Wednesday, June 17, 2015

Making A Scene

For 9 years daily, I have been curating photos of the day, seeking in more recent years what makes an outstanding image in a world of billions of photos.

SOS human chain around an oil drill in Ecuador posted in photo community I helped set up in 2007 for Al Gore's Live Earth, an event which had an audience of 2 billion. 

For 3 years daily, I have been mapping notable photos of movie scenes, music history,  and art history, threading them with other photos for locative context.

Bob Dylan is watching Dean Martin, Aug 1964. Dylan was house-sitting his manager Albert Grossman's Woodstock home (18 Striebel Rd, Bearsville, NY).  Photo by Douglas Gilbert.  

Bringing It All Back Home (1965) cover with Bob Dylan was photographed in 1964 by Daniel Kramer at Albert Grossman's Woodtock home. Sally Grossman is pictured with  items time-stamping that year. 

For 1.5 years daily, I started mapping biographies of notable people, where they made history, where they lived, and where they went to school. For some who made history, I mapped scenes of books they wrote. Quotes of the day were mapped in literary history.

It became fascinating to check-in people who never had an online profile. You could check-in a dead person at an extinct place to immortalize a story.

After visiting Woody Guthrie at Greystone Hospital (59 Koch Ave Morris Plains, NJ), Bob Dylan went to Woody's home at 3520 Mermaid Ave (Coney Island, NY)  in January, 1961, to look for songs Woody said Dylan could have. Arlo Guthrie answered the door. No one could find the songs. The house is gone now. Woody lived here 1943-50 before being hospitalized. The songs were later found by Nora Guthrie and recorded by Billy Bragg and Wilco on the album Mermaid Avenue (1998). 

Historically (before social media)....
I  started to map photos and stories from my dark room journalism days and pre-social media days.  Historical items and cultural artifacts were being geo-tagged for the first time.

Gwyneth Paltrow with Owen Wilson at 3rd Ave and 128th St (East Harlem) in The Royal Tennenbaums (2001)

Then and Now...
Geographical intersections with future imagery became profoundly interesting.

Deeper hidden history made wordless photos relevant.  Things we never knew were nearby started to emerge on a regular basis geo-tagged at the same place or vicinity.

I could see a Movieview of a street for a new perspective. Or where Patti Smith lived, where Steve Jobs went and where songs or paintings were composed near a photo I posted. Historical elements local to a photo enriched the viewing experience. 

Little Italy in Godfather (1974) was on E 6th St (between Ave A and Ave B) in New York City. The building with an arched entrance in front of Robert De Niro's head is 524 E 6th St. 

March 10, 1974, photo by Allan Tannenbaum on set of Godfather (1974) at 524 E 6th St. See the circle on the sign in Allan's photo and  see it again in the photo with De Niro. 

Photo sharing became enhanced with rich stories and alternate visual perspectives.

Personal Connectivity....
The most popular photos, I observed, were typically not the most beautiful photos or most artistic but rather photos that had the greatest personal connections. Viewers had been to those places, knew those people or identified with the story. And stories related to those photos  illuminated those personal connections. It was time or place viewers knew. Time and place defined taste and vice versa for a scene.

Fab Five Freddy  aka Fred Brathwaite was also featured in Blondie's video for Rapture: 

I started to think about an algorithm that made personal connections relevant in story networking.

Intelligent story connections...
A locative database I grew by hand curation  became increasingly powerful connecting stories in a way never seen before at this depth and breadth.  It could do for stories what Pandora did for music, correlating stories. 

The database became smarter at connecting narratives through cumulative curation. The more notable stories were tagged at places with photos  (accurately), the more powerful a streetview of a neighborhood became. Locative intelligence of future layers of stories became only possible with knowledge of prior geographical layers.

The quality of locative intelligence, visual imagery and narratives drove the capabilities of this locative media engine.

The only other search engine I have seen this engaging is Google.

Exclusive location IDs from geo-sleuthing...
A record 20,000 hours was spent charting where notable art, invention and history got made. 1.6 million pieces were geo-tagged into 160,000 posts, covering the most relevant spots in the world.

