The actual bandwidth is 500kbps, or about 350kbps usable after FEC. I'm encoding video at 640x480, 12FPS. It's almost entirely custom software. I'm using a Gstreamer pipeline to take in the video stream, encode it, and encapsulate it. From there, I packetize it, run it through an LDPC encoder (for error correction), data whitening, and finally send it to an FSK modem to convert it into radio waves. On the ground station I basically do the same thing in reverse.
At 1W transmission, I've gotten video over 140km from a balloon to the ground. Most of that distance is horizontal - the winds tend to blow the balloon quite far. The math says the system should be good to at least 250km.
Our payload has no stabilization either, and in past launches we've had quite a bit of rotation and swinging. Here's a video of one of last launch, for reference: https://peertube.scd31.com/w/7CQCYB4BmJzngoZTiMociY
To point the camera we have a lot of clever software and hardware. I didn't have any part in it, so I'll try to explain it as best I can. There's a diagram here, which hopefully will help you to follow along: https://eclipseplus.ca/Project_Details/Payload_Design/
We have a fixed camera mounted above a mirror. The mirror is on a special gimbal, which can be maneuvered using two servos and a bit of math. For coarse aiming, we have an IMU on board which uses the magnetic field of the Earth to figure out its orientation. For fine aiming, we do a bit of image processing to try and center the sun in the field of view of the camera. It's not perfect, but since we're only capturing still images, it's okay if not all the images have the sun in view.
Also of note is the filter, which starts in-place (to protect the optics) and automatically moves out of the way during totality. It also moves back into place afterwards so that we can continue safely taking images after the eclipse.
The balloons on the landing page are accurate! Those pictures are from past launches, which used the same payload (with some changes between each launch)
Absolutely. I've never experienced a total eclipse before but I hear the difference between 95% and 100% is stark; due to the logarithmic nature of the human eye. I'm not even sure if 95% coverage would even be noticeable without eclipse glasses.
I'm from NS myself (Halifax) and have been trying to convince my parents to drive up to NB to see the eclipse. I'll be home for Easter so that'll probably be my last attempt!
As someone who's paranoid about missing the eclipse due to cloudy weather, I've been working with a few people to launch a high altitude balloon during the eclipse: https://eclipseplus.ca/
We have several cameras on board, including one that points at the sun and one that takes video. It all gets live-streamed to the ground (actually, that's my contribution to the project - the communications system) and sent off to YouTube.
We're launching from New Brunswick, Canada, and we'll be going up to about 30km, or 100k feet. So the clouds definitely won't be an issue for us.