It’s a wrap.

Yo. This is the last post for this blog.

If you don’t know anything about this project, please do check out the about page and all of the posts here about the progress of the project. Enjoy your stay!

Also check out my blog thing, Do Things, to see other stuff I work on. You can find my contact information there as well if you have any questions about this project or about life or whatever.

Final Project Summation

I liked this project. Here’s some things.

Process Challenges

The scope of this project was very large. I started a few months early in the fall semester, which definitely made the rest of the process way easier. I was able to basically finish the pan/tilt module and get a few tests down before the actual project semester even began. I think this was actually pretty crucial, and if I did not have that kind of time management early on, attempting to do the entire thing in one semester would have been super high pressure. I was able to be pretty relaxed about my pace the entire time, which made everything much more enjoyable.

It was the most extensive exploration of 3D printing I’ve so far done. It included lots of gears, ABS printing, prints lasting up to 23 hours, integration with lots of other hardware, and some very tight tolerances. With careful planning and design, I was able to get most of these prints off successfully on the first attempt.

This was also the first time I’ve ever used stepper motors. They’re fickle little beasts, but I learned a lot about how to make them behave, and I highly value the addition they’ve made to my actuator toolbox. I’m certain I will use them more in future projects.

Funding was a big challenge and was hugely damaging personally. DCC was no help in supporting my project. This is a spiteful statement.

Theoretical Future Advance Stuff

If I were to continue this project, there is plenty of room to improve.

The 3D printed parts, mostly the gears, could benefit from being fabricated from a material like aluminum. Over time, the gears would wear down and some play would start to be noticeable.

The pan/tilt module could be improved if it was a bit more compact. It’s currently quite large and clunky. The linear module is also clunky. However, I think that being a bit larger is a manifestation of the 3D print & bolt approach that I used. The approach would be a bit different for a more commercial product, and I think that would allow everything to shrink down to very comfortable sizes.

The cable carriage is one of the biggest areas that could be improved, even though it’s technically not a part of the four module system that the core project consists of. Currently, to mount it on on the cable, I have to remove the pulleys, lift the heavy assembly over the cable, rethread the bolts through a washer, the carriage body, a spacer, the pulley, another spacer, the carriage body again, another washer, and then use two wrenches to lock it down with a locknut. The carriage could be improved with an easier means of mounting onto a cable, perhaps by fixing the pulleys to the carriage from only one side instead of two, so that it could simply be lifted and rested onto the cable. Then some sort of safety could be applied so it would not slip off.

The carriage could also be improved in its assembly. It’s less of an issue than the problem of putting it on the cable, but it is a bit difficult to tighten the bolts to mount the modules. It is also not easy, or rather wasteful, to remove and remount the batteries since they are attached by zip ties.

Also, the cable rig would benefit from a second stabilizing cable so that there would be less swing in the shots. This way, less stabilization would need to be used in post and the camera could afford longer exposures, which make for better timelapses.

 

Regent’s Drive Garage

Above is the shot that I set out to get today. A 287 foot span between a corner of Regent’s Drive parking garage and a tree in front of J.M. Patterson, overpassing three sidewalks and two streets. My roommate and I got all the lines set up and ready to tension when I realized that we might not be high enough on the tree in front of J.M. Patterson that the camera wouldn’t hit the tallest vehicles that might drive under that point. We couldn’t go any higher in the tree because of branches that would be in the way. So reluctantly, we decided to make a change of plans. Also, this guy helped convince us:

The new plan was to put the cable between two columns of the parking garage. This was possible due to the weird trapezoidal shape of the garage, but it made the lower southeast unusable due to being too close to the side of the garage.

There was also some unpleasant scraping of cable against concrete, which wasn’t nice. But this setup was still significantly less stressful than the original plan, given that it did not go over any roads or pathways.

It was a pretty close shave toward the lower end of the cable.