A fair amount of geo-sleuthing was conducted to uncover mysterious locations of notable imagery and stories. Stanley Kubrick's first film was geo-tagged for the first time.  As were other notable endeavors.

Meaningless photos offering geographical clues - a streetview of that time - became relevant to help identify locations nearby where history was made.  Live community posts helped solve location mysteries or correct mistaken locations in comment threads.

Points of Interest  Illuminating Trails of Trailblazers....
Stories that put places on the map were put on the map.  The roots of history makers visually illuminated future trails they blazed. The first extensive map of JD Salinger's life was created.

Last home of Catcher In The Rye author JD Salinger 
(GPS  43.511931, -72.346789)

Locative patterns were gleaned for the first time mapping where trailblazers made history. "Geo-patterns" emerged.

400+ homes of Greenwich Village writers. Greenwich Village has had the highest concentration of Nobel and Pulitzer Prize winners in the world as well as people on the cover of Time. 

This is an emerging field called locative media. A new way to exhibit media and art by location as points of interest. Stories paint the topography of a neighborhood and time in history.  Ghosts visualize a neighborhood.

Content gardening for a geo-tagged feed...
For coverage, Google, Apple, Facebook, Instagram and other geo-players are challenged with finding compelling content to create something astounding by location. But user-generated content has a track record of being more social in nature and very brief. Posting by phone is limited by time and scope.

The difference is between showing this scene from The Godfather was shot where you are and Ernest Hemingway wrote this inspiring quote here versus  just saying Jill met Jack here and enjoyed the food. Convenient user-generation produces lightweight insta-results. A community of curators produces heavyweight results.

Content gardening is critical for a decent viewing experience in locative media. The feed has to look very distinguished from social media. The question is how can a highly intelligent locative database produce something beyond Instagram  and Conde Nast for viewing, publishing and conversation pleasure.

Cafe Terrace (9-11 Place du Forum. Arles, France) as an Instagram for good coffee...or as an 1888 painting by Vincent Van Gogh. 

Designing a new locative medium...
It became obvious you cannot design a great new locative medium without the ideal content already populated. So that had to happen first. One needed to have a better idea of how locative content behaved and what the topics were. Enough relevant stories and visuals needed to be indexed first.

Otherwise, it is not so obvious how to architect and focus a compelling locative viewing experience. And that is just for viewing. Compelling community engagement and publishing have their own behavior.

GPS and maps seemed like natural tools for locative viewing. Stories and photos become  viewable by GPS, browseable on map and searchable by address.

But a map was never intended to be a browser for media. It is too slow to scroll regularly by the hour, minute or second. People use them mostly only when lost. Maps also present stories as obscure pins (sometimes nicknamed "chicken pox").  That's something a web page or newspaper would never do in information architecture by design.

Pins on a map show no stories up front. Compelling content is not presented.  What is there is unknown at first glance.

You even need to zoom in to see that more pins exist. Each pin can only represent one story per place even for a place with 100s of stories.  Invisible pins and stories are due to slow page loading. Maps are too slow to load already with just visible pins. 

I needed to think of new ideas for how to view locative imagery faster at a much higher volume level daily.

Maps also  could not facilitate community engagement readily. That posed yet another challenge. Maps lacked "stickiness."

Content refreshment needed...
GPS (also slow to load) can only show stories nearby, which is usually home, work or at a personal hangout. The experience is myopic and often stagnant if no new stories are added nearby. You can be stuck seeing the same old stories nearby all the time without exploring new stories.

Featured daily/hourly new stories from anywhere in the world worked better than GPS for content refreshment. Cultural affinity and personal connections are still key for stickiness but are not dependent on being nearby.

While on the go, there is also very little time to use GPS unless you are lost.  While travelling abroad, you might not have a data plan for GPS. GPS also consumes battery power fast.

I concluded GPS and maps (as they are today) would only be used as reference tools when needed, but not as a habit.

