Snow crystals and stereoscopic vision.

I haven't given up on this blog, but I'm going through a somewhat difficult time and my mind is elsewhere. It's thus on hold for an underdetermined period.

Nevertheless, I wanted to post something special for the Holiday season with these stereo images taken with an electronic microscope by the Beltsville Agricultural Research Center.

I'm copying here one image and I'm sending you to their website for the other ones.

You must first cross your eyes, relax your vision until there are three images, then focus on the center image (that will appear when you cross your eyes). Don't stand too close from your screen (about 60 cm). You can also enlarge the image by clicking on it.

source: BARC-USDA

The entire image set can be found here.

I'm skiving off, as there are some many fascinating things to tell about crystals, stereo vision, and microscopy...

Happy Holidays and I'm hoping to come back in more favorable conditions in 2011!

Women matter.

A few days ago, I read (in french) about a study showing that companies having more women  (over 35%) in leading positions were more successful and were also doing better facing the current economical crisis. This will be published in April in the french Journal "CNRS¹ Travail, genre et sociétés." This phenomenon has also been studied by the McKinsey company, whose website reports their full study.

1. French Center for Scientific Research

The knowledge-ignorance paradox

I thought this moment would never come, but I finally passed all the steps and stages required and I've been slowy getting used to putting a Dr. in front of my name. The feeling of freedom wasn't as good as I expected: A new degree means that job hunting is around the corner and I've been asking myself many questions such as Who am I? What's my future? What do I want to do with my life? How I am going to get a job in this unfavorable economy?

I've been proceeding through elimination (which seems to be a chemistry term, but it's not in French, excuse the sketchy translation). I do not want to keep working in academia: I'm more of an engineer than a researcher. I do not want to work alone all the time or in front of a screen 50+ hours a week - not counting Internet time. I've also come to realize a paradox of the doctorate: on one hand, one's becoming extra specialized on a specific case of a sub-branch of their domain, while on the other hand, the time spent working on the specific case of a sub-branch of the domain makes having a general scientific background much harder. By that, I mean being more or less up to date on the important current topics of science. For instance, I've heard on the radio that the CERN  had some trouble with the LHC, but I'm actually more counting on The Big Bang Theory rather than Science orNature to tell me if scientists  managed to find Higgs bosons.

The specific knowledge takes over a more generalist one.

Culture is what's is left, after you've forgotten everything.

Édouard Herriot

My scientific culture was never as good than when I just graduated from my physics studies. But after all those years, is there much left besides an ability to understand and process information faster?

The most frustrating part is that I've never been surrounded by so many brilliant people than during the past few years. If I take some distance, I also realize that teaching, despite being the last thing you want to do when you have to do it, allows having a real knowledge of what you're actually researching on.

So, my program for this spring is to go back to the reading list I abandoned a while back:

• The Undercover Scientist, by Peter J. Bentley
• Des atomes dans mon café crème, by Pablo Jensen
• La physique de tous les jours, by Istvàn Berkes
• Histoire mondiale des sciences, by Colin Ronan
• Surely you're joking, Mr. Feynman, by Richard P. Feynman
• Big Bang, by Simon Singh
• Lumière et Matière, by Richard P. Feyman
• Lectures on Physics, by Richard P. Feynman

That is, a few thousands pages.

Or maybe I will just start reading Science & Vie - a good vulgarization magazine.

Plus, if I don't start reading soon, I won't have anything to blog.

Holidays fun.

Via Richard Wiseman.

Poetic science.

While I'm at it, I'd like to share two things today.

First, the world's tiniest snowman measures 0.01 mm and even has a nose!

This isn't snow though. He's made of two tiny tin beads bounds by some platinum. The face features have been designed with a focus ions beam. The link is here.

The second thing I'd like to share is a video showing droplets falling, but imaged at a rate of 2000 images/second...

I'm must confess that I don't remember much from my surface physics class, except maybe that interfaces require more energy than bulk to be created and that this is why nature tends to reduce the surfaces extent by creating spheres. I would have been more interested by this class if it had started by such a video.

Back.

No, I haven't given up on this blog. Writing my thesis and the numerous reading-correction-rewriting cycles became strenuous and obsessive. The only way not to completely loose my mind - and my friends - was to have a life as normal as possible during the rare moments of rest. This required not to blog on anything about my thesis. Or on anything closely - or loosely - related to science, technology, computers, libraries, books, paper, pens, etc. And during that period, I had absolutely no desire to read anything scientific.

So, here I am after a couple of months. I'm most likely talking in the big internet emptiness, as my rare posts haven't opened the doors of google for me - yet. But I do want to keep writing, even though this may be too time consuming to be done regularly right now.

I submitted my dissertation about a month ago, which means that I'm only a few days away from having to defend my work. My presentation's ready, my slides are pretty. My feelings about the present situation are mixed. I do know the people in my jury - at different degrees - and they have already submitted their report. Which means that the outcome is more or less already defined. Moreover, a thesis director who does their job doesn't let their student submit a work, whose quality is not sufficient. But still, here I am.  Nervous and full of questions. Nothing's never sure, even though the most likely question is how much correction I'll have to bring to the final version of my dissertation.

And once this is behind me, so many questions will arise: where do I want my career to go? Will I  go through unemployment? Will I find a job in Switzerland or will I have to move abroad? Do I want to move abroad? I'm trying not to think too much about this. Yet.

A few more days.

Let's rehabilitate HDR

This is not HDR.



Seriously? Engineers do their best so the user can have nice images of sunsets or photographing grandma while facing the sun and here comes a group of photoshoppers ruining everything.

I think it's time to set the record straight and explain what is actually HDR - high dynamic range.

An image is generally encoded using 8 bits per channel. 8 bits = 2⁸ = 256 levels. A black and white image can thus be encoded using 256 grey levels. Is that a lot? Not really. Really not. The ratio between the lightest point - the sun - and the darkest point - no, not grandma - may be around 10,000 to 1. That's precisely what's called dynamic range. That means that it's not possible to simultaneously image details in shadows and highlights, which will be most likely "burnt."

Good news is that camera sensors can capture more that 256 light levels. Typically, in raw format, an image has 12 bits per channel, i.e. a little over 4000 grey levels. Getting better. But the problem of transforming a 4000 levels image into a standard 256 levels image remains. This must me done is a smart way, keeping details in both the dark and light parts of the images and a natural appearance by avoiding the artifacts apparently cherished in "artistic" HDR - see image. There's actually an in-camera tone-mapping allowing the user to display a decent looking 8 bits images in her/his screen without any additional work.

There's also another way of doing HDR. Take a series of pictures with different exposures. Typically, an underexposed image containing details in highlights, a correctly exposed image containing details in midtones, and an underexposed image containing details in the shadows. Then merge the images in a way that preserves a maximum of details and get a nice image. Easier said than done... One must work with linear images, correctly merge the image, and applied a "good" tone mapping...

Artistic HDR is probably much simpler....

For those who are interested, the serious stuff's here.