Photogrammar
September 3, 2014 5:11 PM Subscribe
Photogrammar is a web-based platform for organizing, searching, and visualizing the 170,000 photographs from 1935 to 1945 created by the United State’s Farm Security Administration and Office of War Information (FSA-OWI).
There is an interactive map for the approximately 90,000 photos that have geographical information. For example: New Orleans
There is an interactive map for the approximately 90,000 photos that have geographical information. For example: New Orleans
You can also search by keyword. Parade.
posted by ColdChef at 5:23 PM on September 3, 2014 [1 favorite]
posted by ColdChef at 5:23 PM on September 3, 2014 [1 favorite]
Historical time sink! Love the plunge! Thank you!
posted by Alexandra Kitty at 5:40 PM on September 3, 2014
posted by Alexandra Kitty at 5:40 PM on September 3, 2014
From photos of Lancaster County, Pennsylvania: Mrs. Royer's preserves in the cellar of the house on the Enos Royer farm. Thank you, ColdChef!
posted by MonkeyToes at 5:40 PM on September 3, 2014
posted by MonkeyToes at 5:40 PM on September 3, 2014
The map has some issues. It seems to think St. Louis, Missouri is in California. And it's confused St. Charles, MO with St. Charles, Louisiana. But hey, checking out the rest of my state: He really knows how to handle some hot buns. This looks like a place I want to visit. This looks like a sweet ride. And she is making me feel really lazy.
posted by BlueJae at 5:46 PM on September 3, 2014
posted by BlueJae at 5:46 PM on September 3, 2014
I was fortunate to meet Russell Lee when he was teaching photography at the University of Texas. This is the most complete collection of his iconic Pie Town photos I have seen online...
posted by jim in austin at 5:49 PM on September 3, 2014
posted by jim in austin at 5:49 PM on September 3, 2014
I went right to Duval County, Florida to see what there was to see of Jacksonville back in the day. It's mostly military stuff from the bases here in town, so no landmarks that I recognized. However, at the bottom was six photographs taken in the African American section of town. The photographer? Gordon Parks.
posted by AlonzoMosleyFBI at 5:53 PM on September 3, 2014 [1 favorite]
posted by AlonzoMosleyFBI at 5:53 PM on September 3, 2014 [1 favorite]
These "evicted sharecroppers" photos by Arthur Rothstein struck me so that I felt compelled to look up the story.
posted by BlueJae at 6:20 PM on September 3, 2014
posted by BlueJae at 6:20 PM on September 3, 2014
Search for things like "children" and "Main Street" turns up terrific results.
Jack Whinery, homesteader, with his wife and the youngest of his five children, Pie Town, New Mexico
Rural school children, San Augustine County, Texas
Poor whites, Georgetown, D.C. (And the intentionally labeled contrast: Healthy white children, Washington, D.C.)
posted by beagle at 6:24 PM on September 3, 2014
Jack Whinery, homesteader, with his wife and the youngest of his five children, Pie Town, New Mexico
Rural school children, San Augustine County, Texas
Poor whites, Georgetown, D.C. (And the intentionally labeled contrast: Healthy white children, Washington, D.C.)
posted by beagle at 6:24 PM on September 3, 2014
Fox chained to automobile. Moorehead, Minnesota.
Fort Kent, Maine (vicinity). A typical dog in Aroostook County, This dog is on the Gagnon farm.
Black cat in snow. Ross County, Ohio.
Farmer's wife feeding chickens, Scioto Farms, Ohio.
Farmer and wife shingling roof of farmhouse. Old roof was of earth. They had to wait five years before they could get shingles to correct the leaking roof. Williams County, North Dakota.
From the About page:
posted by MonkeyToes at 6:26 PM on September 3, 2014
Fort Kent, Maine (vicinity). A typical dog in Aroostook County, This dog is on the Gagnon farm.
Black cat in snow. Ross County, Ohio.
Farmer's wife feeding chickens, Scioto Farms, Ohio.
Farmer and wife shingling roof of farmhouse. Old roof was of earth. They had to wait five years before they could get shingles to correct the leaking roof. Williams County, North Dakota.
