When I first started thinking about the housing, I thought it was going to be a quick and easy process of simply buying the Polystyrene, placing the cameras inside, cutting some holes for the lenses and then duct-taping it closed. However, after the apparent parachute failure of the first test sounding I observed, I gave the box a bit more thought as it was essentially the protector of my entire project…
Initially I got a few cheap Polystyrene cooler boxes from Merrypak (on the way back from getting the GoPro and the Powershot SX620), which I used to play around with ideas. When actually placing the cameras inside for the first time, it was apparent that the small 2cm thick walls were definitely not going to be enough to protect the cameras on impact with the earth. This is when I first started playing around with the idea of building one myself. Sometime later, my dad gave me a link to a website: www.hotwired.co.za. This proved very useful as the sole purpose of the company is to build custom Polystyrene pieces. I sent them a message through their website, but it seemed to be broken. I knew this because the screen went red and a popup said “this site is broken”. Okay it didn’t actually say that but it did in my mind. I decided that the best thing to do was to give them a visit, it would be the fastest and most efficient way of getting the information instead of back and forth with emails.
So one rainy afternoon, I went to their workshop in Observatory to find out what they did. What I found out is that I should create a 3D model of what I want, send it through to the email they gave me and ask for a quote. In my mind I wanted a hard outer case (Polystyrene), and then something softer on the inside, which would insulate and protect the cameras on impact. My dad suggested I use Styrofoam, which is a plastic based material that is light, hard to break and could cushion the fall nicely. I asked them if they used this material and to my disappointment said no.
“However…” they said.
“Yes??” I replied, a bit too enthusiastically,
“We do use a material called Sondor”.
“What is it?” I asked, this time more reserved.
They proceeded to tell me that it is that soft stuff found in most delicate equipment boxes, that’s usually cut out to the exact shape of the thing you’re using and is incredibly spongy and is very good cushioning, as well as a good insulator, which is exactly what I need.
Pleased with myself on a successful mission, I went home and got to work on designing my model. This proved trickier than first anticipated. I thought it would be a breeze, knowing my history with 3D modeling, however it was not. I started off the designing by first trying to perfectly balance all my cameras onto an A4 sized-board, which would simulate the base of the box.
Being the efficient person that I am (another term is “lazy”, but I prefer “efficient”) I found the nearest object to my seated location that could simulate a pivot point for the board. This happened to be a book, which was a French version of the “The Little Mermaid” (titled “La Petite Sirene”). It was roughly the size that I needed and required the least amount of energy to obtain. Efficient. I then grabbed a bunch of books on a shelf in the office that I didn’t even know existed until then, mostly on computer graphics, networking, introduction into advanced calculus, etc. I used those books, all that knowledge, to prop up the French version on the little mermaid, so that I could balance a board on top of that, all with a bunch of cameras in the balance (literally). I tried again and again to find a setup of the cameras that would result in perfect equilibrium, however, every time I thought I had achieved this, the thing fell over, or a camera would slip, forcing me to do the whole thing again. Finally, after messing around with this for about 2 hours, I gave up and resorted to something I knew: Blender.
Blender is a 3D program that I have been using since I was 12, so I thought I would take my chances with that, rather than La Petite Sirene. I built a rough version of each of the cameras in the program, assigned each a weight, set up simulated physics and attempted to do in the program what I was trying to do in the real world. However, as if mocking me, as soon as I ran the simulation, the objects just danced around the screen and didn’t want to simulate properly. By this point I couldn’t be bothered to look up the problem and fix it, I was tired, it was dark and cold outside, and I had had enough [series time].
Once done with my series for the night and I was getting ready for bed, my dad made a suggestion: instead of trying to create a perfect center of gravity, rather build it so that it works, and then create different tensions in the holding lines afterwards, which could counter balance the imperfect weight distribution.
Slowly, one cog at a time my brain worked until it understand what the other person in the room had said (as I mentioned earlier, I was tired). The only response I could think of was: “Couldn’t you have told me that 8 hours ago…”
And with that I went to bed.
