Archive for February, 2016

jim_2013_07_25_8001

Jim Watters has been a friend and fellow pano-maniac for many years. Even when I discovered the 360 imaging world over a decade ago, Jim was already deep in it, having contributed code to the panotools software stack among other things. There are not many people in the world who have a deeper understanding of the mechanics and math behind image warping and 360º imaging than Jim Watters.

You can find his personal website at http://photocreations.ca/
“This is my personal site and is full of experiments that are not always finished but posted because I felt they were worth sharing.”

Where did we first meet in person?
I first met Jeffrey in person in Prague. I changed my return flight from Vienna and took a train up so I could pick up my copy of the original Sphericam I ordered from him on Kickstarter. Jeffrey at the time just received a pre-production version of the original Ricoh Theta camera. I got a personal tour of Prague as we walking all around testing it out.

When did you first become obsessed with photography? And what else are you obsessed with, besides photography?
When I first started taking pictures, digital photography was not a thing. I won my first camera when I was in grade school in a swimming competition. It was a 110 film camera. My parents upgraded me to a 35mm film camera a Christmas soon after. In Middle school and high school I joined the year book and photography clubs. I had a camera with me on every vacation I can remember.

Right out of high school I studied photography for three years, receiving a Diploma of Fine Craft in Photography.

I had a Commodore 64 the first year they came out. I spent days writing programs for that thing. I started to learn assembler programming on it.

Instead of photography I was also considering studying mathematics or computer science. With no knowledge of a way to combine my two passions of photography & computers at the time, I abandoned computers and began a career in photography.

As time progressed I was using computers more and more to manipulate photographs. My passion for computers increased. With the goal of combining photography manipulation and computers, I added a Bachelor’s degree in computers from SAIT to my list of completed education.

While at school, the first megapixel digital camera was released and I knew I wanted to combine my knowledge of photography and computers at manipulating and enhancing digital photography. The internet became commercial and witched from mainly sites of .gov and .org to .com.

I have always loved being creative and I challenge myself at making things with new materials and technologies all the time. I made my first pan head, I designed and 3D printed several multicamera rigs. I have hacked and modded electronics to get them to do things I wanted instead of what the manufacture intended.

How did you discover 360 photography? When was it and what happened next?
360 photography was a natural progression from the large photo collages and wide angle photography I was doing at the time. My photo collages had overlapping enlargements glued down on 4 foot by 6 foot board:

guitar

In 1991 I learned the proper way to tilt and rotate the camera around the NPP (No Parallax Point). I created my first pano head out of wood. I was still shooting film and cutting the prints to piece them together with tape. These were not even full 360° horizontal.

As soon as software and computers were able to stitch digital versions of panoramas I was shooting panoramas all the time. I purchased a film scanning so I could create seamless cylindrical panos for the purpose of printing.

When I discovered interactive panoramas I jumped on board. I was hooked at first sight. In 1997, soon after Apple released QTVR I went back to some of my old negatives that I shot to be printed and made them interactive on the computer. I realized that partial panoramas was not enough. Panoramas must be full sphere, 360° and consumed on a computer.

This is when I realized I wanted to specialize in software for creating panoramas not just manipulating images to make them pretty.

I started updating the open-source Panotools software that Helmut Dersch started, the same frameworks that PTGui, Hugin, PTAssembler, VideoStitch, and others use for aligning and assembling images into panoramas. I was distributing the software from my website, but after others started submitting changes too I help move it to SourceForge for easy updating and distribution.
Over the years of making my own changes, mentoring Google Summer of Code projects, making Windows Photoshop plugins I have probably acquired the highest level of insight into that code than all other contributors.
Are you a professional or amateur photographer?
I have recently switched from being a software designer for panoramic photographers to working as a consultant as a panorama guru in the 360° panoramic video specializing in stereoscopic 360° video for the purpose of viewing in HMDs.

Do you travel much to do your photography?
A current client wants to show off some exotic locations in VR video. I have been travelling as Director of Photography and Panorama guru on those shots. As a result I have been travelling every few weeks to new locations. I am enjoying this part of my new career choice.
What kind of photography do you like the best? and of what kinds of things?
For still images, I like panos of geographic locations that have many layers of depth and detail. Shows intriguing detail when zoomed out and in. For video I like stories about people. I want to be drawn into their lives and experience what they are experiencing.

I want to tell stories in VR video. I am not the best director and I am not a good writer, but I am kick-ass at the technologies needed. I plan on teaming up with as many projects as I can to make VR video the best it can be.

What is your opinion on today’s state of VR? Will VR, as we know it now, hit the Mainstream in the next 12 months?
I love printed wide angle panoramic images. Panoramic images are improved by making them interactive for viewing on a monitor with a mouse. Interactive panorama images become awesome when viewed in a HMD as VR. The current pixel density of HMDs are no where near what they need to be. The images on screen are either soft or pixelated. But this is overwhelmed by being able to just rotate your head around and see the scene as if you were really there. When the hardware catches up to the need things will get real. 🙂

It will take several years before people will be able to consume VR as we do the Internet now. Right now you can look at just about anything on a monitor or smart phone navigating from one page to the next. The internet is used is ways that were not imagined 20 years ago. VR is not like that yet. It is very limited but growing at an fast pace. The technology will improve over time too. We need faster processors, larger higher density screens, and more content.

