Russell Brown’s recent foray into quadracopters (darn, now everyone is going to try what I was playing with) raised the drone volume in my In Box. But one email was particularly on point. What is the ultimate camera for a drone, and how do we control it? Here’s how one commenter put it:
It should be light enough to get good air time, and perhaps enable me to save a few thousand by not needing super-heavy lift. It should have a good radio communications ability and full functionality from the ground. I should be able to operate / tether it through a remote control body that acts like the camera but has no sensor and has an EVF and the normal controls one would expect on an interchangeable lens camera. It should integrate well with controls for the camera platform so that moving the clone camera moves the platform. The camera needs no display, no viewfinder, and must be weather-resistant. APS-C is fine, though FF would work as well. Much of the weight savings are through what is not needed. It does not need a heavy battery so long as the battery life of the drone is the limiting factor. It needs no buttons, or dials other than on/off because all control is exercised remotely. This also helps constrain cost and weight.
The camera is not limited to drone use, but could be used for any remote application with the remote control body. It should also be designed with the idea of an open architecture software solution to a smart phone, tablet, or computer, or even full integration into the remote flight system. Let the end user choose how they want to control the camera.
The camera makers don’t see the world like users do. This user sees something he needs (and I would argue a number of us could use): a simple "sensor in a box with a lens mount" that communicates to a control system. Technically, the control system doesn’t need to be a camera body without sensor as the person here indicates, but could be just a set of control wings that mount onto your smartphone with the smartphone providing the communications to the remote box and a view from its sensor.
We can actually do bits and pieces of this today by kludging things together. For instance, I’ve been using CamRanger with my Nikon DSLRs to do something similar. But I don’t need all the features (and size and weight) of the DSLR body, I’ve got multiple gizmos and a few wires to deal with, and I could use a more integrated control with slightly tighter timing, too. So it would be really nice to have a truly integrated and well thought out camera/controller system, for sure.
The minute I read this person’s email I thought of Blackmagic Design and the Pocket Cinema Camera (PCC). Make it m4/3 sensor size (the PCC is smaller than m4/3 and a crop), configure everything for stills, make the remote view/control really tight, and put it in a box that’s designed to be mounted in all kinds of ways (note that the PCC, for example, has tripod sockets on both top and bottom, but most of us want a full cage mount around the box).
There’s an opportunity here, and it’s not a difficult one. Do we need autofocus? Probably not. Most aerial and remote setups we’d use this box for are either far enough away from the closest thing (the ground) that you can just use fixed focus set right. We’re probably not going to put zoom lenses on these things, either, as that’s just size and weight we don’t need or want. Given the lens sets that are out there, I’d build this box as m4/3 or maybe Sony E-mount. Consider a box not much bigger than a GoPro but with m4/3 primes being mounted out front as need be. That box communicates back to the second component that plugs into your smartphone for display. I suppose it’s possible that you could just write a straight smartphone app and not do the second component, but I suspect that there are some timing advantages to doing it all with proprietary gear, and this also means that you can build dedicated systems to place in remote places without tying up your phone (i.e. controller is firing the remote camera box for time-lapse).
Would Canon or Nikon do something like this? Nope. Not until some Silicon Valley entity figures out the opportunity, puts a device on the market and proves it will sell (see GoPro, et.al.). By the time they react and do something themselves it would be too late for them if the startup knew what it was doing. Sony’s discovered how tough GoPro is going to be to dislodge, for example: it’s not just putting a sensor in a waterproof box that’s the opportunity, and GoPro gets the marketing done better, too.
So how about one of the money-losing camera companies? Nope, they don’t see the opportunity to do niche, either. If all you look at is your other six primary competitors, then almost all of what you invent are the same things they others are doing.