From the About page:
In order to build support for and justify government programs, the Historical Section set out to document America, often at her most vulnerable, and the successful administration of relief service. The Farm Security Administration—Office of War Information (FSA-OWI) produced some of the most iconic images of the Great Depression and World War II and included photographers such as Dorothea Lange, Walker Evans, and Arthur Rothstein who shaped the visual culture of the era both in its moment and in American memory. Unit photographers were sent across the country. The negatives were sent to Washington, DC. The growing collection came to be known as “The File.” With the United State’s entry into WWII, the unit moved into the Office of War Information and the collection became known as the FSA-OWI File.ColdChef, this is a delightful find. Many thanks for bringing this to my attention.
posted by MonkeyToes at 6:26 PM on September 3, 2014
There are over 3000 Dorothea Lange pictures.
A new start. Bosque Farms project, New Mexico
You've seen this one: Destitute pea pickers in California. Mother of seven children. Age thirty-two. Nipomo, California
Child of former tenant farmer, now a day laborer. Ellis County, Texas
posted by beagle at 6:35 PM on September 3, 2014
A new start. Bosque Farms project, New Mexico
You've seen this one: Destitute pea pickers in California. Mother of seven children. Age thirty-two. Nipomo, California
Child of former tenant farmer, now a day laborer. Ellis County, Texas
posted by beagle at 6:35 PM on September 3, 2014
searching "automobile" turns up some amazing images.
posted by TMezz at 6:40 PM on September 3, 2014 [1 favorite]
posted by TMezz at 6:40 PM on September 3, 2014 [1 favorite]
In case anyone hasn't noticed, when you get to a photo page clicking on the Library of Congress call number will bring you to that photo's LoC page containing larger jpegs and much higher resolution TIFF files.
posted by plastic_animals at 7:07 PM on September 3, 2014 [4 favorites]
posted by plastic_animals at 7:07 PM on September 3, 2014 [4 favorites]
This is a project I've been supporting for the last year and a half, as a small part of an amazing team led by a faculty member and graduate student in American Studies. Thanks to the NEH Office of Digital Humanities for their essential support of this project.
One of my favorite parts was building this visualization of the 1942 three-level classification system. Many of the photos were originally assigned a category ("The Land > Farms > Fences"), but until a team member figured out what the code numbers were and how to link them to the separate typewritten list of categories, you couldn't use them to browse. Even the system's creator, Paul Vanderbilt, never saw his taxonomy rendered in d3.js.
Another fun experiment was the "California Metadata Explorer", which I built in dc.js. Click on various charts and graphs to see what it does. We'd like to build this for all states; I started with California because I'm from there. :)
The geocoding, done by our amazing GIS specialist and visualized using CartoDB and the shapefiles of 1940 counties is is a work-in-progress; we're on the third revision or so and hope to make it even more accurate in the future. (Namespace collision is a huge problem; we hope to restrict by state and county to find the right small town and avoid the Springfield problem.) The chloropleth map was just updated last week to normalize for area, which helps make huge size disparities such as San Bernadino (huge!) and San Francisco county (tiny!) less of a problem. I also like the dot map, which we clearly shows Jack Delano driving on Route 66 to take his pictures!
We also recently made public a feature to show semantically-similar pictures to the one you're currently looking at. We did this by transforming about 90,000 captions into vector space -- essentially, turning each caption into a complex line in an imaginary n-dimensional world, where n is the number of unique words in the corpus. Linear algebra lets us measure the cosine of the distance between these imaginary lines very efficiently, and this turns out to be a good proxy for photos that are described in a similar way. So even though there is no explicit category for "German Americans", those two words in the caption of this photo from Nebraska pulls in similar captions (closer lines) that describe Marjory Collins' 1942 picture of the German-American Club in New York City, and a German-American laborer photographed by Gordon Parks in 1943 in Connecticut. So by using text-mining techniques to surface latent structure, we hope to draw folks deeper into the archive!
posted by squid patrol at 7:37 PM on September 3, 2014 [20 favorites]
One of my favorite parts was building this visualization of the 1942 three-level classification system. Many of the photos were originally assigned a category ("The Land > Farms > Fences"), but until a team member figured out what the code numbers were and how to link them to the separate typewritten list of categories, you couldn't use them to browse. Even the system's creator, Paul Vanderbilt, never saw his taxonomy rendered in d3.js.