The addictive habit found....
A Facebook group featuring notable photos in Manhattan by location showed me something addictive. Threading related photos and stories daily in comments created a wider view of a scene.

Locative experiences I soon realized are less about pins on a map and more about creating a "scene." What I call, "scening it." It's about augmenting a post with more interesting stories or photos at a place, creating enriching conversations.

Great commenting (addictive to read)....

Comments in a cool locative community differ a lot from the "all about me" traits of social media and are more about "here's an interesting story you might not know."

People who were/are part of a scene engage in a conversation regularly in this group. They add their exclusive stories and photos of that scene, and their unique takes on it.

Creative and social history gets further documented to augment a scene. I learn something everyday and in almost every post. The photos are also killer to start off, to incite conversation.

Anatomy of a habitual community...
Here's  how a regular viewing habit was formed at this locative photo group for Manhattan on Facebook:

1. FEATURED FEED - A featured feed (with a variety of curators) shows  new scenes (photo, photo credit,  people in it, location, date and story). I don't have to go look for a "scene." Cool new scenes get pushed to me by the hour.

2. LIKE - People "like" the scene to push the scene up in the feed.

3. ADD STORIES - People add personal histories or observations of the scene in comments to move the scene up in the feed. This separates it from a traditional "photo dump" site, as one administrator noted.   The photo draws people to the scene. The personal stories make the scene.

4. ADD PHOTOS - People also add interesting photos related to that scene, including movie scenes there. Photos can show how a scene changed (then and now).  This augments the scene visually.

5. NOTIFICATIONS - You are notified if someone liked a post you liked or added something interesting.

6. LOCATION ID  - If a location is mysterious, others help to identify the scene's location.  Locative clues are highlighted to help everyone search. Viewers found it fun and obsessive to look for a mysterious place in a city they knew.

7. SHARE - The scene is shared to a personal Facebook page. More people join the community as a result. The scenes are also blogged about because of their cultural interest. New York Magazine called this locative community "high brow" and "brilliant."

8. SEARCH - A search engine is available to find scenes by name or place.

Accelerating scene augmentation....
With an intelligent locative database, cool photos and stories can be connected to a scene powerfully fast.  You can also follow people of your taste tagged to other scenes to see other points of interest rapidly.

You can branch off and explore fast.

Creating new scenes (curated addresses)...
Posting a new scene not already in the feed has been the hardest task to date when geo-tagging photos. After determining the scene/image is unique, you  have to figure out who took the photo and where it took place. This is usually done at home on a desktop. The best you can typically do by phone is post a photo of what a place looks like now - or just view a feed (like and comment).

The locative database I  have been curating  doesn't  use a pre-ordained venue database offered by Google, Mapbox or Foursquare.  I had to manually look for and type each new address (or lat/long) into the database.  But the benefit is that many notable addresses (homes, landscapes and new lat/long sites) are not in 3rd party venue databases already. Only this database now has these addresses compiled.

These are premium addresses because history was made there.  The addresses themselves are curated. There were even astrological connections on birthdays and at birthplaces featured daily.

There is no 3rd party  charging for traffic for an off-the-shelf venue database who can also  pull the plug if you don't pay enough.

A style guide was also needed to differentiate locations (and sometimes people) with the same names. There can be common names like Chinatown, Little Italy, City Hall, Main Street, 1st Street or Starbucks in every major city.

Sometimes it is hard to know if a photo has been posted before at places with so many stories.  There was no duplicate image detection other than personal memory of the feed.

Filtering views (ranking by topic)...
On average, history was made twice at a location that made history. But there were also event venues like Madison Square Garden with 100s of stories needing to be ranked for viewing.

Six topics emerged for presentation:

1)  Movie scenes
2)  Music history
3) Photos of the Day
4) History & Biographies
5) Literature & Books (Quotes & Scenes)
6) Art & Invention

Reminded me of Trivia Pursuit ...

