William Eggleston on a swing in Jackson, Missippi

A recent TOP post on Eggleston photo sometimes titled "Jackson, Missippi" coming up for auction at Christie's. This is perhaps one Eggleston's more famous pictures.

http://theonlinephotographer.typepad.com/the_online_photographer/2012/02/eggl...

"Jackson, Missippi" shows an old woman wearing a brightly colored polyester print dress on a swinging couch with an equally garish upholstery.

But you probably haven't seen this photo before

I'd hadn't seen this Eggleston shot before. And Google image searches suggest it's not on the Internet so I'm not sure how many others have seen it.
The image is from a screen capture from the "William Eggleston in the Real World" a documentary released in 2005 on available on DVD. It's in the titles at the end of movie.

That dapper young man on the swing is (a young) Eggleston. It gives a sense of how long ago the "old woman on a swing" photo was taken (given that Eggleston is seventy-ish). It's an image out of time.

The two photos are shot close together in time but not at the same time: check out the leaf litter in front of the couch (certainly the same year and season) and the sun striking the log in the background (it's different in the two shots -- the sun moved). The camera is slightly lower in the Eggleston shot and is perhaps shot with a longer focal length or camera further away? The complicating factor is the video is 4:3 so the picture is cropped. I think the left and right edges are cropped (assuming 3:2 original) so perhaps it's the same focal length but further away from the couch. The other obvious point is the saturation of the print isn't pushed as hard: the upholstery looks faded and worn "self-portrait".

I don't think I've seen a self-portrait of Eggleston before either. Neither has Weski

http://www.egglestontrust.com/hasselblad_weski.html

> In the only self-portrait of William Eggleston known to me, the photographer--only partially visible--is looking at an old fan, its shiny brass propeller corresponding with the warm yellow of a plastic bottle before him. Standing in front of an old, white enamel Coca-Cola sign, the fan harks back, through its antiquated form, to bygone days, but without any hint of sentimentality.

So that's not this one ...

BTW, the doc isn't as good as the BBC Imagine one. A bit too much aping Eggleston's photo and video style (the mundane; the fragmentary shot; the odd angles; getting too close with the video camera). The map isn't the territory. But still interesting and does cover some different areas of Eggleston's life "The Colorful Mr Eggleston".

Fujifilm X-Trans color filter array layout has interesting symmetries

Fujifilm have revealed the X-Trans color filter array layout used on the X-Pro1 camera.

http://www.fujifilm.com/products/digital_cameras/x/fujifilm_x_pro1/features/

Two things come to mind:
1. The reds, greens and blues aren't quite balanced: there is a small excess of green. Not a big issue and perhaps a small advantage if it helps luminance contrast (acutance).

Of the 36 pixels in each 6x6 block 20 are green, 8 are blue and 8 are red. In a Bayer array 6x6 block there would be 18 green, 9 blue and 9 red. Green is where most of the luminance info (and contrast detail)

2. The array has an interesting set of symmetries: the 6x6 block is composed of four 3x3 blocks (of 2R5G2B) with the same layout which are related to each other by 90 degree clockwise rotations as you go around the four blocks in a clockwise order.
This symmetry drives the (small) imbalance in the green which appears in a center 2x2 block of green pels.

Aside from the central 2x2 block all the other 2x2 blocks in the 6x6 array have set of RGGB so they can sample color info with about half the resolution of the array (as you would expect).
When you tile the array you can see the "all green" 2x2 block repeats every 3 pels in both vertical and horizontal directions. The outside of this block is decorated with repeating GRB pels as you go around the "all green" block.

I suspect those 2x2 blocks of green pels improve the luminance sampling rate (reducing aliasing issues) or are possibly also useful (in pixel binning) for CD AF or scene brightness measurement?

All of this must come with a new demosaicing algorithm that makes the best use of the data provided by this "odd" CFA.

Of course Fuji gloss of the other aliasing issues (other than color moire) that come from removing the optical low pass filter but this layout may help more than just color moire. It won't be perfect (you can't remove alias artifact after the fact) but it may be "good enough".

Choosing a misleading favicon ...

Browsing Snapsort (looking for the "ideal" camera for Veleta -- it's an interesting site with a nice browser/comparison UI) I clicked on a group of possible comparisons.

And waited for them to load.

And waited.
And waited.

And waited.

