(I did not make it to
misia's talk, unfortunately,
but I had to go back to Charles Village this evening anyhow.)
There is boring architecture in Baltimore, but most of what counts as boring is only dull in comparison to more fanciful examples in other parts of the city. A lot of this town is an entertaining assortment of styles and variations, even for the basic -- and overwhelmingly dominant here -- row house. While SoWeBo tends to have flat-front rowhouses with small details, it seems the farther north one travels the more the front walls of the houses bulge forth and shrink back, with curves and angles breaking up the line of the sidewalk's edge even where there are no porches. Along North Avenue, there are steep conical peaks on the round protrusions on the fronts of the rowhouses.
Tonight was one of those orange sunsets with lots of that "peeking through a slit in dark clouds" effect going on, and a healthy dose of "reflect off the nearby clouds and make them look shiny, not just pink". And I was driving west along North Avenue, towards that fairytale movie sunset, which silhouetted those steep conical peaks. And in the shadow cast by the further ones, distances became kind of random and the rest of the roofline became somewhat arbitrary, and really one might as well have been looking at the top of a storybook castle except for the lack of pennants streaming in the breeze. Several blocks of North Avenue and its cross streets became almost Disney battlements and turrets as I drove into the pink and orange sunset.
How fitting, then, that as I caught sight of this I was listening to a cassette of Playford tunes by The Broadside Band. Something medieval would have been even better, but this worked. It was the right mix of tootly and tinkly and dramatically oomphy to go with the visions of princesses and white horses that the almost-sight of a castle with tall pointy bits and an orangepink sunset suggested.
When my life ends, I have to find a way to stick around to watch the credits. I like how the special effects are synchronized with the sound track.
There ought to be some mathematical function that takes the shape of the front of a rowhouse and assigns a number prepresenting how ornate its construction is, or at least how curvy. Then it would be cool to generate a map of Baltimore showing the fancifactor and construction date of each house. I could have fun looking for patterns in that.
Of course, it would probably be easier to find a book in a library that just illustrated historical trends with pictures of a few typical houses from each decade, and a general overview of the ages of different parts of the city.
(no subject)
http://www.diydata.com/planning/brick_bonds/brick_bonds.htm
There's generally a strong correlation between brickwork and century:
Flemish Bond --> 18th century
American Common Bond --> 19th century
Stretcher Bond --> 20th century
but there are a zillion variations.
Coming up with an algorithm to classify a limited set of simple geometric patterns would be a lot easier than doing the same for wooden fretwork.
(no subject)
Still, you've given me another thing to look for as I drive around Baltimore (and I'll have to remember to pay attention to the brickwork the next time I visit Boston, too), and it's cool to find out both the names of the bonds and that "bond" is the word to search on. Thanks.
(no subject)
That's a good question. I'm guessing the answer is "sort of". If you had more money to throw into construction, you'll go for both fancier brickwork AND more frills.
If the answer is "no", well I DID say it was a simplified version of the problem.
It was quite common to use a more expensive bond on more visible areas of a house.
A few examples you'll be familiar with:
Savage Mill is a riotous mixture of bonds, but mostly variants of American Common.
Marietta House mixes bonds depending on which way the wall faces: more visible --> more expensive.
UMCP uses Flemish Bond for EVERYTHING, including the SECU ATM structure.
Note that brick-size has been standardized for at least a couple centuries. This should be useful for any automated analysis you do, since once you find the brick-texture ina an image, you'll know that each course of bricks will be 2.5" tall.
Brickwork
http://www.diydata.com/planning/brick_bonds/brick_bonds.htm
There's generally a strong correlation between brickwork and century:
Flemish Bond --> 18th century
American Common Bond --> 19th century
Stretcher Bond --> 20th century
but there are a zillion variations.
Coming up with an algorithm to classify a limited set of simple geometric patterns would be a lot easier than doing the same for wooden fretwork.
LJ user interface
Nothing happened, so I pressed it again.
(no subject)