Maintenance, November 2009

- Today was a beautiful time to work on the Frankenbike, it being Veterans' Day and sunny. Unfortunately, i spent the better part of the afternoon scouring the house for her key. Turns out it was still in the bike--not the ignition, which i had checked earlier, but the seat cover latch. Fantastic. Nonetheless, i now have a drill and a new clutch cable, meaning i can finally get into the crankcase to replace the damn thing.

- The folks let me know that Christmas (my 2007 Camry), and most Toyotas produced from 2004 onwards, have an unresolved "issue" that sometimes results in a stuck accelerator. Ruh-roh. This made news when an off-duty CHP and his family were killed due to a stuck accelerator in their Lexus. This raises the question of why a police officer didn't take the obvious steps to resolve the issue (shift to neutral and kill the engine), but not being in the car at the time it's hard to judge.

- I bought the Windows 7 upgrade at the student rate ($30) and installed it successfully...after eighteen hours of wrangling. Microsoft and Digital River really botched this one. I quite literally had issues at every single step of the install process, and ultimately had to resort to the good ol' double install trick* to get the damn thing working.

That being said, i'm incredibly happy with W7. It's a much slimmer OS than Vista, and Microsoft did a great job ironing out most of the kinks and inconsistencies of both Vista and XP. I also can run in XP mode for the ancient programs geology demands i use. Lastly, i'm up to 64-bit architecture now, and the processing boost is really, really great.

* If you do want to upgrade from 32-bit Windows Vista/XP to 64-bit Windows 7, Microsoft requires a clean install of a full-version of Windows 7, not the upgrade. This means an upgrade key won't work. There's a workaround for this, found here, that will let you use an upgrade key to do a clean install. This means anyone buying Windows 7 can get a full installer for the price of an upgrade--pretty sweet! Read More......

Lost in Color Space

Color space conversion is annoying the hell out of me. Take a look at this picture, taken from the first Portland post. This was taken in Adobe RGB color space, but like all web images, was improperly converted to sRGB (Standard RGB) color space during upload.

This is the same image, except using Adobe's Adobe RGB to sRGB converter, instead of Google's direct translation. Notice the more vivid reds.

Shorter Scott: the colors are different and i'm not happy. A technical discussion as to why they are different follows, and if you process images with Photoshop it may behoove you to read on. Monitors typically define color as a function of three channels, or base colors: red, green, and blue. By varying the "colorfulness" (difference from gray) of each channel, you can simultaneously vary the hue, saturation, and brightness of your particular color. If you quantify the channel variations, the colors can actually be mapped to Cartesian (x, y, z) coordinates--they can be represented as a volume. When calibrated to a known color reference point, this color map becomes absolute--a true color space.

Each channel has 256 layers, so 256 x 256 x 256 = 16.7 million possible colors. Bitchin!

sRGB is by far the most common color space, mainly because our color photoreceptors (cone cells) are most sensitive to those three wavelengths. However, another tetrachromatic (4-channel) color space--CMYK--is preferred for print. These colors--Cyan, Maroon, Yellow, and Key black--correspond to ink used in most printers. CMYK color space encompasses a different area of color space, or gamut, than sRGB. This can result in inaccuracies while printing, and also means some colors possible in CMYK space cannot be represented in sRGB space.

Hence Adobe RGB--a color space designed to represent CMYK space using RGB colors. Speaking objectively, Adobe RGB represents an improvement upon sRGB because it can accurately portray a greater gamut than sRGB. Personally, i prefer Adobe RGB because it gives more vibrant reds and greens, which are important when shooting landscapes. It's also the default setting in Photoshop, so i was using Adobe RGB long before i knew there was a compatibility issue. A big one.

Here's the issue: both Adobe RGB and sRGB have 16.7 million possible colors, but Adobe RGB has a larger gamut. Therefore, it must represent a greater range of color using the same amount of colors. This means that the spacing between colors in Adobe RGB is greater than the spacing between colors in sRGB. Therefore, one color represented in Adobe RGB will have different color values of R, G, and B than the equivalent color in sRGB. Because of this mismatch, a red-155 in sRGB will be less vibrant than a red-155 in Adobe RGB.

With me so far? A particular RGB value in Adobe RGB is not the same color as the same value in sRGB. The sRGB color is less duller than the Adobe RGB color with the same value.