The next day was a different story. I started designing the box in SketchUp, which is much easier to use than Blender, for the designing purposes anyway. I played around with various designs but couldn’t make up my mind on the dimensions of the box, or the configurations of the cameras. Eventually I decided I would use the SX500 (the best one) to be focused on the horizon, the other Powershot to be focused up at a 45° angle, the Hero 6 focused on the horizon as well (but in video mode) and the Hero 4 facing down. Then I started off a standard A4 size plane, which I decided was best, as I could work my way from there.
The box had 4 main functions: 1. Outer rigidity. It had to be able to withstand a beating on the outside without compromising the inside. 2. Soft inner. While taking a beating it needed to make sure that its contents couldn’t feel that it was taking a beating. 3. Stability. The cameras should not be able to move. This will result in clean footage and mint pictures. 4. Insulation. At high altitudes it gets cold. Very cold. So it needed to be able to retain heat easily.
With these ideas in mind, my configuration and my dimensions, I got going. For real this time. I started off by making 3D models of all the cameras to their right size, so I could design around them. I started with the biggest camera and made sure that the lens had a tight, but snug fit. I then calculated the Angular Field Of View (AFOV) for the given focal length and the sensor. Using basic trig, I worked out that the equation for this is:
With this information I was able to derive a conical shape that would ensure that hole made in the side of the box would not be visible to the sensor, meaning clear photos.
Once finished with the big camera, I moved on to the hardest one. This was the one that was going to be pointing up at 45°. I had to apply the same rules to this as I did to the first camera, so I had to make a hole for the lens, I had to make sure that the hole size expanded with the size of the walls, so that the sensor would not pick up any white Polystyrene, all at 45°. I found it is actually rather hard to intersect Boolean 3-dimensional solid polygonal geometry through another solid geometry while operating on an irregular angular plane…
I had to create two pieces of geometry: the wall of the box, and the AFOV, extended to 200mm (This basically means that I took the size of the lens, and I then stretched it out by 20cm, so that I had a long tube. However, if I were to place my camera on the one side of the tube and look through it, most of the image would be dark, do to perspective geometry.
In order to overcome this, I scaled up the other side of the tube until there was no more dark in the picture, and it looked like there nothing in front of the camera. This was worked out with math and would allow me to cut a hole in the side of the box with absolute confidence that nothing would be obscuring the lens’ view).
Knowing that this box was going to be molded, I had to ensure that none of my geometry was 2-dimensional, or in geometry known as a “plane”. This is where I made my AFOV object piece a solid by extruding the walls outwards to simulate solidity. Once I had that, I rotated it by 45° and placed it in its correct location and then intersected it with the wall of the box. This allowed me to delete parts of the geometry that were unwanted and I was left with a hole in the side of the box that perfectly matched the AFOV of the camera. However there was one problem. It turned out that 45° was too raised and I would not see any Earth in the photo, only the sky, which could just as easily be achieved on the ground…
So I started again. Same process, but this time set it to 30°. This also allowed for a lower lid, which meant less volume, which means less weight, which means more altitude, which means better photos, which means better project, which means happy student, which means happy teacher. So all in all a good move.
Once the two biggest cameras were done I ‘quickly’ (10 hours) finished the holes for the two GoPros and hence finished the exterior of the box. Then I modelled the inside, making sure that all the Sondor cushioning was at least 5mm smaller than the size of the camera, which made the camera a tight fit, adding to stability and insulation.
Once the inside was finished and the model was cleaned up (all unnecessary vertices and polygons removed) I added some last touches, such as indents for binding tape that will be used to seal the box, and extra padding on the corners of the boxes, for extra protectiveness and lastly, the lid.
Then, I saved the project, stopped screen recording and sent the model to the workshop where I await the quote. Hopefully I can actually pay for the thing. It would be awkward if I couldn’t…
Here’s what I designed using the Sketchup software:
