What is Bullet Time Photography?
Bullet Time Photography refers to the filming technique where time is either slowed down or stopped completely, but the movement of the camera carries on. For reference: picture a camera on a dolly track travelling along a path, but surrounding factors have frozen for that period of time.
The term Bullet Time Photography was coined back in 1998 / 1999 thanks to The matrix movie. But it certainly didn’t start there, as the photographic technique goes all the way back to the 1870s with Eadweard Muybridge and Etienne Jules Marey.
What Does Bullet Time Do?
Bullet time photography can show moments in time that are otherwise too fast with any other filming style, allowing the camera operator to move around the environment and see it from other angles previously un-shown.
For example, a balloon being shot with a rifle can be viewed all the way around and even focus on the actual bullet that penetrates the balloon from anywhere on the path the cameras were set up.
Whatever the subject, Bullet Time is based on an array of cameras, at regular points along a predetermined path. These cameras are then fired and controlled to capture the photograph in succession. Those images are then put onto a video timeline and animated to be exported as a video.
Bullet Time Workflows
The cameras can be fired Sequentially, one after the other giving a controlled slow motion or controlled time control, or all at the same time, synchronous. giving a frozen moment. They can even be from frozen to controlled slow motion and back to frozen again.
Video editing software is then used to align precisely the images, so they don’t zoom, rotate or move on the X and Y axis as they move through the sequence of images, an imprecise alignment would create a glitch effect. We then grade the images and export them out as a movie in 6K, that can be reduced later on to match the other edits.
Basic Equipment you will need.
To create a professional bullet time movie, you need an array of cameras, the more the better. This will improve the smoothness of the transition from the start to the end of the sequence. Reducing the distance between the cameras to a minimum will reduce the angle of rotation around the subject to a minimum and give a far smoother result.
You now have two methods of powering, triggering, camera control and data collection. The 1st is as cheap as possible for a student project.
1, Student/personal Project
Power the cameras from their own batteries, most cameras will give you about 6-8hrs of battery life with them sat on standby (turn auto power off, off) and only take 50 photos or so. Put in a medium sized memory card (about 32gb should be more than enough) and use an infrared remote control to fire them all at once.
It is a nightmare data collecting, taking the images off the cards, sorting them into sets and then aligning them. Its how we started and we wouldn’t want to be there again for all the money in the world! But if you have time, and a lot of patience, it will work. Make sure all the cameras are on manual, and you set all the cameras settings to identical settings in every way. If anything is different, you will see it instantly as the video plays through the sequence of images. It is almost impossible to grade an image taken on different settings to match the other images… we have been there!
2, Professional shoot
- Camera Power
You want to be looking at powering the cameras, this gives you time to keep the cameras on whilst testing, aligning, calibrations etc, and the last thing you want are cameras going off as a director shouts “Are you guys ready for your slot now?” or you are about to go live on an experiential event with a queue of dozens of visitors half drunk at a party. - Bullet Time Software
You then need to look at connecting the cameras to a computer and using Bullet Time Software to control and the cameras. This means you can change the setting on the computer and press synchronise and the entire array change in one go. You have fewer worries about one being different and can change instantly as the sun moves or the directors asks if you can change it a few stops.As the cameras are connected to the computer, this also gives us the option to download the images directly to the computer through the Bullet Time Software as the cameras are fired. They are downloaded into sets within seconds and allows instant preview to make sure they are in focus, all shooting the same, nothing is wrong and all the cameras have fired. - Bullet Time Trigger
So you will need to fire the cameras in a controlled manner. We cant just use the infrared remote as above, as sometimes it will miss a camera and other times it will maybe not fire at precisely the correct time. We need a far more controlled way of doing this.We have created a custom made trigger that uses an industrial Siemens PCI computer as it works far faster than a normal home computer, and gives us full control. We can put speed fire them with exact inverts down to 1/10,000th of a second (0.1ms). We can link it to a PC and put in time curves, so we can speed ramp and down from frozen to 5 seconds in 1ms intervals. It gives us full freedom to do as we wish, and give the directors exactly what they want. - Bullet Time Fault FindingThe biggest difference between an amateur and a professional is the system and how it is built. We have created tech boxes that have all the equipment needed for 12 cameras. You don’t forget anything and you don’t need to fault find on the job.This allows us to swap out a fault bit of equipment in a moment and carry on. The tech boxes are connected to a wiring loom, we have literally hand made these ourselves, soldered every last connection. This is also hot swappable, so if damaged anywhere, we can swap in the after of moments. Each camera is connected to the loom by a small flyer wire, which as the Power, USB and Trigger connectors on. We have hundreds of these, so if there is any issues with the connectors (they DO get damaged over time), then we can swap it in a few seconds. This makes set up plug and play with everything, and fault finding not needed as we always have spare equipment with us. We will then mark it up and fault find when we get back home.We have had countless phone calls from all over the world from pro/amateur Bullet Time photographers that cant get the system to fire up, and it is always down to seriously bad cable management and just one item such as a USB hub or a power pack causing problems.
