Saturday, October 26, 2013

Google vs. Feet On Location

Philadelphia's 30th Street Station - The setting for a scene in DOHA 12 by Lance Charnes

I met a great Author, Lance Charnes, who conducted a fascinating experiment perfect for Novel Travelist. He mentally built Philadelphia's 30th Street Station via google and then compared his accuracy during a trip to Philadelphia. 

Building 30th Street Station By Lance Charnes

In my international thriller Doha 12, assassins follow our heroes Jake Eldar and Miriam Schaffer to Philadelphia’s 30th Street Station. There, the bad guys launch an ill-prepared attempt to kill Jake and Miriam, which devolves into a three-way shoot-out running the length of the terminal. This is one of the major set-piece scenes, and because the place is so familiar to a large number of people in the Philly metro area, I wanted to get the setting right.

One problem: like most indie writers, I’m not working on an advance. All travel and research is completely on my own dime. I needed to do as much research as I could for free, since I had no idea when or if I’d ever get to Philadelphia to check out the place in person.

Interior of 30th Street Station - View from Stairs
Whenever I select a setting for a scene, I try to harvest as many high-quality pictures as I can from as many angles as possible. Google Images is perfect for this; put in your search term, and you get back a flood of photos from all manner of sources, including Flickr, newspapers, TV, and so on. If you do this, keep crawling through the results; the farther in you get, the more offbeat the sources. 

  • Train enthusiast websites had close-up pictures of the Amtrak information desk and board, and shots of the arrival platforms. 
  • An advertising firm showed an ad placement it had done in the main hall. 
  • Someone thought to take a snap of a women’s restroom. 
  • Another traveler had been there at Christmas (the shoot-out happens in early December), so I got much-needed pictures of the decorations, including the giant, perfectly conical tree at the east end of the concourse.

At the end of this process, I had a big collection of still photos, but no good idea about the layout of the place. I’d found only one small, blurry floorplan on the Amtrak site. From that and basic photointerpretation skills (I used to be in intel), I constructed a reasonable plan for the concourse; everything else was a guess.

I had lots of random still photos, but no clear idea of the layout.

Next, I turned to video. YouTube offered up 23,000 hits on “Philadelphia 30th street station.” Here’s where the weird diversity of the Internet truly came into play. There were tons of trainspotting videos; after digging through these, I found the one I needed, an end-to-end video taken in a NJ Transit commuter train going from Cherry Hill (NJ) Station to 30th Street – exactly the route our heroes take. I found videos taken by people walking through the concourse (note to future videographers: whip pans are lousy to watch), waiting for pickup outside, a flash mob dancing in the concourse, and a guided tour of the station at Christmas courtesy of a Philly yoga enthusiast. I plowed through a lot of truly awful video (too dark/too bright/out of focus/taken during an earthquake), keeping the links for the ones that were the most helpful.

The videos showed me: 

  • How people move through the space
  • What you can see from where
  • Some of the ambient sounds (note to future videographers: shut up and let the location speak for itself). 
I still didn’t have a good floorplan, though. I used the videos to refine the less-than-wonderful one I’d been able to scratch together, then forged ahead and wrote the scene.

A couple months passed. During an editing session, I decided to see if anything new had surfaced on the web. Lo and behold, the website Metro Jacksonville (Florida!) had posted an essay on the Amtrak Keystone Corridor train service, holding it up as an example for Jacksonville transit. The post included a reproduction of 30th Street’s visitor directory. Not only was it a clear, accurate floorplan, but it told which vendors were in each of the commercial spaces. Eureka!

It also showed that beyond the concourse, my cobbled-together floorplan was mostly wrong.

I dragged this treasure into Photoshop and did some measuring. The real concourse is 135’ wide by 290’ long. The map concourse was 177 pixels wide by 352 pixels long. With a bit of fudging, I was able to lay down a 9’ (three-stride) grid on the map concourse. I could finally measure distances and sizes throughout the terminal, time out how long it would take my characters to move from place to place, estimate how far they could shoot and what they could hide behind. I rewrote the scene using this new information and hoped it was good enough.

