Welcome Inspire Pilots!
Join our free DJI Inspire community today!
Sign up

Anyone else planning to fly a 360Hero array?

Yes, I have been doing 360videos for awhile and want to try it with my new Inspire.

I was thinking of building my own gimbal stabilizer or modify an existing.

I can imagine without a gimbal there would be a lot jello action.

FYI I got a 3d printer and built my own 360 gopro mount
 
  • Like
Reactions: SkySav
Yes, I have been doing 360videos for awhile and want to try it with my new Inspire.

I was thinking of building my own gimbal stabilizer or modify an existing.

I can imagine without a gimbal there would be a lot jello action.

FYI I got a 3d printer and built my own 360 gopro mount

Thanks for your input. I would to learn more about the experiences and check out your work. I have access to a couple people who have 3d printers. Where did you get the model file for the 3d print? Care to share? How many GoPros are in your array? I'm also planning on adding an additional 5.8 FPV cam at the nose of my Inspire.
 
Yes, I have been doing 360videos for awhile and want to try it with my new Inspire.

I was thinking of building my own gimbal stabilizer or modify an existing.

I can imagine without a gimbal there would be a lot jello action.

FYI I got a 3d printer and built my own 360 gopro mount
SWEET, man. Yes, please share your 3D file...I'd definitely give it a try. :)
 
Here is a sample. You may need to install koloreyes if you don't have an html5 browser (Chrome).

https://eyes.kolor.com/video/c93e3c47bd01efd5f564aeb16474454c

These are lower resolution version for mobile viewing. Resolutions limit can exceed 8000x4000 videos, that may need ultra high speed data.

I have been contemplating on uploading my .stl file on http://thingiverse.com but it originally got copied from there.

It's really simple.
-Make 6 Gopro rectangle
holder/box.
-Make your own glue. (ABS scraps and acetone). Make it thick or low viscosity. You'll discover why.
-Arrange the holder in cube and glue in place.

It was printed individually then glued together, because of the complex shape it would have produced.

Anyways my design slightly evolved, but it was really no engineering hack. I may share it when I have time to explain it better.
 
I love the 360 videos I've seen shot from multi rotors. I knew I was going to rig my Inspire 1 after I saw this payload test. Thoughts?

where is the video from the 360? i cant see it being very usefull without the gimbol. just shakey footage that couldnt be used except for home fun not pro???? if wrong love to see it.
 
With a full 360° rig, as long as there isn't too much swinging (that would screw up the perspective) you can do all the stabilization in post production.
 
  • Like
Reactions: SkySav
With a full 360° rig, as long as there isn't too much swinging (that would screw up the perspective) you can do all the stabilization in post production.
I don't think you can do all stabilization in Post. Kolor Video improved on it though.

Why not prevent most of it at the beginning?
 
  • Like
Reactions: SkySav
Well in theory you can, just need to rotate the "virtual" ball. Now whether there is aleady available software that will do it today I have no idea. Obviously best to prevent first if you can.
 
  • Like
Reactions: SkySav
Well in theory you can, just need to rotate the "virtual" ball. Now whether there is aleady available software that will do it today I have no idea. Obviously best to prevent first if you can.
Theory is one thing , reality is another.

Have been doing 360videos since 2010 and have been following its evolution. Never heard of the technique that you speak of.I wonder if the guys at Kolor uses it.

Could you expand the theory a bit more.

Thanks
 
Well... take the Inspire cam. It has a "finite" field of view, so in order to keep that view stable we need to heep it oriented the same way with a gimbal.

Now on a 360° setup you are recording everything around you. Say if your camera ball tilts down 10° during recording, all you need to do to get the original view during playback is tilt the view back up by 10°, which you obviously can.

The Parrot Bebop drone is doing a version of that in realtime, albeit only over 180°. The camera has a 180° fisheye, but the image you see on your phone is a narrower, rectilinear view similar to that on the Inspire. The camera crops/defishes/stabilises live in flight using the IMU info (of course with a relatively low quality).
In post you can use either recorded IMU data or image analysis to do that in full quality.
 
Have you considered lens orientation on a 360 rig with 6 go pro array?
Well... take the Inspire cam. It has a "finite" field of view, so in order to keep that view stable we need to heep it oriented the same way with a gimbal.

Now on a 360° setup you are recording everything around you. Say if your camera ball tilts down 10° during recording, all you need to do to get the original view during playback is tilt the view back up by 10°, which you obviously can.

The Parrot Bebop drone is doing a version of that in realtime, albeit only over 180°. The camera has a 180° fisheye, but the image you see on your phone is a narrower, rectilinear view similar to that on the Inspire. The camera crops/defishes/stabilises live in flight using the IMU info (of course with a relatively low quality).
In post you can use either recorded IMU data or image analysis to do that in full quality.

Your explanation may be relative to a single camera , but more factors are involved. (Lense orientation, perspective, time of moment and/or rolling shutter.

A camera turned sideways rolling shutter/jello is different from a camera that is upright.

Stabilizing a single camera is simple compared to an array of cameras oriented in diverse posit ions while in motion.

Nice try, but I was hoping you had a solution I could exploit.

Thanks again.
 
Well, all of these factors already affect 360° video in the first place. Nothing different.

Do the proecssing the exact same way you would usually, end up with the video like the one you posted on kolor earlier. Then in the same way you'd use the mouse to look around, use the same "scrolling" to compensate camera motion.
 
I think
Well, all of these factors already affect 360° video in the first place. Nothing different.

Do the proecssing the exact same way you would usually, end up with the video like the one you posted on kolor earlier. Then in the same way you'd use the mouse to look around, use the same "scrolling" to compensate camera motion.
That process will produce cropping of the final stitch. 360video producers will know the results would not be desired and how it affects the end results for viewing.

For those people new to 360video, stabilizing videos aren't as easy as regular videos.

Please visit Kolor.Com or video-stitch.Com as they serve the niche for producing 360videos.
 
where is the video from the 360? i cant see it being very usefull without the gimbol. just shakey footage that couldnt be used except for home fun not pro???? if wrong love to see it.
This was just a payload test. I agree that a gimbal or at least a hard point attached arm with dampeners are necessary for decent footage.
 
Not bad. Technology is making progress. The lens is the key in getting good overlaps.

Bubble cam is a bit low on video resolution. Consider dividing horizontal final stitch by 4 you'll see that size does not make viewing a joy compared to go pro arrays.
 
  • Like
Reactions: stevet
Not bad. Technology is making progress. The lens is the key in getting good overlaps.

Bubble cam is a bit low on video resolution. Consider dividing horizontal final stitch by 4 you'll see that size does not make viewing a joy compared to go pro arrays.

Yea the quality probably wont be the best but its good to see some smaller cameras that some drones can carry based upon the weight. I might to go NAB this year and see what other sorts of 360 cameras are coming and make some connections :)
 
  • Like
Reactions: SkySav

Members online

No members online now.

Forum statistics

Threads
22,277
Messages
210,655
Members
34,325
Latest member
LewisuhaAbips