I enjoy my video content. Also that of my cycling trips. Truth is: you don’t enjoy it so very much. There are many problems with it, from storytelling to editing. For one major issue, I should quote the girl.ex (the girl who was the girl until two months ago):
You know, it would be nicer if I wouldn’t stare at streets and booties only for five minutes of video.
And that’s a fair point. So, along with my new GoPro Hero 6, I got a selfie stick. Then I went out on a winter ride in the vicinity of Bozen. This is about the best that I got out of it from a solo ride. The video starts after my lengthy introduction with actual footage from on the road.
Arguably the most atmospheric scenes of this short clip are those that I made with the selfie stick. The stuff in between is filling material. There’s always lots of filling material, but little of catchy footage. For one, that’s because I never enter these rides with a game plan. But eventually you get only so much out of a front-mounted GoPro.
A selfie-stick is difficult to control, still rather static, and not safe
With GoPro’s inbuilt stabilization, I can limit shaky footage. The video looks rather smooth. I cannot adjust the length of the selfie stick without stopping again (or riding free-handed). I cannot really see if I’m holding the camera correctly at all times. Hence, the selfie-stick is great to get some more full-body shots of me, but it requires me to stop to being and end recording a clip. In between, one of my hands will be more busy with holding and controlling the selfie-stick than with holding the bike. Since footage is most interesting in more technical situations, selfie-stick effectively distract and therefore impair safety.
It also looks really, really silly to ride with a selfie stick. Too often, I find myself to skip the next opportunity at interesting, but populated places in favor of just taking a random clip of filling material.
360° cameras are a broken promise
At the end of 2017, Facebook made me stumble over a new kind of action cam: the Rylo 360 cam. It’s not that I had not been aware of 360° cameras before, but I’m personally not interested in spherical shots like the one below.
I find 360 videos in which I can move the camera interesting, but since my primary mode of recording is a flat video for YouTube, the use scenarios of those always seemed too limited to justify the expense. (But I don’t understand why real estate agents have not adopted that already widely.) And they used to be terribly shaky.
Then Rylo entered the scene with two promising features: world-class image stabilization and re-framed shots that would just extract a flat image from a 360 lens – but with the neat trick that this flat image could follow a dynamic path and move around the entire sphere. (It’s called Overcapture by some manufacturers and Free Capture by others.) With the tap of a button in the app (during editing), the clip could fixate a person, a house, a traffic light, or whatever, and stay focused on that while the camera continued moving. Enjoy both in the next two clips.
Competitors soon followed. Within a few months, stabilization quality in 360 cameras has skyrocketed, and all major competitors are offering similar editing modes. And I was thinking: isn’t that the solution to more captivating shots, with more versatility?
Well, no. Or at least: not really. It works great in sports during which you can hold a selfie-stick. In road cycling, cameras often end up on handlebars. And then it’s going to look something like this.
Bare with me, but I don’t think that’s much improvement over what I did. In way too many angles, the bike is in the way. You still end up in a situation in which you need to take the camera in your hand. It has a lot more potential when you ride in a group. Even then, you’ll end up with things being in the way.
Fine, you can argue, but at least you get more options while holding it in your hand, and you get a bit more out of it even without taking it into your hands at all. But then (at least to this date) this is on the expense of lesser video quality and smaller frame rates. Before you add editing complexity and rendering times.
If I shoot 360 to then use things like Overcapture etc. (which most of consumers will eventually want to do), then processing and cutting the final video will have to go down in time and in complexity. Basically enough to run on a normal notebook. That’s one more reason why 1080p video is so popular and why people want 4K, but won’t use it. You don’t need a dedicated graphic card and an i7 processor to create videos, and rendering doesn’t take hours. Some say that modern smartphones have already reached the capability to render clips smoothly. True. Smartphone editing works nicely for single clips, but not so much for larger projects. From my perspective, the future should not exclusively be in having to move the files physically from the camera to the phone, then again having to move them to the desktop (or notebook). For larger projects, a better workflow is to move them to the PC (or insert the SD card), pair phone and desktop, and then via an app use the phone only as a controller to find the optimal viewing angle (alike to what Insta360 is offering), but do the rendering on the PC.
Drones are slowly coming of age, but they need a diet
360 cameras are but one recent breakthrough in camera technology. The other ones are drones. Here’s a quote of DC Rainmaker at the beginning of his 2017 review of the DJI Spark camera drone.
I’m always on the lookout for a better way to capture cool footage while out running, cycling, swimming, hiking, or doing whatever the heck it is I’m doing. Be that a better GoPro or a better way to get a unique shot, there’s always room for improvement.
Over the past few years we’ve seen drones become one way for folks in sport to get those shots, either autonomously or manually. Each has their place in life, though I eagerly await the day and product that can quietly, safely, and cinematically pull together an automated edit/movie of an epic downhill mountain bike ride or skiing run. But we are getting closer and closer to that.