Also, this was my first time trying out this new remote control configuration. This is the first iteration of the remote control setup that has the Playstation controller mounted solidly to everything else. I also found that I wasn’t using the other LCD monitor as much as I thought I would, so I removed that from the rig. The LCD monitor had the battery pack on it though, so I unscrewed the battery mount from the back of the monitor and screwed it onto the remote control unit, which worked surprisingly well. The whole thing felt super solid really nice to handle.

I took two timelapses on this setup, one in each direction. I think they turned out pretty well!

Maryland Day Timelapses

Today was Maryland Day, and I took advantage of the opportunity to shoot a few timelapses of the massive amount of people and activities going on here on campus. I took a total of four timelapses in two different location. The first was a 135 foot span in Hornbake Plaza from 10:50 – 11:30 AM with 829 frames.

At one point during the capture of the Hornbake timelapse, a lady approached me and asked me about what I was doing. She introduced herself as the organizer of Maryland Day. I’d think there would be more than just one person behind this event, but okay. She asked me about my thing and who I was with at least twice; I don’t think she quite grasped that I wasn’t affiliated with anyone but myself. She wasn’t aggressive about my presence there, but she did seem a bit concerned about the low height my camera was hanging over their flag exhibit, which was a totally valid concern. She asked if I had gotten permission with anyone, and I said no, and she said that next time I should get permission and approval for my rigging next time (eh okay but rather not do the bureaucracy thing). The last thing she asked me was if the camera would hit any of the flags. I charmingly and confidently responded “of course not.” This ended our conversation, and I turned around to resume watching my camera, only to find that at that very moment it was scraping against the top of one of the flags. I found this hilarious.

The other three were a shorter 55 foot span diagonally between two trees on the plaza in front of Kim. I started shooting at 3:00, and the last timelapse, which was cut short because the camera battery died, took its last frame at 4:31. It was okay that it died on the last one because it was mostly only cleanup at that time anyway. These timelapes faced Kim and captured all the exciting engineering demos on the plaza.

Here, an extra carabiner is required for each carabiner connected to the rigging plate in order to rotate the carabiner by 90 degrees. The top carabiner had to be rotated in order for the plate to be in the correct orientation when connected to the sling wrapped around the tree, and the two pulleys had to be rotated 90 degrees so that they would bring the cables downward so I could do the tensioning assembly vertically.

I got a few people asking me questions at Hornbake, but the people at Kim were especially interested in what I was doing. I think most of them thought that I was another Maryland Day exhibitor. It was fun to talk to so many people and explain my project to them. Most people were super into it and thought it was cool. Some people just didn’t get it. But hey.

Super Windy Test

Today I rigged up a 380 foot span between a tree in front of Van Munching and a column on the North side and on the third level of the parking garage in front of Prince Frederick. The only other cable timelapse I’ve done before this one was at Bitcamp, which was indoors. This one was outside, 60 feet longer, and winds were averaging 15 mph with gusts of up to 25 mph.  The wind was definitely not ideal, but it was a good opportunity to stress test the system. Spoiler alert: results were not stupendous.

It’s normal for there to be some amount of bumps and instability. For the tests at Bitcamp I was easily able to correct this with digital stabilization in After Effects. But here, stabilization could only do so much; the swings were often too extreme to handle. But even though the video didn’t turn out great, it was still a great test of the system in extreme conditions.

First Complete Shots at Bitcamp

It’s been a crazy week in anticipation of Bitcamp, a 36 hour hackathon in Cole Fieldhouse. Not only have I had to deal with finishing this project in time, but I’ve also had to work on another project for the event. So I’ve been pretty worked, but everything came together.

For the timelapse rig at Bitcamp, I wanted to span a cable across the full 320 foot length of Cole, anchoring the cable on the roof supporting columns. The pair of cables were simply anchored to one side, and the ran through a pair of pulleys on the other side.

The pulleys brought the cables to the side to an adjacent column where they were tensioned and anchored.

The rigging held up well and felt solid. Nobody died. I’m definitely happy about that.

