Live data from Hacker News

Soloshot: Automatically film outdoor activities without a camera operator

shop.soloshot.com

101–110 of 175 posts

Re: Soloshot: Automatically film outdoor activities without a camera operator

#102

Earlier quoted context omitted.

Shouldn't the camera let you track exactly where it is so you can find it?

Tracking might not work when it's ten feet underwater..

Less actually.

It will loose tracking on a duck dive and on wipe outs. It even has 'barrel mode' so that the camera will keep tracking on your vector should the signal cut out while you're shielded by the wave lip. Haven't had much use for that feature :(

Re: Soloshot: Automatically film outdoor activities without a camera operator

#103
As a hobby, I drive racecars. I have been thinking about something like this near the track to video me going by. (Far away vantage points look boring.)

It seems like for much less than $500 you could build an RPI and just track moving objects via optical flow (assuming you are alone on the track.)

Re: Soloshot: Automatically film outdoor activities without a camera operator

#104

This is badass. As an athlete whos heavily into extreme sports I'd buy one in a heartbeat. I think you've got some JavaScript issues going on with your page, though, It keep jumping to the top of the page when I'm scrolling down.

We kitesurf and use GoPros, often on tripods on the beach. People come up and take them as soon as we're more than 200m away. It's quite easy for us to spot them, confront them and get it back. But it happens all the time No. 1 rule - don't leave anything of value on the beach.

So next we need to add some stabilization to this platform and then you can just stick it out on a bouy ^_^

Re: Soloshot: Automatically film outdoor activities without a camera operator

#105

Would it be possible to mimic these features in post prod? Essentially gather a giant high res shot, using the money spent on fancy feature on higher res. Then run the same algorithms in post to get a similar video?

Not really. The level of resolution necessary is an obstacle. Additionally, having a rotating tripod mount as part of the flow means the angle of view is changing, as opposed to a static angle you describe. That said, it could be faked, but wouldn't be the same.

Makes sense. Thanks!

Re: Soloshot: Automatically film outdoor activities without a camera operator

#106
Awesome, I always wanted to make one of these. I had the idea about 10 years ago, when I was really into making stunt videos and timelapses as a teenager. This was way before Raspberry Pis and Arduinos, so I think it would be a lot easier today. Of course there's a huge difference between a hobby project, and a commercial product. But watching all of those Youtube videos was like seeing my idea come to life.

Re: Soloshot: Automatically film outdoor activities without a camera operator

#108
It's very cool, but the demo videos suggest that the tracking isn't ideal yet. Which susprises me, because we only have what, 50+ years of R&D with tracking moving targets. As in, if you can figure how how fast the target is moving across the field of view, you should be able to predict where it is going to be in the next frame, PLUS, take into account the latency of the Soloshot motor, and calculate that all together in real-time such that the target stays as close as possible to the center of the field of view. Right now, based on the demo videos, it appears that either the CPU is slow, or the Soloshot motor is slow, or both, with the result being that the target gets ahead of the Soloshot and gets to the edge of the field of view regularly, making for less than optimal viewing afterward.

Seems like a totally, well, 99% fixable problem, and given Soloshot's been out for a while I'm surprised the demos aren't more stellar.

Still, it's impressive and I'm surprised GoPro hasn't acquired the company already.

Re: Soloshot: Automatically film outdoor activities without a camera operator

#110

Like many ideas I have had, I'm both happy and sad. I had this idea a good 5 years ago. Kicking myself, but then again, I' not a hardware guy.

I had the idea too, maybe even 10 years ago. But don't worry, it's not like you narrowly missed winning the lottery. Building a product like this takes years of dedication and hard work.
Post reply on HN