I had a lot of ideas for this project. I tried to make my fishbowl into a camera obscura, then I tried to mod my Sims 4 game to track the sims location. I settled on creating a panorama of the bridge on Schenley Drive– it’s covered in locks that signify love. I have very limited experience with photography and photo editing and photo stitching so I wanted to learn about the process.
I started by using a telephoto lens and a tripod. I moved the tripod along the sidewalk across street. This would minimize distortion. I made two passes along the bridge, moving towards Oakland to capture the top part, and back towards Schenley to capture the bottom. Unfortunately as the sun dipped behind the clouds the lighting was a bit inconsistent between photos and because the sidewalk was uneven, distortion became inevitable.
I planned to use Autopano, but because a large number of the photos contained similar landmarks, the Autopano algorithms were confused and stacked the photos on top of each other, ignoring the location of the locks. I used Photoshop instead, which was more successful.
My final plan was to create an annotate-able photo of the entire bridge so that the owners of the locks could identify and write something about their locks.
Comments from the group review.
General feedback: strong choice of subject. Lots of room for material quality to improve (blurry etc.). A better capture rig could be developed. Gigapan? Interesting subject matter- ‘social objects’ Is crowdsourcing annotations the right direction, or would it be richer if you gathered and searched for the annotations yourself? I.e. you piecing the story together (even if invented) Kickstart the process of talking about these objects. Could be organized by time, color, think of ways to filter these objects. Thinking about locks as a dataset. Index of emotional attachment. Database for storytelling.
Clearly you learned about DSLR photography, telephoto lenses, failure of Autopano (because of nearly identical landmarks in each photo)
Stitching this with autopano involves manual image adjustment, manual control point adjustment, and more. It’s a pain. One way to start, before manual intervention, is under ‘optimization’ in the group settings, uncheck off “clean up bad links”. Since your links will be bad due to the planar movement of the camera. The core problem of stitching with planar movement is parallax, but the fence is mostly flat to the face of the camera; so this should be able to work. Go to the control point editor and delete any points that 1) are wrong (finding too much similarity) and 2) are not of the flat fence facade, which – due to parallax – will mess things up.
Would be cool to have some data with position on bridge and frequency of locks 🙂 +1
Graph the bridge by number of locks? Color of locks? Size of locks? Graph is the literal length of the bridge?
Manually select out all the padlocks from the bridge and graph them by color and location or something? A-la Golan’s collection lecture
^^ Each of these locks could be thought of as a data point
The final idea is interesting but for now it feels like it is still in a very early stage, feeling void of content and concept.
Why maintain the current arrangement of the locks?
Wait why have a physical lock if you can have an eternal annotation in the digital world.
There is slight bowing—easy to fix and straighten.
What if you did an analysis of people also still together (and graph with position on bridge?? :O )
Do you have any personal link with/interest in the lock bridge?
Interesting final idea for the project → could be taken further
Have you at all researched some of the “politics” of lock bridges? Esp in Paris, other places in europe, the trend of putting locks on bridges is actually destroying the bridges themselves, and
cities have to go in and cut off all of the love locks. Your idea for creating a digital, interactive version could be a way of preserving the gesture of these locks after the cities have to cut them.
The end result also looks like a scatter plot
I love the annotating locks idea. Kind of like how they put their initials though?
I feel you about the camera thing. “I’ll just take some photos!” << so hard So will this be presented as a panorama? Is this it’s final form? Kind of want to see the failed autopano panoramas lol You don’t need a telephoto lens for this, and you would probably get better results from using a corrected wide angle. It would be interesting to see the result of ‘brightening’ the locks on the bridge or auto-masking them away from the bridge. I would love to see somehow also a way if they could be sorted by date? I would love to see if my parents put a lock on it when they went to college here Re:annotations. These people have already annotated the bridge but putting their lock on it – the locks themselves are the annotations (!!!!!) → there is often enough data paired with the lock itself that you could use if need be