I know that gimbals are really cool and make great things possible, but I just can't wait for them to die. They are heavy, draggy and add complexity. It would be great to just record everything and then scroll to the portion on video you want to see. If you are capturing video for a production, you can fix it in post. If you are doing surveillance, you could completely stabilize your view with software. Maybe this will get us one step closer:
http://bit.ly/1gV7n2c
Comments
Thanks Joe! Very cool!
https://www.kickstarter.com/projects/1307511016/centr-interactive-p...
@Ilya: I like your remark about the available photons and lens size. Makes it very clear that capturing all and everything comes at a cost.
Another limitation - apart from available light - is communication bandwidth. Assume you want to use your drone for inspection / video / FPV / etc. and have a live view of your on-board camera on the ground. With a gimbal you can choose what you want to see. With a "360 camera" you'd have to transfer the full video. That's a lot of data.
Well, on second thought, you could also tell (via telemetry) an on-board computer which view / frame to transmit to the ground. And still have the full 360 video after you land for detailed analysis / inspection / fun.
Let's see what future brings...
@R_Lefebvre, what you are describing sounds like a compound eye of an insect.
Also, I don't think that anyone has mentioned this, but the human eye essentially has a gimbal system (it even "rolls" when we tilt our head from side to side). It's amazing when you realize that nature has already "experimented" with all the possible approaches =)
@R_Lefebvre:
The closest thing available to hobbyists is the Panono, but it currently only takes stills. I can't imagine the CPU that can capture video at 108 megapixels and still launch in my Bixler.
Rob - Yeah, his type of camera really needs a Curvilinear Image Sensor.
But as you say John, low noise, high contrast and low light performance calls for more glass and a larger sensor with bigger photo buckets/diodes.
What if they make a spherical imager chip? Wouldn't that solve the problem of fisheye lenses? Thing is in that case... I'm not sure how a lens would work, to focus light on the chip.
Then instead of that, an array of non-fisheye lenses, and flat chips. With overlapping viewing angles. Then you combine them in software, and allow the user to pick the framing from the combined image.
Joe, as always price is the biggest decider. You can make incredibly sharp lenses, but they quickly become bulky and typically cost more then your entire copter system. Same with high resolution sensors. More pixels means more processing, thermal challenges etc. adding cost at every step.
I think a good hobby grade acceptable camera might be available in 5 years. But a true 1080p ready easily integrated so you cannot see the quaility difference in cinema movies? No way. The industry is just now moving away from tracks cranes and dollies! They are just now getting gimbals good enough to carry the cameras required to take the raw data for movies. And it still takes a very skilled operator. Even the Gopro is mediocre at best. Some tv shows use them, but the quaility difference is immedialtly noticable. For youtube and most online uses they are for sure one of the best bangs for the buck. If they could every movie producing house would much rather use a $399 gopro rather than a 100,000 plus camera that requires highly trained operators. Each time you "correct" a picture with software you degrade the image. In order to get a good end product image if you are going to use software for correction you need a very very very high quality sensor. And it isnt going to fit in that camera anytime soon.
-
1
-
2
-
3
of 3 Next