Google Cardboard is a good method of getting people an idea of what VR might be. GearVR is currently the best mobile option for a VR headset. My local cell phone company is giving away GearVR headsets with Samson S6 or Note5 purchase. Mattel View-Master is a high quality plastic Google Cardboard for only $30.

Using Oculus order day stats it will take several months to fulfill its preorders for the Rift. By next Christmas VR will be huge.

I am still looking forward to a time when everyone can afford to capture video using lightfield technology. Having your head rotate at a fixed point in space can be disorienting in VR.

At first only the large motion picture houses will be able to afford the equipment needed to capture and process lightfield footage such a camera will acquire. Then no one will be able to download and watch because the file sizes and processing power needed to watch it will exceed the hardware uses will have. But eventually, it will happen.

I am looking forward to get my Sphericam 2. Synced global shutter will make stitching so much easier. It is exactly the specs of the minimum camera that is needed right now.

It was nice to see a couple new cameras for VR video introduced at CES.
Who are some of the interesting companies or people who are getting into VR / 360 Photography these days?
The VR porn industry will be big. They are pushing at the boundaries of what can be done in VR video. They have short turnaround, making lots of videos as simply as possible. Most companies in this space have standardized on 180° stereo shot with two synced cameras. Although not as immersive as full spherical stereo panorama it is still very immersive. Reading customer comments of what they do and don’t like about each video is very informative.

Camera position is the most important thing about VR video. I have shot panoramic video that works ok when viewed with a mouse on a screen but fails when tried to view in a HMD. I have a poker video shot from above the center of the table. In VR it is too weird of a position to watch a poker game. It does not work on many levels.

What is your opinion about 360 Video?
360° video is my livelihood these days. I want to capture everything in 360° stereo but the extra time and other issues limit stereo to only the best possibilities scripts. I try to watch a 360 video every day. Not always in VR but often.

As a pioneer in building the equipment that panoramic photographers use, what kind of trends have you noticed in the last years / decade that might not be obvious to other people?

I see myself as an evangelist of everything panorama and recently also VR. I want the masses be to able to create and consume VR content. In this drive I have searched for inexpensive ways for photographers to create acceptable content with off the self components.

I created my first 360° panoramic video recorder rig with webcams back in 2005. I have been actively looking at every small action camera that comes out to see if I can use it to improve an existing design.

Project_Beyond

above: Jim Watter’s stereoscopic panoramic camera design; below: Samsung’s Project Beyond design, revealed some time later

My most popular design I created was with 7 overlapping pairs of Mobius cameras creating a stereo 360° camera rig that resembled what Samsung came out with Project Beyond a few months later.

In the fall of 2013 I acquired a bunch of these small Mobius action cameras that I intended to used to make the smallest 360° video rig that I have seen. Before I had the rig assembled I tried a Oculus Rift DK1 for the first time and realized that all panoramas needed to be in stereo. That panoramic video needs to be in stereo. I did not even attempt making a mono rig and went to work on building a stereoscopic video rig. No one had one available for sale.

The result was a proof of concept and a lot of knowledge about what hardware is needed to make such a rig. http://photocreations.ca/3D/mobius_camera_rig.html

I shared what I discovered on my website and I hoped that someone or some company with more resources than me would take my idea and make a usable camera rig. The lenses pairs need to be about 65mm, the same distance as the human eye, apart. To be enjoyable to watch on HMD, the frame rate needs to be at least 60fps, to utilize all the viewable pixels of current HMD each pano needs to be about 6k wide. To be portable the cameras need to write to internal storage and run off batteries that would last about an hour. For stereo it is important to have the cameras genlocked together so the image captured by both the left and right eye are captured at exactly the same time. The Mobius cameras I was using solved some but not all of these issues.

Samsung came out with Project Beyond a few months later. On paper it sounded like everything I was hoping for. I used 7 pairs of overlapping cameras and they used 8 pairs. Their rig is a single camera with 8 pairs of sensors covering the horizon and a single sensor covering the zenith. 18 months later and there are no sample videos to view. They put a request out to have a few 360 video producers to test and give feedback on their rig. I heard rumors of problems with image quality but have not been able to confirm anything. I was surprised that they did not have any big announcements at CES this year.

Everything happens in incremental steps. Both new methods and new components are needed. Over time the components get better and cheaper. Methods get faster and easier. I was reviewing an email discussion I was having in 2001 about live panoramic video on the web. Things have improved a lot since then but it is still not an easy thing to do. In the last 15 years panorama photography has been a niche market but now with VR becoming bigger we will see things change at a faster rate.
How do you think our panoramic medium will evolve over the next years or decades?
Software updates will make stereo panoramas easier to make. They will be the award winners for HMD viewing.

Lightfield capture which is the holy grail of image capture for VR will start to become possible in two years. Distribution and viewing lightfield video will will not be easy for a couple more years.