Then I noticed they had already loaded but the favicon (this sort of fading shutter thing) looks so like the "I'm loading" spinning rosette that Safari puts in the same location that my visual system had grouped the two together.

Interesting bug.

Snow White and the Apple in Vivace, Seattle

Firefox 4 and TenFourFox: web apps and restarting the app

I finally turned off "Show my windows and tabs from last time" and instead used "Restore Previous Session" to bring back the previous session.

I quit and restarted TenFourFox to recover memory (dontcha love that!).

And I noticed something that I haven't seen spelt out anywhere when TenFourFox restarted: the previous Web App tabs all reappeared (as app tabs) even though you're not asking it to remember any session. Interesting.

Though what seems even more odd, on TenFourFox at least, is the web apps restored to a "not current" version of the web page. The page that came up was from some time yesterday (not even the one I had at quit time for the app). Weird.

That was true for all 5 web app tabs: they must be caching and reshowing old data rather than going to the web (supporting for locally cached HTML5 style web apps that you can use off line?)

I guess Firefox 4 is a bit like Perl: there's more than one way to do it. And some of them are wrong or at least unexpected.

Why did the Barred Owl's range expand into the West?

I asked earlier on TWEETERS: Why did the Barred Owl's range expand? Anthropogenic or natural causes?
Anthropgenic causes seem to get the nod in the literature (as Rob has pointed out) for their expansion across the Plains. But there is no evidence for or against logging as being involved in the Barred Owl range expansion.

Two of Kent Livesey's papers deal with this in a detailed way looking at the chronological progression of the Barred Owl's range in "Range Expansion of Barred Owls, Part I: Chronology and Distribution" (Livezey 2009a.pdf) and "Range Expansion of Barred Owls, Part II: Facilitating Ecological Changes" (Livezey 2009b.pdf). They're PDF files zipped up into an archive (without a password) on rapidshare.

http://rapidshare.com/files/447227260/Kent_Livezey_-_Spotted_Owl_papers.zip

Two quote:

Elizabeth G. Kelly 2001 thesis for an MS in Wildlife Science at OSU ("Range Expansion of the Northern Barred Owl: an Evaluation of the Impact on Spotted Owls") has a nice intro and literature survey on the problem. It's a good intro to the issue.

The intro discusses the literature on hypotheses on why the Barred Owl's range expanded.

Quote:

> Historically, the range of the barred owl was limited to the eastern United States and eastern Canada (Bent 1938). In the early 1900s the range of the barred owl gradually expanded westward across Canada to British Columbia, then north into southeast Alaska and south into western Montana, Idaho, and Washington (Grant 1966, Campbell 1973, Reichard 1974, Shea 1974, Boxall and Stepney 1982, American Ornithologists' Union 1983, Sharp 1989, Dunbar et al. 1991, Wright and Hayward 1998). Barred owls first appeared in Oregon in 1974 (Taylor and Forsman 1976) and California in 1981 (Dark et al. 1998).>> Avifaunal range expansions in North America are common and have been attributed to anthropogenic or natural causes, or both (DeSante and George 1994, Johnson 1994, Root and Weckstein 1994). Woodlands that developed on the Great Plains during glacial periods became grasslands during warmer interglacial periods, thereby creating a barrier to the east-west movement of forest-dwelling species (Knopf 1986). More recently, changes in climate and natural forest cover, as well as tree planting and fire suppression, have created forest cover in grasslands and prairies that may have facilitated range expansions of forest birds (Ehrlich et al. 1988, Johnson 1994).>> Hypotheses that have been proposed to explain the range expansion of the barred owl across Canada include changes in climate (Johnson 1994), tree planting and development of riparian forest (Dark et al. 1998) or increased adaptation to coniferous forests (Boxall and Stepney 1982). Range expansions of other North American owls during the same period include the apparent expansion of the boreal owl (Aegoliusfunereus) from Canada southward to Colorado and northern New Mexico (Johnson 1994) and movements of barn owls (Tyto alba) into parts of British Columbia and California (DeSante and George 1994). In Europe, the tawny owl (Strix aluco) was first documented in Finland in 1875 and has since expanded its range northward, becoming common in southern Finland (Mikkola 1983).>> Although some have suggested that the range expansion of the barred owl into the Pacific Northwest was facilitated by forest management practices (Hamer 1988, Root and Weckstein 1994, Dark et al. 1998, Konig et al. 1999), there is no data to support or refute this hypothesis. In fact, the range expansion may have occurred regardless of forest management activities (Johnson 1994). Barred owls appear to be habitat generalists that can occupy a broad range of forest conditions, from highly fragmented forests in managed landscapes to pristine forests in Wilderness Areas (Shea 1974, Hamer 1988, Dunbar et al. 1991, Wright and Hayward 1998). In a study of habitat use by barred owls in the North Cascades of Washington, Hamer et al. (1989) found that although some individual barred owls selected young forests, the majority used old-growth, mature and young forests in proportion to availability. This suggested that in general, barred owls did not select any particular forest age class among those that were available to them. Thus, it is by no means clear that the range expansion of the barred owl has been facilitated by forest management activities. In any event, to state unequivocally that "Logging actually favours the expansion of this species... " (Konig et al. 1999:328) without mention of other equally probable hypotheses is misleading and possibly erroneous.
You can find the whole of her thesis here (including the references) in a PDF

