What's wrong with the status quo?

Capturing high quality 360° photos is hard, slow, or expensive.

‍Professionals use high end equipment like DSLRs and glass fisheye lenses. They have learned and practiced proper technique and esoteric photo stitching software.

The barrier to entry is $780 and plenty of time producing photos that don't stitch well before you get good enough to produce photos that come out looking nice.

Then there are other problems like having to carry around 10 pounds of camera, lens, and tripod. Or not being able to see the results until after stitching, which can take a long time to process depending on how expensive your computer is. If you're really unlucky the stitching software will fail because you jiggled the camera off the nodal point during capture. (Professionals pay even more for boutique nodal point fixing devices)

Casual photographers do have an option. Smartphone based 360° capture. It's called Google Photosphere for Android. Apple also has apps with similar functionality.

This family of solutions isn't hard or expensive, but it's slow and the quality is poor.

Don't get me wrong, casual capture is faster than the method used by professionals, but it still takes about a minute to capture a single photo. Compared to snapping a photo for Instagram, which is instant, status quo casual 360° capture is too slow for mass adoption.

There is a new option that sits in between the previous two. Self contained 360° cameras. These are fast, instant capture, but the output resolution is low and the price is high enough to be an obstacle for casual daily users.

What should exist but doesn't?

Faster capture method. Aside from being slow, you look ridiculous while capturing a photo. Looking ridiculous is fine for novelty use, but doesn't scale for daily casual capture. Instant capture is ideal.

Cheaper 360° camera. Mid hundreds is still too expensive for casual photographers.

Painless stitching process. Nothing ruins a great photo like a dog with 3 heads. Instant stitching is ideal.

It's a tradeoff triangle. It seems possible to achieve 2 of the above at the same time, but there is not a 360° camera that is faster, cheaper, and produces error free stitching.

How should the solution look and feel?

Before landing on a real solution, I tried and failed several times.

First attempt: Maybe we could sell a device that reduces stitching errors with smartphone based capture. Instead of spinning in a circle, let a robot do it for you without shaky hands.

Learnings: Cost and bulk would still be hinderances, still too niche of a solution.

Second attempt: Maybe we could spin a GoPro at high speed on only 1 axis since it has a fisheye lens built in. The blurred output video could be de-blurred since we know it is only blurred on 1 axis. This would be fast, cheap, and portable.

Learnings: The de-blurring algorithm and stitching software would require further steps that re-introduce unnecessary complexity. Any gains in faster capture would be lost to these extra steps in post production. Furthermore, for this to work effectively would require running the GoPro at its highest framerate, which unfortunately also means a significant reduction in brightness.

Third attempt: Maybe we can just speed up smartphone based capture with pocketable fisheye lenses for smartphones. Less photos required because they capture wider angles, 6-10 instead of 24+. Improve stitching quality because there is more overlap between input photos. Also, the lenses are cheap at <$10.

Learnings: This actually worked pretty well! Huge points for portability and low cost. Also great that the stitching quality improved. This was a better user experience than the status quo, but I still wasn't satisfied with the overall results from smartphone camera sensors. The sensors were just too small to capture enough light, and this issue was amplified by attaching fisheye lenses to them. Output was grainy from high ISO if you wanted to be able to see what you were capturing.

The problem I kept having was that I was trying to build a camera system that couldn't exist in the real world.

‍Whether it be the limitations of smartphone camera sensors or the bulk of robotic assistants, physical constraints were unavoidable. My goal of building an impossible camera was, well, impossible!

It seemed impossible until one photographer using my VR photo sharing platform uploaded this...

Final attempt: A user just uploaded this 360° screenshot created manually by spinning around, taking many screenshots, then using the same esoteric stitching software as professional 360° photographers.

What if we automated that 10 minute process down to 1 click?

Learnings: It's possible to do the impossible if you're dealing with cameras that weigh nothing, cost nothing, and are invisible. Trifecta achieved! Instant capture. Free. Perfect painless stitching.

How can I test the solution?

I hired the smartest person I knew in the VR industry to write wizard level code and collaborate on the features we'd include in this impossible camera.

Chiara Coetzee developed the camera to spec, but went even further by adding support for stereoscopic 3D 360° and video capture.

As a VR photo sharing platform, my company didn't need video capture, but there was clear value to users to include it.

To test the solution, Chiara published the camera which we called "360 Panorama Capture" on the Unity Asset Store under a free and open source license. We got some press coverage and I started promoting the solution to VR developers.

The idea was that the ability to capture 360° photos for free and instantly could be passed on to the users of software made by these developers.

One key feature was the ability to upload to my VR photo sharing platform with the same 1 click it took to capture a photo. I pitched VRChat, a social VR platform, on testing out this functionality with their user base.

Did the solution work?

360 Panorama Capture became the most recommended virtual 360° camera. It was the first free, instant, and easy to use consumer 360° camera with high quality output.

It was used by >500,000 people to capture photos in VRChat. This is 50x the estimated number of professional 360° photographers worldwide.

Beyond VRChat it has been employed for architectural pre-visualization, education, medicine, and art. The use cases are limited only by the imagination of VR content creators.

Beyond photos it's been used to capture some of the most viewed 3D 360 videos in the world.

This solution created a new market of weekly active virtual 360° photographers from 0 to 16,000 with $0 in paid marketing.