Bullet Time Photography

I am starting to come to the end of my testing stages, as I have found pretty much all the problems you could ever have with this rig, and those problems were/are EVERYTHING!!

Everything from camera alignment, focal length, focus, white balance, colour balance, camera level/tilt, rig positioning, plus a few other juicy ones involving extreme electronics. But after 6 months of trial and error, and hours upon hours of setting up and taking down, we have kind of got it perfect.

Almost every part of the project has been hand made, from the frame that the cameras sits on, the ball and socket heads that are between the cameras and rig, to the internal electronics of the cameras to remove shutter lag that was up to 1.3 second in some cameras. We then had a firing system produced that allows the control and firing of the cameras. That was produced by an electronics company who makes satellite systems for governments. The trigger part cost more than my car, and the whole lot to date would cost more than a brand new BMW M3 to reproduce, so if your thinking of doing this really think hard before going for it.

When I started I thought ah this will be simple, a 2 core cable, that connects the cameras to each other in parallel through simple jack plugs where the remote shutter goes and it will work. Well I can tell you now that failed massively as the cameras don’t work as you think on an on/off system, I spent almost a month trying to work out why, and eventually contracted it out to a micro-electronics electronics company.

With a huge amount of work that took over 3 months, including the cameras having to be stripped down, 2 chips replaced and the actual circuits to be rebuilt to rewire them, we overcome most of the problems of time lag and joining them all together

We now have it working correctly and the time delay between the firing mechanism down to 50 billionths of a second, but there are still slight delays we cant remove as that is down to electronics of the cameras we cant touch. These delays are less than 2000th of a second, not a massive amount, but I suppose with a fast moving object like a spinning bottle from a cock tail waiter you can just assay see the difference.

We have now done quite a large amount of testing, but as I don’t have a studio, I have been using a local college who have a studio when not in use, which is not often, so I get a small window of time every now and then, where I have to set up, take images and take back down again before the college closes at 5pm.

The real problem Is this, imagine trying to learn photography with no light meters, nobody to ask for advice, and no information on the cameras. Then every time you take a photo you have to wait until it is developed to find how it looks. That’s where I am now!

Every time we take the images, we have to then spend almost an hour taking all the images from the cameras, another 3 or 4 hours trying to work out which image is from which set, then another 6 or more aligning the images so they don’t jump all over, to find where you need to research next. By that time though the rig is all taken apart, and I have to wait until the next available time slot, where it takes another 4 hours or so setting the rig back up, to do more test shots.

I have even made a laser system, which works similar to a rifle sight, so it puts a dot where the camera sees exactly, to help align the cameras.

I have some examples here, that you may like, they aren’t perfect as they are another tester, we need to now move onto lighting the environment it is in. I really need to find a semi permanent place to keep it all set up so I don’t have to set it up each time.

I’m not sure how we can light any of it without one camera or more seeing a light?

But ill keep you posted when I have done more.

http://www.quicktimevirtualreality.com/bullettime/portfolio.html

All the best
Ian

PS
If your wondering, what the hell has this got to with Virtual reality? The whole project is designed to make Object VR movies, but without a turntable :)