http://ir.library.oregonstate.edu/dspace/bitstream/1957/9653/1/Kelly_Elizabet...
Or read it in the Google PDF viewer online

http://docs.google.com/viewer?url=http%3A%2F%2Fir.library.oregonstate.edu%2Fd...

BNA has some notes on historical range change (written in 2000). See the link for details and references.

http://bna.birds.cornell.edu/bna/species/508/articles/distribution
http://bna.birds.cornell.edu.ezproxy.spl.org:2048/bna/species/508/articles/di...

> Historical Changes
>> During twentieth century, range extended across boreal forest of Canada to montane forests of w. Canada and the w. U.S. (Grant 1966, Taylor and Forsman 1976, Sharp 1989). First record for s. Manitoba riparian forests in 1886 (Seton 1886). First recorded in Saskatchewan boreal forest 1948, and first nest 1961 (Houston 1959, 1961). Earliest record for Alberta, 1932 from the boreal forest (Preble 1941); first nest 1966 (Jones 1966). First record for British Columbia 1943 at Liard Crossing; first nest 1946; has spread south and west to coast (Grant 1966). By 1965, observed in ne. Washington; breeding by 1974 (Rogers 1966, Smith et al. 1997). Spread west across Cascade Mtns. by 1973 (Sharp 1989). First record for Idaho 1968 (Stephens and Sturts 1997) and Oregon 1974 (Gilligan et al. 1994). Expanded to nw. California 1981 (Evens and LeValley 1982), first breeding 1991 (Dark et al. 1998). Early record from Colorado, nest observed 1897 (Aiken 1900), and Wyoming, Crook Co. 1905 (Cooke 1909). Mechanism that possibly facilitated westward expansion not clear; explanations are anthropogenic in nature: fire suppression in boreal forest, increasing age of forest and size of trees (for nests), and establishment of riparian forests and planting of shelterbelts in n. Great Plains.
>> Populations and range reduced locally in e. Canada and U.S. due to forest clearing for agriculture over last 200 yr (Cadman et al. 1987, Erskine 1992, Jackson 1996, Nicholson 1997).

Combine journalism with the technical report format to make engineers write concisely with sentences?

Tools for Thinking: What scientific concept would improve everybody’s cognitive toolkit?

David Brooks at the New York TImes (no link as it's now paywalled) says:

A few months ago, Steven Pinker of Harvard asked a smart question: What scientific concept would improve everybody’s cognitive toolkit? 

The good folks at Edge.org organized a symposium, and 164 thinkers contributed suggestions. 

http://www.edge.org/q2011/q11_1.html#paulos 

The results of the symposium are at http://www.edge.org/q2011/q11_index.html

Scroll down after the newspaper/blogger blurbs and you get (bite sized) contributions from each of the smart folks that start here

http://www.edge.org/q2011/q11_1.html

Some of these ideas just point you in the right direction (e.g. Paulos' blurb on probability distributions doesn't tell you too much ... so use Wikipedia and Google to search for more.)

If you like that you might also like: http://en.wikipedia.org/wiki/List_of_cognitive_biases

My favorite heuristic: if you are having a difficult time making a choice between several options just pick one at random. The reason you're having difficulty choosing is there is (obviously) little difference between the choices so a random choice is as good as any other choice. If you can't stick with a random choice is there a hidden reason not to? What is it? And how does that inform your revised choice?