Part of those requirements is getting a unit that’s small enough to be easy to carry in any scenario, and while the DJI Mavic last year seemed to fit that bill – it’s got nothing on the Spark this year from a portability standpoint. Thus I was eager to put it through its paces see how well the automated tracking modes (theoretically ideal for sport) handled here compared to that. Plus, I wanted to see how it handled in general when taking it with me for a ride just as an extra long selfie-stick of sorts.
And then, still, the DJI Spark did not live up to the needs of everyday filming for cyclists.
I’ve seen a few drones passing by. Many of them on Indiegogo. None of them was transparent with respect to their actual performance. I’ve made bad experiences particularly with Indiegogo-based crowdfunding projects, so I didn’t there jumping into any of them. The one I was most seriously thinking about funding indeed got stopped by Indiegogo before the end of the campaign. And in any case: standard features of pocket drones would include about 10min of flight time, 1080p/30fps video quality, and at most semi-autonomous flight. Particularly the last point is a hassle for using drones in cycling.
Then came the Skydio R1 drone.
That video is a consumer shot, not a marketing clip. It has brought drone videos to a whole new level of autonomy. The main feature is it’s world-class obstacle avoidance. To that end, 12 of the 13 cameras in the drone do nothing but recognize objects in the path of the drone.
Unfortunately, they are huge. They work nicely if they are more or less the only thing that you put in a backpack while you reach your filming location. Take them on a mult-day trip, something like a Transalp? No way. And that doesn’t even mention the hefty $2500 price tag on the Skydio. It’s five times as much as I’d be comfortable to pay.
Now, this big format has a good reason. Drones consume a lot of electricity. Small drones will have shorter flight times. Drones are exposed to wind. Small drones have inherently more unstable flight patterns. Drones occasionally crash. Small drones tend to be less sturdy. The Skydio drone carries technology for 13 cameras. Small drones can support fewer internal electronics. (Think space, but think also overheating.)
What should the future hold?
Usability. The future needs to focus on usability. For action cams, 360 cams, and video drones, usability is the key to keep them using for longer and more often – and to consider upgrades to more advanced products. With action cams, we’ve now reached video quality that exceeds the needs and demands of most consumers, but they are stuck to boring angles. They will become 360 cams eventually. 360 cams give some freedom back to the movie maker, but they suffer from mediocre video quality, less robustness, and clunky editing experiences. Drones finally start to get somewhat decent in in-flight operation (or great, but then you can’t afford them), but they are too big and too bad at filming. So, what needs to change?
Bring fast 360/4K rendering to desktops
I’ve cut a video. I’m willing to let it render for 30 minutes. Maybe one hour. That’s about it. Computers get faster. But video quality gets better, and that requires more resources. Overall, both computer performance and computer demands develop in parallel. Maybe distributed rendering using cloud services could be an option. But in any case: rendering needs a speed boost.
Make 360 editing on desktops a natural experience
One neat feature of the Insta360 One camera is how exactly it uses the smartphone for editing. Basically, you use the phone screen as a lens into your 360° world. The extracted clip will show the image from where you hold your phone. When you move the phone, you change the camera angle.
Desktop editing instead is a mouse-based experience. By it’s very nature, that’s not as natural. So, think about an integrated editing software which can use the smartphone as an external editing controller.
Something needs to happen with the selfie-stick
Other than that they selfie-sticks are embarrassing, they are somewhat useful. But switching the camera from the handlebar mount to the selfie-stick: Boy, that’s cumbersome!
Now, imagine a mount that is a selfie-stick by itself. Something which includes the possibility to increase the distance between camera and rider. Something that can swivel and tilt around a pivoting point. Something that allows me to keep riding with both hands at the handlebar. Something that I can remove fast for more flexible shots when I need to. Something that’s just a plain handlebar mount when I’m fine with that.
Bring contemporary 360 video to drones
The Skydio has 13 cameras to be able to see everything around it. I’m wondering if you can achieve worse, but sufficient performance with fewer cameras. Could high-resolution 360 video be enough for a consumer level camera?
Even if not: there’s more ease in flying a drone that you just keep stable at a certain level and distance if you don’t also need to worry if the camera is pointed in the right direction. Stabilization similar to that of Rylo and Insta360 One would also help to overcome that problem of shaky drone videos when wind catches the drone. 360 for drones is an obvious step.
Develop obstacle avoidance for the jersey pocket
From my perspective, this is a deal maker for drones. A camera with half the performance of the Skydio that weighs no more than 200g and fits in my jersey pocket and/or in a small backpack that is filled with other gear as well: I’d take it immediately. (If it appears on Indiegogo, at least I’d seriously consider it despite my bad experiences.)
What’s your point?
Anyone could come up with fantastic fantasies of future features. So, why do I write about it? Honestly, because I did. At least I hope that some guys who blog about drones and 360 cams read along and keep the usual use case in mind. Mind-boggling technology deserves reviews of praise, but we do need to keep an eye on usability.
Maybe it’s also just a very long post to excuse myself for my contributions to the pile of boring bike videos? I simply haven’t found the technological solution yet that I can take along on a solo ride where it’s still riding first, not filming first. So, I apologize. Mea culpa. And I promise that I’ll keep my eyes open for the product. I’ll be back with another post.