This is the remote control assembly a put together. I used a flash bracket with a handle on it and mounted the monitor where the camera would normally go, and the remote control module where the flash would normally go. The video receiver was strapped on the back of the remote control module. This was a really nice setup, and it was easy to carry around one handed and run around with. I think I’ll keep setting it up this way.

I captured a total of 11 timelapses over the 36 hours. Most of them took place over time spans of about 75 minutes, since that was about the limit of how fast the linear module could move. By minimizing the capture time, I was able to maximize the amount of timelapses I could capture.

Some of them turned out really well, others were more experimental and didn’t necessarily work out so well. Overall, it was a huge success, and everyone there really appreciated the work I was doing and enjoyed the timelapses.

Cable Carriage Assembled & First Test

Today, I had some friends kindly help me cut out the wood parts for the cable carriage on a CNC.

The first event I want to do an aerial timelapse over is Bitcamp, a hackathon in Cole Field House this weekend, 4 days from now. Before then, I need to get at least one test hang of the entire rig at ground level to make sure it works. However, it’s supposed to rain for the rest of the week, so I had to rush to assemble the carriage so I could get that done tonight.

There were some hiccups in assembly with parts that did not fit together the way I had foreseen, but I was able to get it all together. After that, the rigging went up fairly quickly. The span we were able to get was about 180 feet. Cole will be more like 300 feet, so another test would be ideal to ensure that the distance is feasible. Otherwise, the test went very well. The device was able to advance itself along the cable smoothly.

It was a little windy, so the camera was swaying a bit. I had foreseen this problem and expected this to happen. It should not be much of an issue indoors, but for outside shots it will be a challenge. I’ll also need to get spacers for the two large white pulleys that carry the rig so they don’t slide back and forth. Also, the bolts that I’m using with those pulleys were slightly too big for the pulleys, so I attempted to drill them out the holes in the pulleys a bit, but I ended up ripping out the metal bearings. Now it’s just plastic spinning on the bolt, and it’s getting eaten up pretty fast. I think I’ll have to order new pulleys and do something else about the big bolt.

Stakeholder Interview

For my stakeholder, I chose to interview Jared Mezzocchi, a faculty member in the school of theatre, dance, and performance studies who works in media and projections design and direction. I chose Jared because he’s a really cool guy who I thought would be interested in the project, which he was, and who would be able to offer creative input.

With his background in theatre and media design, he also offered some interesting perspectives on how the project could be used that I would have never thought of. He pitched a few different ideas on how my device could be used in different ways.

His first idea was to adapt the system from being used for timelapse to being used for stop animation. Instead of delaying a fixed amount of time between each image capture and camera movement, it would instead wait for a button press from the animator. It would be an easy adaptation to make, and it would open up an entire new realm of camera movement for stop animation that I hadn’t even considered.

He also had an interesting idea for a performance piece in which between each image that would be 30 to 60 seconds apart, groups of performers would attempt to build a set, perform a complete story, strike the set, and be gone by the time that next image would be taken. It could be like a competition, in which different groups have to all tell their stories in the space, and if anyone gets caught in an image, they’re out. That the camera is continuously capturing image in a timelapse that is supposed to be a representation of what is going on in the space over time, but the performers are betraying that idea, attempting to remain invisible and create a misleading final product.

We also discussed some ideas on different types of shots we could do. We had the idea of a long timelapse down the length of the Clarice atrium as people flow in and out of the space during the day. The timelapse would be helpful to me for my final video, and also helpful to the Clarice media team, so everyone would win. So Jared is working on talking to Clarice administration to get permission to get rigging points dropped from the ceiling to pull it off.

Speaking of administration, I also inquired for ideas about how to navigate the bureaucracy involved in rigging up cables between trees outside on campus. He agreed with my ask-for-forgiveness-not-permission philosophy, so that’s the policy I’m going to start myself on and we’ll see what happens. He also connected me with a former student he knew who pulled of a heist to set up a zipline on campus, which apparently remained up for several weeks before being discovered.

Jared was really into the project, and he asked me to keep him updated, so I’ll be doing that. I’ll also be letting him know when I’m setting up shots so he can come by and check it out.