Weekly since 2007, I have also been documenting music communities in Toronto and New York (as well as cross-country tours) - music history as it unfolds. I have snapped 100,000+ photos following the lives of several hundred Artists over 8 years who have each grown social media communities. It showed a geo-pattern for  where music history got made today.

Map of events chronicled for one singer in one city over 8 years. 

Photo Radius...
Notable stories and photos local to a new photo posted  gave macro-context. Something you cannot find in a typical social media photo feed.  Any place, person, photo or story  becomes enriched by nearby history.

Hidden stories get connected. People, art and places get connected in new ways.

In a prior post, I also noted a way to explore based on mutual interests to discover even more hidden stories driven by personal connections.

Saturday, May 23, 2015

Taxi Driver Diner Location Solved via Hell's Kitchen Mystery

There's no better place than Manhattan Before 1990 for a mysterious location of a photo to be solved. The detective work is  beyond anything a museum or library would do in New York City.  There might be stellar geo-sleuths out there but collectively this group is unrivaled.

The mysterious location of this 1979 photo (above) by M. Joedicke  took among the longest to find. The photographer didn't even know where it was.

There's a Pronto gas station sign  but let's just say Pronto came and went and nobody remembered it having even arrived in New York City.

The next clue was the liquor store telephone number with a 563 exchange which we later deduced had to be in Midtown:

But hold on. Some members wondered if this was even in New York City. How do we know for sure?  The bus stop is unique to New York:

That brings out the next obvious clue. This is a two-way street and it had an uncommon slope which also led the search towards Inwood, The Bronx and Harlem. That was far too wide of a search perimeter to find anything fast.

So another investigative approach was deployed. Another clue is the road is wet. A member then looked at all the rainy day photos in the original album.  He immediately found rainy photos on 34th Steet (this would later prove to be the correct street). But the photographer traveled to both East 34th St in Kips Bay and to West 34th St by New Yorker Hotel. He stayed at Hotel Mansfield on 44th St near 5th Ave and had also photographed 42nd Street all the way across. 

But that narrowed it down to only two, 2-way streets he photographed on a rainy day. Additionally, 34th St and 42nd St only sloped closer to the rivers. 

The taxi, garage and gas station led the member to believe it had to be near 10th Avenue where cabs often fueled.  42nd St is very well documented in New York photos so the member suggested focusing on 10th Ave and 34th St where there is an empty lot:

This turned out to be the correct spot but two notable members said it could not be the spot. One  respected  geo-sleuth said 2 buildings and a gas station could not fit. But the member suggesting 10th Ave and 34th St noted the slope, mentioned the photographer's 34th St and 42nd St rainy day photos and found a BP gas station on site there at one point: 

The member suggested 1979 gas stations could be skinny and noted how small the Pronto sign was. There was also a speculative guess by another member that a blurry address above one door was in the 400s. This site was 461 W 34th St. 

All the circumstantial evidence for taxis to fuel up, former gas station was here, two-way street, empty lot (open to possibilities), bus route, 563 exchange area, and a two-way street the photographer photographed on a rainy day (1 of only 2 two-way streets he photographed that day). The bet was the photographer (from another country)  had to have taken another photo during his stay nearby this photo. 

But still there were no firm believers of this spot being correct. Later, a former taxi driver and notable photographer  almost put it to rest by saying he had pumped gas here into his cab and it cannot be the spot. 

Then David Cohen revived the theory this could be 10th Ave and 34th St facing north...6 days later! There is a billboard which meant this had to have been a vacant lot, parking lot or gas station for a long time. 

Somehow he found video footage of the place  in a DJ mix on Youtube at the 2:11 mark. Here are stills at 10th Ave and 34th St (NE corner) with the same gas station sign holder and deli building (a coffee shop here), with a liquor store, by the extant 1929 building at 455 W 34th St  (behind signs): 

A comparison by Ruben Iglesias: 

The video footage is a beautiful document of 10th Avenue from 30th St to 46th St.  That led to another connection. 

For the longest time, quite a few Taxi Driver (1976)  geo-sleuths were wondering where the diner across from HESS was with Robert De Niro (r).