I presume another Brockman book will be on the way based on this: "What scientific concept would improve everybody’s cognitive toolkit?" or something similar

[Tweeters] Bird Abbreviations

On Mar 13, 2011, at 8:44 PM, Bill Anderson wrote:

> I have no problem with abbreviations as long as the bird is first fully named at the beginning of the e-mail. Agreed.

> I have seen the term SORA used quite often and still don't know what it stands for.> Bill Anderson; Edmonds, WA.

It's either a cryptic rail :-)
Or the Searchable Ornithological Research Archive
http://elibrary.unm.edu/sora/

For example, it's the place where you can find the original 1978 suggestion in a paper by Chandler Robbins for a standardized 4 letter species abbreviations for bird banders.

http://elibrary.unm.edu/sora/NABB/v006n02/p0046-p0047.pdf
http://elibrary.unm.edu/sora/NABB/v003n01/p0016-p0025.pdf

The official list is kept at US Bird Banding Laboratory.

http://www.pwrc.usgs.gov/bbl/manual/aspeclst.htm

This has been critiqued by quite a few people, especially by Shipman, as the four letter codes are not unique, easily confused by the casual user (and the occasional expert).

http://infohost.nmt.edu/~shipman/z/nom/bblcrit.html
http://infohost.nmt.edu/~shipman/z/nom/homepage.html

The four letter part of the spec came from the preexisting standard banding field recording forms. From an information design point of view it is a poor design. There are more than 100 "collisions" in the US BBL code when applying the rules to a name gives the same code for two or even three different bird species. A lot of the collision codes are used as real codes so it's easy to make mistakes (pick the incorrect code for the bird you are observing that matches another bird) and difficult to catch or correct those errors.
Perhaps worse still the BBL codes are incomplete as they don't include game birds. So there aren't 4 letter alpha codes for all the birds on the AOU list which leads to other potential collisions.

Other 4 letter alpha systems have been proposed and used. Curiously, the Breeding Bird survey has different one from the BBL (collisions are resolved differently)
http://infohost.nmt.edu/~shipman/z/nom/bbs.html

And the "Birder's code" was suggested in Birding in 1992 is not much different from BBL codes.

This was standardized for banders by Pyle and DeSante in 2003. The Institute for Bird Populations maintains an updated list of 4 letter alpha codes for common names and a 6 letter alpha codes for binomial names based on the current AOU species list using Pyle and DeSante's updates.

http://www.birdpop.org/alphacodes.htm

If you are going to use a 4 letter alpha code it should be from this list so everyone will understand what it means.

There is an easier alternative for non-banders that avoids these problems that was suggested by Shipman: use a 6 letter alpha code.
http://infohost.nmt.edu/~shipman/z/nom/6home.html

Bowman regularized that original idea and maintains a list of Six-letter Code for ABA Bird Checklist Species

http://www-personal.umich.edu/~bbowman/birds/sixlettercode.html

I'd encourage people, at the risk of seeming less cool, to use Bowman's 6 letter code. See the end of the email for more detail. There are four easy to remember rules which generate only 9 collisions on the AOU list. You just have to remember the nine exceptions. None of the collision codes are "legal" so if you use an incorrect code in the field you can tell when it is incorrect. Very useful for field use and for use when abbreviated field notes are exchanged by people. It also permits Shipman-like extensions that are not included in the 4 letter codes e.g. an unknown buteo, "buteo spp", would be BUTEO.
This scheme is also extensible to bird fauna lists in other parts of the world: just follow the rules and define the standard alternatives for collisions against the standard checklist for that region.

Birdbanders are cool after all and if you have the brainpower to accurately remember all the Pyle and DeSante alpha codes then one can use them. But as the new Crossley ID book, which uses 4 letter "alpha" codes extensively through the guide (to save space?) it is difficult to get them all correct even for as good a birder and author as Crossley.

It turns out Crossley gets a couple wrong in the guide that passed through a few minds before publication. If he can't get them right ...

http://www.crossleybooks.com/comments-corrections/

> p.464 The alpha code for Sharp-tailed Sparrow is SSTS and not STSP. Sharp-tailed Sparrow is now called Saltmarsh Sparrow.
>> p 462 Grasshopper Sparrow – abbreviation LESP near end of text should be LCSP (Le Conte’s Sparrow)

I really encourage people to think about using the Bowman 6 letter code. It works well. And take little effort to learn it.

