All the scenes were filmed, close-up shots were taken, and editing is done. Now, the Sparrowscope project is officially on Kickstarter. Check it out!
The Sparrowscope has been an engineering project for most of its life, but for the past few weeks, horizons have expanded. I’ve decided to put the Sparrowscope as a product on Kickstarter, which has involved filming, audio recording, product photography and video editing. I’ve done none of the usual programming, soldering, finite element analysis, control loop tuning, or any of the other engineering tasks I got used to, which has been a nice artistic break. The Kickstarter project will be public soon enough (and I’ll post a link), but I thought it would be fun to share some of the behind-the-scenes looks at what I’ve been doing, and maybe help people out who are working on their own crowdfunding campaigns.
With friends to help me and lend some equipment, as well as a few rentals, it was possible to put together a decent-looking video without spending a fortune. What you need are patient friends, good lighting, a decent camera and some good sound equipment.
A bunch of my friends made this video possible—on set were Alyssa, Anthony, Audrey, Dana, David, Evan, Juliana, Justina, Mojan, Nic and Winnie. Thanks everyone! Sadly, a few of them spent a long time doing some great acting and didn’t make it into the short video, but I really appreciate their work. And when they weren’t acting, they were an amazing crew: monitoring sound, running the camera (when I was in front of it), doing set design, giving advice, holding DIY reflectors… even some costume design! Check it out.
The reflector you see in the first photo is just a piece of cardboard with a coating of spray adhesive and aluminum foil. The dull side of the aluminum foil turned out more useful than the shiny side, but I had both just in case. This light stopped the shadow on the side of my face from being too dark in the video; but we learned that too much light can make the scene look unnatural as well. If you can, bring a laptop with you and check the test footage on set.
Will lent me the greatest camera I’ve ever held, but loads of DSLRs these days will shoot video and if you can’t borrow one, look for a rental. I rented some audio gear so that wind noises wouldn’t be louder than me, which really pays off. That fancy gear as operated by my very talented friends pushed the production value through the roof, but this was really all done on a shoestring budget. The zeppelin (the fuzzy thing covering the microphone), boom pole, microphone and portable sound recorder were rented for a total of around $60. Totally worth it.
I’ve never done product photography before, but thankfully there was enough instruction online to get me most of the way, and Anthony answered my photography questions. Here’s what the scene in my apartment looked like:
Those are just clip-on light fixtures from the hardware store fitted (temporarily!) with PAR-20 halogen flood lights; with one mounted on my microphone stand and the other on Anthony’s tripod. The idea is to have a well-lit background reflecting light—after some adjustments, you’ll get it to the point that there aren’t harsh shadows around your subject and it’s picture-taking time. Foreground lighting is provided by a halogen desk lamp diffused through tissue paper.
I used aperture priority (Av on the camera dial) with the smallest possible aperture (biggest number) to maximize how much is in focus. I adjusted the exposure so that everything looked bright without having blown out (ie, completely white with no detail) areas of the photo—cameras have a feature to highlight over-exposed areas which comes in handy here, check your manual for how to turn this on.
The resulting picture only needed a little bit of white balance correction; Gimp‘s automatic white balance feature did a great job for that with no effort on my part.
The tools for making a good-looking video are cheaper and more accessible than ever. I hope this gives you some ideas and inspiration for your own project. Let me know if you find this post helpful when putting a video together; I want to see the video you make! Good luck.
I got the printed circuit boards (PCBs, in electronics speak) mid-week for my production prototypes, and finally soldered everything yesterday.
For the prototypes, I picked the bigger packages of components because I would be soldering this board by hand. Eventually, the production version will use smaller parts to keep the board size down, and they’ll be machine-assembled so hand-soldering won’t be a concern. However, the charger IC (integrated circuit, that’s a chip) I had my sights on only comes in the minute WSON package. I kid you not, that acronym stands for Very Very Thin Small Outline No Lead Package. That’s no typo, there are two Verys in the W of that acronym. The gap between pins on a WSON package is 0.25mm (for reference, the gap between pins on the microcontroller is 3.75 times wider) plus the pins don’t extend beyond the chip body… and the anticipated difficulties soldering something so small by hand is why I ordered a couple of extras. I’ll admit that’s a lesson I learned the hard way; a few years ago I worked on a project and ordered my normal extras rate for a WSON part, then and had to order a few more later. You can see this tiny charger IC in the next picture, to the left of the USB connection at the bottom left side of the photo.
After a few attempts at soldering the charger IC that didn’t work out, I managed to populate some working boards. Here you can see an orange light on the finished board, as it charges a lithium polymer cell from the micro USB port on the board. That light changes to green when charging is done.
One of my favorite parts of this circuit board isn’t even functional—the other side has the Sparrowscope logo on the copper layer. 99% of Sparrowscope owners will never see that on their Sparrowscope, but I like that it’s there.
I’ve been busy lately preparing a production model of the Sparrowscope. A production model not only functions in the same way as the final product, but it’s made in the same (or close to the same) way as the final version will be as mass-produced, from the same materials. I have several components being machined, laser-cut, bent, welded, and 3D printed, all of which will make it to this blog soon.
Getting to this stage has taken a lot of testing, in approximate calculations, simulations and physical models. It’s possible today to—without leaving a computer—design a part, specify the forces acting on that part, and measure the stresses and deformations. But only a part of the academic engineering world thinks that physical prototyping is history. I think Boeing even changed their mind about this one. There are all sorts of things you find out when you put all of the components together in real life, which you may have missed on a computer. Sure, there are some simple products where our ‘virtual prototyping’ capabilities are sufficient, but I’ve certainly learned lessons and made improvements to the Sparrowscope by going to the shop and making things.
I had to pick a gauge of aluminum for main structural components in the Sparrowscope. Running a few simulations was a painless way to get a sense of an acceptable range, given a multiple (that’s the safety margin) of real-world forces. I knew these forces not from simulations, but from actual measurements in experimental tests. But rather than try to simulate all of the possibilities of loading the part, I took a high-strain area and roughly manufactured it in a few different metal gauges. That gave me a real sense of how stiff this part feels. I didn’t need perfect test pieces, I scribed a segment of the design onto each sheet of aluminum by tracing a pattern, cut them out with tin snips, and took then to a sheet metal brake to make the test parts.
Now of course, to anyone who has dealt with sheet metal, you know there’s a bit of a trick here. You can’t make these three bends together on the sheet metal brake I’ve shown; the two bends in one direction would put metal ‘in the way’ for doing the outward bend. The last bend of the tapered segment was done by hand in a vice. Like I said, it didn’t need to be perfect, rough edges were fine. I just needed approximate versions of these shapes to play around with. It’s like a sketch, but made from sheet metal.
The last picture shows the orientation of this segment in the actual design, but I realize that writing this doesn’t explain how the part works… soon I’ll have pictures to show yo how the whole Sparrowscope comes together! I’ll also talk a little bit more about prototyping in a future post.