The same member who suggested the 10th Ave and 34th St had previously suggested this diner was across from the HESS gas station at 10th Ave between 44th St and 45th St. He couldn't figure out the angle because the architecture of HESS has changed since 1975 when Taxi Driver was shot. Bob Egan of PopSpots found a photo of how it once looked:

Using geometry, a city plan and directory listing, he was able to deduce the Taxi Driver (1976) diner had to be at 632 10th Ave:

There was an M&N Restaurant there. But we had no photo of the exterior or visual confirmation this place was correct....until this video footage at the 7:47 mark (used to solve the gas station mystery at 10th Ave and 34th St at the 2:11 mark).  There were images of 632 10th Ave at 45th St:

The ceiling light, door position and neon sign seemed to put it in the right ballpark for the Taxi Driver (1976) diner across from HESS. Bob Egan of PopSpots immediately replied.

He first stitched together two images for a wider view of 632 10th Ave:

He then used digital fingerprinting to confirm it's a match: 

Geo-sleuthing takes a lot of patience to wait for the clues to arrive and geo-artifacts to manifest. It also requires one to hold on to a circumstantial theory when others think it is wrong. There are many times a theory is wrong. Without the smoking gun, it could be wrong. 

It's never fully solved until you have visual confirmation.

Very humbling.

UPDATE - The  yellow circle should be over the bar on the left in the Taxi Driver scene and the green square should be over the H in HESS. 

UPDATE:  The DJ Mix video footage is originally from Chantal Ackerman's documentary News From Home (1977).  

Tuesday, May 12, 2015

Mystery in the Bronx

One day, I hope Manhattan Before 1990  on Facebook becomes a museum.

The level of photographic curation and location identification is unrivaled in New York City.

Members dig for rare photos and photographers share exclusive photos of New York City that incite compelling storytelling or investigation.

Last night, I thought this Facebook group would be stumped for the first time as I watched people make guesses at where a park and building were in a 1940s photo by Mike Katz posted by Ruben Iglesias.

The immediate clues were the address 35 and the circles in the metal fence, Ruben noted:

Other clues noted that turned out to be true...a store is to the right (via Ruben) and this entrance has a  "T-shaped inset" (via Matthew Rohn)  which is not that obvious in the photo. This is what it looked like after it was found:

Like I said,  not so obvious.

Those clues could be anywhere  in New York City and a lot  can change since the 1940s.  

The group looked at almost every fence and park in Manhattan facing an apartment building. But still no cigar. Only so many buildings in Manhattan could have the address 35 by a park/square. Almost all of them were noted in the comments. 

My instincts told me it was either demolished uptown (or in Harlem)...or even further from discovery  in The Bronx.

One member started looking in the Bronx because he remembered fencing with similar circles there.

Eventually a man who had worked for Ma Bell in the Bronx started looking there at buildings that felt familiar to him by a park.

He suggested 35 Mc Clelland across Mullaly Park.  However, even though he nailed the place, it was not so obvious.

The streetview looked like this:

He was about to move on, after thinking the entrance was not the same. But another member recalled the "T-shaped inset" clue for the entrance and felt the window pattern was similar (but he photo was not as detailed): 

Additionally, the fence at Mullaly Park had circles and the park had benches along a diagonal path in a similar angle as the one in Ruben's mysterious post:

The right side at 35 Mc Clellan also looked similar with a fire escape and store in the right position at 1147 River Ave:

There was still doubt. This could easily be architecture at other locations.   Investigating older streetview was needed from 2007 and it showed a matching entrance and ground level window (r): 

This is the kind of work few museums with photo collections do.  And that is why Manhattan Before 1990 gets my vote for becoming a museum one day. 

Locals of all kinds contribute clues to identifying locations for photos and add enriching stories. 

In this case, the telephone man Martin Balassi (a "Bronx sherpa")  solved it  after 15 hours of group searching:

The 3 benches along the diagonal path of Mullaly Park are now facing 35 Mc Clellan. Here is an aerial: