A digitally controlled KAP rig, chapter 1
I like those really small, simple KAP rigs. Really I do. The problem is they just won't do what I need.
When I started in KAP, I had a goal in mind. I wanted a panoramic aerial shot of Old Lyme Shores. This is a private beach in Connecticut my family has been visiting since my father was a child. A few years ago, I made a panoramic photo of the beachfront from the ground that was popular with the OLS community. Now I want to do the same thing from a KAP perspective.
So last summer, I built my first KAP rig to attempt a shot of the beach but the winds didn't cooperate. (And I didn't have enough kite.) Before I got a chance to fly my rig, it was time to go back home. I didn't want to quit there, of course. I got a bigger kite and successfully flew my rig many times. I figured it was all good practice for my next chance to shoot Old Lyme Shores.
But practicing with my rig was revealing some problems. The controller on my rig was pre-programmed to shoot overlapping images of a wide area. I planned to stitch these together with Photoshop or other software to produce the final image. But the constant swinging of the suspended rig was wreaking havoc with the camera's aim. The photos didn't overlap perfectly. There were always gaps in the final merged images.
So I gave up on panoramas for a while. Instead I adjusted my technique to try to get better single-frame shots. By changing from telephoto to wide-angle settings I was able to improve my yield a bit. Also I found I got more good shots in a vertical orientation so I adjusted the camera cradle to shoot more downward aiming angles. Finally, I learned to set the exposure controls properly before each KAP flight.
But I still hadn't solved the problem of getting my panoramic shot.
Other KAPers were creating panoramic shots with their R/C controlled rigs. So I decided to try to incorporate an R/C receiver into my rig. But I still didn't want to jettison the automated controller. I wanted the R/C receiver to be an input to the microprocessor which would then control the servos. With that setup I could do both manual shots and sequenced shots.
The main problem with combining the two would be getting the 360 pan servo working. I had resisted modifying a servo for continuous rotation because that defeats the feedback mechanism of the servo. In effect it creates an open-loop control system. This means it may suffer from drift on the pan axis. For a manually controlled system this is just inconvenient but for an automated one, it would be a huge flaw. But a solution to this was sitting under my nose -- we had a surplus digital compass module in our lab at work. If I used the compass as a heading reference, I could close the feedback loop and eliminate the drift. And when operated in automatic mode, the controller would shoot in the intended direction regardless of shifting wind direction or a swaying rig. Hey, that means not only
can I use a servo with continuous rotation, I can eliminate the gaps in my sequenced shots. Eureka!
After a few days consideration, I realized, though, that I wouldn't be able to fit the compass and R/C receivers on my old rig. Nor would I be able to modify it for 360 degree operation on the pan servo. The old rig's design was too rigid. Not only that, it was starting to show signs of weakness (literally). It was time for a new rig -- one that incorporated all the lessons I learned from the first rig and from studying other's designs.
So the new rig would have to be different in many key respects. These are some of its requirements:
* Use continuous-rotation servo for pan axis.
* Incorporate digital compass and implement closed-loop control for pan axis.
* Able to accomodate a digital SLR camera.
* Wider range of motion: 360 degree pan, 45 to -135 degree tilt, -135 to +135 degree left-right motion.
* R/C manual operation in addition to full-auto and semi-auto shooting modes.
So those were my requirements to start from. With this rig I believe I can get the beach panorama I've been gunning for as well as many other interesting KAP shots.
Stay tuned for more since design and construction have now begun. There are many more interesting aspects of this rig I would like to share.
When I started in KAP, I had a goal in mind. I wanted a panoramic aerial shot of Old Lyme Shores. This is a private beach in Connecticut my family has been visiting since my father was a child. A few years ago, I made a panoramic photo of the beachfront from the ground that was popular with the OLS community. Now I want to do the same thing from a KAP perspective.
So last summer, I built my first KAP rig to attempt a shot of the beach but the winds didn't cooperate. (And I didn't have enough kite.) Before I got a chance to fly my rig, it was time to go back home. I didn't want to quit there, of course. I got a bigger kite and successfully flew my rig many times. I figured it was all good practice for my next chance to shoot Old Lyme Shores.
But practicing with my rig was revealing some problems. The controller on my rig was pre-programmed to shoot overlapping images of a wide area. I planned to stitch these together with Photoshop or other software to produce the final image. But the constant swinging of the suspended rig was wreaking havoc with the camera's aim. The photos didn't overlap perfectly. There were always gaps in the final merged images.
So I gave up on panoramas for a while. Instead I adjusted my technique to try to get better single-frame shots. By changing from telephoto to wide-angle settings I was able to improve my yield a bit. Also I found I got more good shots in a vertical orientation so I adjusted the camera cradle to shoot more downward aiming angles. Finally, I learned to set the exposure controls properly before each KAP flight.
But I still hadn't solved the problem of getting my panoramic shot.
Other KAPers were creating panoramic shots with their R/C controlled rigs. So I decided to try to incorporate an R/C receiver into my rig. But I still didn't want to jettison the automated controller. I wanted the R/C receiver to be an input to the microprocessor which would then control the servos. With that setup I could do both manual shots and sequenced shots.
The main problem with combining the two would be getting the 360 pan servo working. I had resisted modifying a servo for continuous rotation because that defeats the feedback mechanism of the servo. In effect it creates an open-loop control system. This means it may suffer from drift on the pan axis. For a manually controlled system this is just inconvenient but for an automated one, it would be a huge flaw. But a solution to this was sitting under my nose -- we had a surplus digital compass module in our lab at work. If I used the compass as a heading reference, I could close the feedback loop and eliminate the drift. And when operated in automatic mode, the controller would shoot in the intended direction regardless of shifting wind direction or a swaying rig. Hey, that means not only
can I use a servo with continuous rotation, I can eliminate the gaps in my sequenced shots. Eureka!
After a few days consideration, I realized, though, that I wouldn't be able to fit the compass and R/C receivers on my old rig. Nor would I be able to modify it for 360 degree operation on the pan servo. The old rig's design was too rigid. Not only that, it was starting to show signs of weakness (literally). It was time for a new rig -- one that incorporated all the lessons I learned from the first rig and from studying other's designs.
So the new rig would have to be different in many key respects. These are some of its requirements:
* Use continuous-rotation servo for pan axis.
* Incorporate digital compass and implement closed-loop control for pan axis.
* Able to accomodate a digital SLR camera.
* Wider range of motion: 360 degree pan, 45 to -135 degree tilt, -135 to +135 degree left-right motion.
* R/C manual operation in addition to full-auto and semi-auto shooting modes.
So those were my requirements to start from. With this rig I believe I can get the beach panorama I've been gunning for as well as many other interesting KAP shots.
Stay tuned for more since design and construction have now begun. There are many more interesting aspects of this rig I would like to share.
posted by Dave at 8:02 PM
0 Comments:
Post a Comment
<< Home