Gigapixels and giga headaches

Like Aldo and Ian, I have been working the last 8 months on an automated panohead to make gigapixel panoramas. I had some basic specs in mind, designig it:

- I wanted to have the option to attach the camera both in landscape or portrait position.

- It had to be sturdy enough to carry the 5D with a telephoto lens.

- It had to be operated from a laptop, so I could also use the DSLR Pro software to check the images.

- I wanted a remote zoom function to operate it from a pole.

- It had to be relatively cheap to build.

Here some impressions of the first result:

And here it is on the pole, complete with the first cable spagetti:

It all worked fine, so it was time to put it through the test. I found a great site to make a pano and this was the (HDR)result:

It looks great, doesn't it? But it sucks awfully, for there was not one image that was really sharp. And so I found out that the problem designing a head like this is not in the positioning, nor is it in the software. The main problem is to keep the camera still. There will always be a transmission between the motor and the camera and every transmission will have a surtain play between the wheels. As there is no way to avoid that, I went back to the drawing board and created my first brake. As the pole is moving due to the wind, I also had to find a solution for that.

The first brake was not good enough. It appeared to be very hard to kill any movement of the head. Especially when there is more wind. I recently built a better brake and this one seems to do the job well. Here an example of a pano stitched from 108 portrait images, about 40.000 pixels wide. You can click the image to watch it in HD View. I have scaled it down to 50%, saving myself a lot of upload time.

It is not remotely as spectacular as Aldo's Red Bull pano, which is still the greatest action pano that I have ever seen, but I am pleased with the result. At least it is sharp now and this is a solution that I can sell to make large format prints. I will do a new test with the pole soon, rigging it with kevlar wires. Testing with other wires proved that they stretch too much, still allowing the pole to move a bit. Hopefully the kevlar will keep it still enough to make even long time exposures.

I am completely redesigning the panohead now, for it catches too much wind as it is and the sleeker I can design it, the better it will be. I will also try out an idea for a smaller an even better brake.

To be continued.....