Another fun experiment was the "California Metadata Explorer", which I built in dc.js. Click on various charts and graphs to see what it does. We'd like to build this for all states; I started with California because I'm from there. :)
The geocoding, done by our amazing GIS specialist and visualized using CartoDB and the shapefiles of 1940 counties is is a work-in-progress; we're on the third revision or so and hope to make it even more accurate in the future. (Namespace collision is a huge problem; we hope to restrict by state and county to find the right small town and avoid the Springfield problem.) The chloropleth map was just updated last week to normalize for area, which helps make huge size disparities such as San Bernadino (huge!) and San Francisco county (tiny!) less of a problem. I also like the dot map, which we clearly shows Jack Delano driving on Route 66 to take his pictures!
We also recently made public a feature to show semantically-similar pictures to the one you're currently looking at. We did this by transforming about 90,000 captions into vector space -- essentially, turning each caption into a complex line in an imaginary n-dimensional world, where n is the number of unique words in the corpus. Linear algebra lets us measure the cosine of the distance between these imaginary lines very efficiently, and this turns out to be a good proxy for photos that are described in a similar way. So even though there is no explicit category for "German Americans", those two words in the caption of this photo from Nebraska pulls in similar captions (closer lines) that describe Marjory Collins' 1942 picture of the German-American Club in New York City, and a German-American laborer photographed by Gordon Parks in 1943 in Connecticut. So by using text-mining techniques to surface latent structure, we hope to draw folks deeper into the archive!
posted by squid patrol at 7:37 PM on September 3, 2014 [20 favorites]
Oh, this is completely and utterly fantastic!
I went straight to San Francisco and got dozens of photos by Dorothea Lange. ( ... I'm pretty sure that first one's supposed to be "North Beach", not "North Bean," although now I really really want to go find the North Bean neighborhood.)
The Salvation Army - Self-Denial and The interview for unemployment compensation are amazing.
Thank you for this!
posted by kristi at 8:00 PM on September 3, 2014
I went straight to San Francisco and got dozens of photos by Dorothea Lange. ( ... I'm pretty sure that first one's supposed to be "North Beach", not "North Bean," although now I really really want to go find the North Bean neighborhood.)
The Salvation Army - Self-Denial and The interview for unemployment compensation are amazing.
Thank you for this!
posted by kristi at 8:00 PM on September 3, 2014
Child of former tenant farmer, now a day laborer. Ellis County, Texas.
posted by donajo at 8:08 PM on September 3, 2014
posted by donajo at 8:08 PM on September 3, 2014
On preview, beagle beat me to it, but it's near my hometown, so.
posted by donajo at 8:11 PM on September 3, 2014
posted by donajo at 8:11 PM on September 3, 2014
Thank you, plastic_animals about clicking on the Library of Congress call numbers to access larger resolution images. Excellent!
posted by rmmcclay at 11:14 PM on September 3, 2014
posted by rmmcclay at 11:14 PM on September 3, 2014
Beautiful work, squid patrol. Keep it up. And thank you for your lovely in-depth description of how you organized your similar images suggestions. Sets my nerdy heart aflutter, it does.
posted by BlueJae at 5:26 AM on September 4, 2014
posted by BlueJae at 5:26 AM on September 4, 2014
1937 Vico Motor Oil Map
You never have to change it; it just recourses
posted by thelonius at 5:47 AM on September 4, 2014
You never have to change it; it just recourses
posted by thelonius at 5:47 AM on September 4, 2014
"Lumberjack turning handspring near Littlefork, Minnesota": delightful!
http://photogrammar.yale.edu/records/index.php?record=fsa2000011352/PP
posted by wenestvedt at 7:19 AM on September 4, 2014
http://photogrammar.yale.edu/records/index.php?record=fsa2000011352/PP
posted by wenestvedt at 7:19 AM on September 4, 2014
« Older Bah! | A Democrat and an Independent run together against... Newer »
This thread has been archived and is closed to new comments
posted by chatongriffes at 5:21 PM on September 3, 2014