.

Join me, then, in cheering the dawn of a new Twitter—a Twitter in which words are relegated to their rightful role as captions. It's a Twitter for the latter 90 percent of the 21st Century, and for the lazier 90 percent of society. Or, perhaps more accurately, for the 90 percent of society that's willing to own up to its laziness.

Or, perhaps more accurately still, for the people who account for nearly 90 percent of Twitter's revenue: