360° Filmmaking Challenges

360° Filmmaking Challenges

 

I'll be the first to admit it, I am not a filmmaker. Truth be known, I have a B.S. in Marketing, and absolutely no formal training in video editing, script writing, graphics creation or filmmaking. I've just played with editing for the past 10 years, and although I can say I am pretty happy with the edits I did for ATV racing, they are basically music videos with random "cool shots" from the race edited together.

So when I decided to make my first short film, and do it in 360° video, I realized just how much learning I would have to do.  Just the basics would include how to create a compelling storyline, how to compose a shot list, legal issues such as releases and copyrights, as well as the technology behind it all, your camera selection and editing platform.  Oy Vey!  That's a LOT of information to learn!

But I was inspired during a dive trip to Utila, Honduras, and I had to do something.  You see, this particular trip was amazing in every way, the people, the diving, the culture and the island.  And on one particular dive we came across so much plastic trash floating, it actually saddened me right there.  Later in the trip I would see trash scattered all over the north beaches, and that just reinforced it.  Here I was, in tropical paradise, and there was all this trash where I couldn't otherwise see any signs of human existence.

So, I knew I wanted to do something, but had no idea what, or how to do it.  Then I saw a great 360° short film called Valen's Reef which is about some islanders in the Pacific who take back control of their reefs and change the fishing practices from commercial to more sustainable methods.  And I had just begun to play with 360° video in order to show my friends the wonders of the reefs and diving.  So it hit me!  Make a short film about plastics pollution in the ocean, and use Utila as the site, but do it with that same twist, use 360° video to tell the story.

When I first began to research hardware, there weren't many options out, certainly none of the "all-in-one" rigs were released yet, and I had the additional requirement of wanting to be able to film 360° underwater as well as on land.  The basic premise was that you purchased 6 GoPro style cameras and put them in a cube array.  Supposedly adding a 7th camera was going to help with underwater stitching (more on that later). 
 
Well, I found a 3D printed, 7 camera mount on eBay for under $200.  Hmm, I had seen SJCam's offerings with the SJ4000 and SJ5000 cameras, and they certainly weren't the worst GoPro knock-offs, they actually had knock-off's being made of the SJCam cameras! (Ironic, as the SJCams are GoPro knockoffs).  At $40-$60 ea (getting refurbished cameras from B&H Photo) they were a steal compared to $399 per GoPro Hero 4 Silver.  The cameras had 4:3 aspect ratio settings, so they seemed like they could be used for 360° video, but after purchasing 7 cameras I learned that their FOV (Field Of Vision) was too small, and there were literally gaps between the areas that needed to be stitched.  Nothing I did, short of getting a rig that used more cameras, and buying many more SJCams, was going to make those cameras work and get me a proper stitch.
 
Enter 6 more GoPro Hero4 Silver cameras (I already owned 1).  Now we're cooking with gas... I think...  I have no issues getting proper stitches above ground, on dry land, using the 7 GoPro cameras.  I make a few practice recordings using a monopod, and manage to get successful stitches.  I'm feeling pretty good about this setup, decent quality cameras, and although the 3D printed rig is rather fragile, that's ok for now.  So I decide to take the rig scuba diving in Cozumel, that's the primary reason for it anyway, and I fail to get a good stitch, and I have TONS of issues with parallax, no matter what I did.The best footage I got was when I left the camera on a monopod, stationary, while we were snorkeling, and even then there are many instances of parallax being a problem.  To the best of my knowledge and research, it seems like you will only get a good stitch underwater when you use domes.  Older rigs like the 360Rize from 360 Heros, as well as newer rigs like the Abyss V4, all share one thing in common, they all have domes on their GoPro cameras.  And domes aren't readily available, and the Abyss is quite expensive, so this doesn't work well for me.  7 GoPro's, and I can only get 360 video when the rig isn't underwater.  Anybody want to donate an Abyss for me to use?
 
Next on the list to try is a pair of Kodak SP360 4K cameras.  After some preliminary tests in the koi tanks at work, it seems like these cameras hold some promise!  They have a dual underwater case designed for 30M depth, so we shall see..  The benefits of this rig is that it is only 2 cameras now, and that's a lot fewer memory cards and video files to manage.  Stitches are also much easier to do.  The weight of 2 cameras, and their individual coverage angle also allows this setup to be mounted to a drone to be able to do aerial 360° videos, while making the drone disappear, and the viewer appears to be floating thru the air. Bonus! 

There are a variety of options available to you for rendering 360° video, sometimes it is bundled with the camera (as in the PixPro Software with the Kodaks), but generally, if you want software that gives you more control you will have to pay for additional, third-party software.  I have experience with both PixPro for my Kodaks, as well as Kolor's Autopano Video Pro 2.x and Autopano Giga 2.x, and have used both software packages to stitch the same footage, and personally I like the control given to me by Autopano Video Pro 2.5 with Giga.

Autopano Video Pro 2.x and Autopano Giga 2.x are made to work with each other.  Giga is used to help define the stitch points, and straighten your horizon and verticals, while Video Pro does the rendering.  I find that by adding control points, my seams are less noticeable, and videos match up better.  Pixpro doesn't give you any of those advanced processing capabilities.

Once I have my footage stitched, horizon set and basic video rendered its time to start using some editing software to put it all together.  Unfortunately for me, Pinnacle Studio 19 did not support the 2:1 aspect ratio of 360° video, and Studio 20 wasn't out when I assembled my first rig.  This meant that I had to use an even more professional piece of software, Adobe Premiere Pro, to do any editing without adding a zenith or nadir.  Holy crap!  Premiere was so much more powerful than Pinnacle Studio (but it also comes at a much higher price), the power and control you have over effects and transitions was much improved over Studio, and the interface was much smoother.  Sure, it looks very complicated in the beginning, but once you use it for a little while, it becomes easily navigable and familiar.

I quickly learned that if you want your titles (and logo overlays) to look proper in 360, you would need to do some post-processing to have them appear undistorted.  For this task there is a plugin for Adobe After Effects called Skybox Mettle.  This plugin distorts your flat image, or text, or even video clip, so that when the 360° video is wrapped, it appears flat and not distorted.

So by now you have a stitched, edited and titled 360° video that you can view on your phone.  You probably want to share this video, and to do that you will need to inject some metadata first, otherwise YouTube 360° and Facebook 360° will not render the video properly.  For this task I use the Spatial Media Metadata Injector.  This tiny program injects a snippet of code into your file that tells YouTube and Facebook to play the video in 360°.  If you upload a file without metadata, it will look like a distorted video twice as wide as it is tall because you are viewing the panorama, in flat view.

One of the hardest things to adjust to in 360° video, is that everything is always in the shot.  Sure, you can do some post production editing to remove the monopod, or mask a small area, but there are no "behind the scenes" areas.  Everything is in view, all the time.   This can present many challenges in choosing a location for your shoot.

The most obvious challenge here is lighting, as any auxiliary lighting will be seen in the shot.  You need to try to take advantage of all the natural lighting that you can, and hopefully it works well with the video.

Anybody that you don't want in the shot also has to hide, around corners, behind bushes or cars, otherwise they will be in the shot as well.  Things that get to close to the camera will be affected by parallax where the cameras are seamed, this was a particularly big problem with the 7 GoPro rig underwater without domes, as divers would disappear as they swam across a seam line, only to reappear moments later coming out of that seam.

Leave a Reply