And now here's the problem: sRGB is the web standard--every photo site that i'm aware of, from Flickr to Facebook to deviantArt to Picasa, uses sRGB. When images are uploaded, the host will interpret Adobe RGB color values as though they were sRGB colors. This means that colors are changed during the upload process. The result? Images processed in Adobe RGB appear bland and flat, compared to sRGB images--even though Adobe should show a better range of colors.

Furthermore, because web browsers will automatically do this conversion, you simply cannot use Adobe RGB images for the web, without risking the loss of color information. Within Photoshop there is an option to accurately convert an Adobe RGB image into its sRGB equivalent, but this means all of my images to date have come out different than i've viewed them on my hard drive. Really this is just me being anal about my images, but really i just have this thing about when i'm not portrayed accurately [PDF].

You can read more about this problem here.
Read More......

The PJG

Portland has the best Japanese garden i've yet to see in the states. Situated in the hills west of town, Portland did a simply great job of isolating the park in its urban setting. Esteban and i took an afternoon off and went to investigate.

We were fortunate to visit Portland during their peak leaf-peeping time. The weather was surprisingly good; temperatures were balmy, the clouds strangely burned off by midafternoon, and it rained only once during the entire conference period. On this day Portland's perpetual cloudiness gave a great diffuse light--perfect for foliage.

One thing that differentiated PJG from other gardens is their noncommercial photo policy. I've agreed not to market any of these pictures, and it's easy to see why they're so anal about this--it's nearly impossible to take a bad picture in the place. This says quite a bit about the Japanese aesthetic, and even more about how that spartan ideal of beauty has influenced contemporary photography.

Also in the Washington Park area: the International Rose Test Garden and the Hoyt Arboretum, among other venues.

For those in Chicagoland / Beloit, Rockford has a surprisingly good garden, even if it is a bit noisy--the Anderson Japanese Gardens. Read More......

Portland

Three weeks ago i was in Portland, Oregon, basking in what most people agree is the city that best approximates Pinko Commie Heaven. Bush I dubbed it "Little Beirut," which makes little sense given the city's demographics. UC paid my way to Portland because i was presenting results from the Death Valley work; you can find an abstract here. The entire conference was at the Oregon Convention Center, which is capped by those two glowing spikes in the foreground. The skyline of Downtown Portland lies behind.

Geographically, Portland is divided by the Willamette River, with the downtown on the west bank. Someone compared it to Cincinnati, which i suppose works if you move all of downtown Cincinnati across the Ohio. Barring that, Cincinnati is contemplating putting in streetcars, using Portland as a model.

Portland's Pinko Commie status, as well as Cincinnati's streetcar plan, are based on Portland's simply stellar environmental record. There's pretty much consensus that Portland is the greenest city in the United States (1 2 3 4 5). Perhaps the most tangible aspect of Portland's green initiative is the seamless integration of their streetcar, bus, and light rail systems. Public transit within the downtown area is free, meaning i was constantly shuttling across the Willamette to catch lunch before the afternoon technical sessions. The area is also fully supported by Google Maps. Add Portland's bike ethos and you have a metropolitan area that was, while i was there, quieter and completely free of traffic snarls.

People often lose sight of the fact that green initiatives can benefit in ways that have nothing to do with being "green." Portland's just another example of the quality-of-life improvements that green initiatives can bring to a city.

On a completely separate note, Portland has an excellent Japanese Garden, which is where Esteban and i spent an afternoon after our sessions.

Can't forget the food carts. This place does an authentic Loco Moco! Read More......

Lapse Rate

Hi friends and family, i'm back from a week of geology in the Pacific Northwest. Presently i'm up to my nuts in grading and organizing pictures for the departmental photo contest, so for now you'll have to content yourself with this totally mediocre view of Mount Hood, "Oregon's most recently active volcano." Read More......

Pre-Chicago Nerd Pron

Pollster Strategic Vision LLC is in a bit of trouble after a failure of transparency in regards to their survey methods. Folks at 538.com took it to a new level when they found seemingly non-random non-uniform distributions of trailing digits (e.g. 53-47 McCain-Obama). In a recent post, they use Fourier analysis to determine the likelihood of the discrepancies to be the result of a systematic methodological error. The result:

The Strategic Vision result, on the other hand, or something more extreme, would occur by chance with probability only 0.00019. That’s not as low a p-value as the results obtained without filtering the non-uniform components, but it’s still very low -- less than one chance in 5000 to have occurred by chance alone.