Bowman's Six-letter Code for ABA Bird Checklist Species
-------------------------------------------------------

STANDARD ABBREVIATIONS
"Words" in a species name are those parts of the name separated by spaces or hyphens. Ex.: "Bay-breasted Warbler" has three words.

¤ ONE-WORD NAMES Use the first six letters, or use the entire name if it is less than six letters in length. Ex.: Canvasback = CANVAS, Sora = SORA

¤ TWO-WORD NAMES Use the first three letters of each word. Ex.: Pectoral Sandpiper = PECSAN, Fish Crow = FISCRO

¤ THREE-WORD NAMES Use the first two letters of each word. Ex.: Great Gray Owl = GRGROW, Broad-tailed Hummingbird = BRTAHU

¤ FOUR-WORD NAMES Use the first letter of each of the first two words and the first two letters of each of the last two words. Ex. Black-and-white Warbler = BAWHWA, Black-crowned Night-Heron = BCNIHE

Because nine instances of duplicate abbreviations result when the above rules are applied to the 969 species names in the American Birding Association (ABA) Checklist for the Birds of North America, special-case abbreviations are necessary for the 18 species listed below:
SPECIAL CASE ABBREVIATIONS

Barn Owl BARNOWBarred Owl BARROW
Leach's Storm-Petrel LEACSP
Least Storm-Petrel LEASSP
Blackburnian Warbler BLBUWABlackpoll Warbler BLPOWAPalm Warbler PALMWAPallas's Warbler PALLWABlack-throated Gray Warbler BTGYWABlack-throated Green Warbler BTGNWAWilson's Warbler WILSWAWillow Warbler WILLWACommon Redpoll COREDPCommon Redshank COREDSYellow-breasted Bunting YBREBUYellow-browed Bunting YBROBUGreen-breasted Mango GNBRMAGray-breasted Martin GYBRMA

Note that in each case of duplicate codes the duplicate code is used for neither of the special-case species. For example, BAROWL is used for neither Barn Owl nor Barred Owl. This lessens confusion and decreases errors.
http://www-personal.umich.edu/~bbowman/birds/sixlettercode.html

[Tweeters] Interesting Coyote Tale

On Mar 12, 2011, at 7:08 PM, Casacummins wrote on TWEETERS:

On Wednesday, March I was driving on Highway 104, North of the Hood Canal Bridge at about 3:00 pm when a larger-than-usual coyote crossed the road about 100 yards ahead, headed Southeast. He was rangy and tall, but much darker colored than the coyotes I’ve seen before. The most striking feature however was the large, snow-white tip of the tail, The white tail is what made me notice, and for a moment I thought I had seen a fox. As I passed, he paused on the road embankment and looked toward the highway. Dark grayish-brown and the facial features of a coyote – somewhat thin face and long muzzle, but somewhat large. Then it was into the forest and gone. Has anyone seen a color phase like this one on a coyote, or could it have been a coyote/dog cross?

Nice observation.

There are melanistic coyotes out there. Including those with white tipped tails. Like this one:

It does look more "rangy" and "dog-like" than my mental version of a "normal" coyote but I think that's more in my mind than in reality: it's colliding with mental images of black dogs.

The white tip is interesting. Non-melanistic coyotes typically have a black tip Does one ever see white tips in non-melanistic coyotes? Yes, but perhaps it isn't as obvious as in the black coyote.

http://www.desertusa.com/june96/du_cycot.html

Is it a cross with a domestic dog?

Probably not recently but perhaps at some time in the past it was. Or may be not. The technical story is interesting. There is some argument that they get their black color from cross breeding with dogs rather than a straight mutation of MC1R gene.

The mechanism in coyotes (and wolves) appears to have a different mechanism: a specific mutation (at the K locus) on another gene CBD103 but it acts on the same MC1R receptor. Most of the time coyotes are (two tone -- "sable" in dog speak) phaeomelanin colored (tan/brown) but if they get a MC1R receptor turned on they color their pelage with eumelanin and appear (single tone) black.

The papers linked below claim the melanistic K locus mutation in North American wolves derives from past hybridization with domestic dogs. It also has high frequency in forested habitats and looks like it underwent positive selection for wolves and coyotes in "dark" environments.

http://www.sciencemag.org/content/323/5919/1339.abstract
http://www.canislupus.it/public/AndersonScienceExpress.pdf

Or the quick version in a blog

http://evolutionarynovelty.blogspot.com/2009/06/whos-afraid-of-big-black-wolf...