Cameras to use or avoid in a Bullet Time Rig
The Canon range are excellent, they are realistic in price, lots of spares about, attachments, lenses, and are easy to link systems into. The way they are built allows software developers to link software into it very easily. Other companies such as Nikon and GoPro also allow this, but nowhere to the level you can with Canon.
Avoid gimmick cameras such as GoPro. Yes they are “cool”, but they are famously flaky, they often go wrong and crash. They are much more expensive than DSLR, and a lot less controllable. We have multiple sets of lenses for our DSLR cameras, you cant do that with GoPros. Also it is almost impossible to fire a GoPro at precise moments. We have a set of GoPro cameras and tried for years to make it work. They are great for fun, but not on a pro set.
Bullet Time Structure Frame
The cameras cant move at all, any movement and the effect is ruined. It is paramount to the entire project, so you can not cut corners here. We use heavy duty lighting rigging and custom made solid steel clamps with good quality ball heads. We often destroy the ball heads as we tighten them with pliers to make sure there is no movement at all. During a shoot lighting guys, public, actors and other people often bump into the rig. Any movement fro this shock and you will see instantly the glitch. Also knocking a lens knocks the glass elements, which will actually change the view point.
As an example set a camera up on a tripod, lock it off zoomed into something and turn live view on. Now digitally zoom in and whilst watching on the back, just tap the focus part of the lens, the actual front. Just sort of move it up and down, wiggle it, not zooming it. You will see the view moving all over and probably the focus will be way different.
You need a structure solid enough to not be effected by minor knocks or vibrations from the flooring as actors jump about, the frame should absorb this. Add weight to it, make sure it is very strong and sturdy.
Focal Points and controlled moves
The entire process is similar to how a camera would be on a dolly. The cameras move along a path you cant change easily after set up, so putting similar thoughts into setting up, such as put a handful of cameras on with batteries in them. Then with the rest of the production team walk around and show them what the cameras can see on the rear screens, find out asap what they want to capture and what is in or out of shot. If somebody will be jumping high, get somebody to that hight or more on ladders for example. Will their head or limbs be cut off. Remember you are probably shooting in 6×4, will there be more crop when exporting at 16×9? When aligning the images, you could lose another 10%, will you still have everything in shot?
It is much easier to move around a subject that is static for alignment. It is also simpler to keep an equal distance between your focal point and cameras, such as a radius. This can make post production work hours or days shorter afterwards.
Resolution and captured formats
In general there are 2 files the camera will create, JPEG and RAW.
JPG is a compressed format, and is very flat with little data. For experiential or test shots, this is more than enough. It is quick to download and simple to work with.
RAW is an uncompressed file with all the data still that the camera could capture. It allows far greater grading (colour correction) later on, and can be digitally zoomed in further and keeping the quality. It will take longer to download the images from the shot, but it is worth every moment waiting. With 120 cameras, to download the JPGs only, takes about 12 seconds or less. To download JPG and RAW takes about 60-90 seconds per shot. The bottle neck is the USB 2.0 wire that most cameras use. You can reduce this by putting less cameras on more computers, but the costs and setup and strike time is already very high, so you need to make a balance of the 2.
Conclusion
There is much more than even this, and some trade secrets we of course cant release on here. The post production alone is very difficult and can take years to master doing it even when doing it every day. The cost of setting up your own Bullet Time Rig, testing it, creating samples and making sure you know how to use it before a shoot to hiring us for a day can be many times the cost of your rig before even counting your time. But you know that with us doing it, it will be done correct first time and with no technical down times. It’s the equivalence of buying an Arri Alexa with a full set of Primes for one shoot.
Please call us or drop us a line, we are more than happy to go through everything with you and give you a very competitive quote. Whilst you will have the confidence we have worked with the worlds largest companies on everything from Feature films, hundreds of Experiential Events and even some tiny low budget independent film makers sets.