In other words, statistics demonstrate that Strategic Vision has been cooking their books. "Nerdgasm," indeed.
* * *
On a slightly related note, Tom Levenson (director of the science writing grad program at MIT) pretty thoroughly dissects Megan McArdle's entire mode of thinking. McMegan inexplicably shares blogspace with TNC and Sully over at The Atlantic. She is known for blogging as "Jane Galt" and lying. Levenson establishes McArdle as a truly professional cherry picker of knowledge, not knowing--or not caring--enough to think critically about the data sources she uses. If you can get through it (4,000 some words), you get a great analysis of how the phrase "conservative think tank" (or liberal think tank, for that matter) is irrevocably oxymoronic.

Shorter version at BJ.
* * *
Chris Mooney (different Chris Mooney) has the latest debunking of climate myths. Unfortunately this is something we will never get enough of. 27%! Read More......

Represent, Michigan.

[US Senator John] Kyl: "I don't need maternity care" in my benefits package. [US Senator Debbie] Stabenow fires back: "I think your mom probably did."

(HT: Feministing) Read More......

Ticino

Ticino, my favorite Swiss canton, is featured in the New York Times' fall travel issue of Thtyle. Awesome! I did my undergraduate research in Ticino, which comprises most of the Italian-speaking part of Switzerland. It is without a doubt the best place in the world to do research. I got to sample lots of the local cuisine, which is predictably a hybridization of northern Italian and "mountain" food. This article brings back warm, satisfying memories of steaming plates of polenta and lots of wine.

Another thing the article nails is how freaking rainy a Ticinese summer can be. This raininess was the focus of my undergraduate research, which correlated the local precipitation high with more intense glaciation over the last 2 million years. Hooray for science and venison stew! Read More......

Census worker hanged, "fed" written on chest

Of course we shouldn't jump to conclusions. Southeastern Kentucky, especially the Public land, has a bit of a problem with drugs. Or, the guy could have secretly been an ACORN worker who killed himself to smear the Teabaggers. Or, he could be the first victim of a serial killer named Fred who also just happened to be a horrible speller, as well.

I definitely hope it's not related to Michele Bachmann and her census-concentration camp link. Even Glenn Beck backed off that bomb a while ago. And i hope it was just coincidence that the body was found on 9/12.

Damn. To think that i suggested this job to friends, and now this. Whatever happens, it's a pretty sorry state of affairs when a meth killing is the positive outcome of the story. Read More......

the R word

Maybe not so much "Racism," but "Reconstruction?"

I like to call the following maps the "How Race Matters" map series.

This is a map (metadata) that illustrates Presidential voting shifts by county between 2004 and 2008.

This is a map (metadata) that illustrates the percentage of African Americans, by census tract.

aaand this is the Hispanic population (metadata), percentage by census tract.

Watch carefully how vote shifts to the Republicans in the South eerily corresponds to the "less black" census areas. It appears that, for the South, there's a pretty strong negative correlation between vote shifts to Obama and % of African Americans. Most notably, see Arkansas, Louisiana, eastern Oklahoma, northeast Texas, and the Appalachian areas (northeast Mississippi, northern Alabama, all of Tennessee and the mountain regions of Kentucky/Virginia/West Va/Pennsylvania).

Hispanics are also notably absent from counties that shifted Republican, the exception being Florida, with its traditionally conservative Cuban voting base. It's interesting how vote shifts in Texas parallel the Hispanic population to a greater degree than the African American population.

The caveat here is that Indiana, North Dakota, and Montana all recorded major shifts towards Obama, yet they're arguably whiter than the rest of the South. Within the South, the race correlation isn't much seen in the Old South, mostly VA and NC, but partially SC, GA - although SC/GA may be due to a larger concentration of African Americans. The correlation is strongest in the Deep South (SC-GA-AL-MS-LA). It's almost completely absent everywhere else.

So: In the South, it's clear that large numbers of black people voted for Obama, and equally large numbers of white people voted for McCain. Elsewhere this dual relationship just doesn't seem to exist. Race definitely plays a part, although maybe in an oblique way: Southern politics still smarts at the effects of Reconstruction and the Northern betrayal, which stems from the Civil War, which stems from race. A lot of these voters were alive when segregation was still active, and undoubtedly that experience colors a lot of peoples' political thoughts. Context, it appears, is key. Read More......