Fast-forward to October 2011. Through a series of circumstances I won’t bore you with, I got to go to D.C., Philadelphia and New York City. A few days before Halloween, I found myself standing in the concourse of the real 30th Street Station.

First, it’s a tremendously strange feeling to finally be in a place you’ve known only through photos. (Going to the Parthenon felt exactly the same way: damn, it really looks like the pictures.) Secondly, it’s very strange to go someplace you’ve never been and know exactly where everything is. I spent the next ninety minutes roaming the station, taking pictures and making notes. I traced the steps my characters ran, took cover behind the obstacles they used, checked the sightlines, confirmed which windows would get hit by the missed shots. I have no doubt I’m now on some Amtrak Police watchlist for all the suspicious things I did that morning. What kind of law-abiding citizen takes pictures while crouched behind a bench?

Statue - Great for my hero to hide behind.
The upshot? I had to make only minor adjustments to the action. What I’d pasted together off the Internet turned out to be about 95% right. The other 5% involved the passage of time and the tricks camera play: signs and trash cans had moved, some of the stores had changed out, the half-walls around the stairs leading to the tracks were lower than they looked (or I’d been measuring them against short people). I took notes and made these tweaks when I got home without causing another rewrite.

I also used this research method for some of Doha 12’s other settings, such as the Manhattan Diamond District, Central Park East’s Temple Emanu-El, and Brooklyn’s Green-Wood Cemetery. It worked time and again. Still, there are limits. Photo lighting isn’t always normal lighting; some places are darker or lighter in real life. You can’t feel the air temperature (chilly in the station), and until online Smell-o-Vision happens, you can’t get the ambient smells (cleaners and donuts in the station). This wasn’t a major drawback in my case. However, if your scene is set in a Kolkata meat market, the missing information may be crucial.

The take-home lesson: just because you don’t have an advance doesn’t mean you can’t accurately describe a setting in your writing. Another bonus: you can surf for hours and call it “research.”

Lance Charnes is an emergency manager and former Air Force intelligence officer. He’s the author of the international thriller Doha 12 and the upcoming near-future thriller South. He tweets (@lcharnes) about scuba diving, shipwrecks, marine archaeology and art crime, among other things.


  1. Great post, Lance! Totally hits home. I'll also add that even if you do travel to the location of your setting, it's difficult to foresee every little detail you will need for the scene unless you've already written the scene. When I was researching my first book (The Death Row Complex) I had almost finished the book before traveling to Washington, D.C., so I knew *exactly* what I needed to look for once I traveled there. But when researching The Vesuvius Isotope, I was still writing. So I didn't know every detail I would need, and was just taking in as much as possible. But after the travels, when the writing resumed, I still didn't have every single piece of information I ultimately needed. Putting the scenes together involved a combination of reviewing my own notes and photographs (some of which were odd and found nowhere else on the web) and also the perspectives of others. And yes, Youtube videos are golden.

  2. You're right -- it's far easier to do onsite research once you've written the scene. In a previous (yet-to-be-published) work, I ended up with a scene set in the restroom of a train station. Unfortunately, I'd never thought of checking out the restrooms when I was actually *in* that station. Now I make a point of it. (A mature bladder helps remind me to do this kind of field research.)

  3. HAHA! I'm cracking up because I learned the same thing when doing Vesuvius. Bottom line: When doing field research, ALWAYS check out bathrooms! They are important! (incidentally, there are multiple quasi-important scenes in Vesuvius that take place in the bathroom.)

  4. I'm totally giggling at both your bathroom research! Now I'm suddenly realizing that in all four of my books, I have important bathroom scenes. LOL!

    From here on out, when I make my check list of things to view at a museum for research, I'll add the bathroom!

  5. Haha!! I had totally forgotten about this little dialog, at least, in the forefront of my mind I had forgotten. Came back to this page as a brief distraction from NaNo writing. Guess what I just wrote this morning? A bathroom scene!

  6. Haha!! I had totally forgotten about this little dialog, at least, in the forefront of my mind I had forgotten. Came back to this page as a brief distraction from NaNo writing. Guess what I just wrote this morning? A bathroom scene!