I have always been a keen photographer and more recently have been captivated by slow motion video which is in actuality high speed filming. An example might be filming for 1 second at 1000 frames per second (fps) which when played back at a more ‘normal’ 25 frames per second would give a movie about 40 seconds long.

Typically cameras used for this type of work are extremely expensive for example the phantom range as used by the YouTubers The Slow Mo Guys. Way beyond my budget in this life time so I had to look for an alternative. There are a few options out there but as usual the best options are US based only. Over here in the UK primarily we have still cameras that can shoot at up to 1000 fps e.g. the Sony RX10 bridge camera and RX100 pocket camera ranges although often at reduced resolution the higher the fps and for a limited time, usually just a few seconds. One other UK option is the FPS1000 produced by Graham Rowan of The Slow Motion Camera Company (open in a new window/tab)

I bought my one about a year ago (Xmas 2016) after it was developed from a kickstarter campaign. Graham has been developing the firmware steadily and it improves with each release. There is a way to go but as they say, Rome was not built in a day

Anyway, for some strange and inexplicable reason I thought it would be very cool to try and film an air rifle pellet in flight from barrel to target. A scan through YouTube didn’t give me any results that were similar to what I was thinking of. There were lots of videos of pellets hitting things but the camera was stationary, I wanted to track the moving pellet.

The magnitude of the challenge?

So, maths on the back of a fag packet time. I plucked some figures out of the air to work with. I thought a triangular setup might work best so at the corners I would have the air rifle, the target and the camera. I said 30 feet from gun to target (an innocent egg) and the camera about twenty feet from the closest point of the flight path. My rifle shoots its pellets at about 525mph so pretty quick. This means the camera will have to rotate from the gun to the egg in about .04 of a second or 40ms. Turns out this is not much time and the camera is way too heavy to rotate at this speed (about 350rpm). If my garden setup is smaller the camera will have to move faster. Halve the distance from camera to flight path and the camera rpm has to double. This is clearly not practical so a plan B is needed.

Plan B. This consists of filming the pellet through a mirror and rotating the mirror rather than the camera. On the face of it this is more achievable but still tricky. Whatever I do it won’t work first time so it will need to be repeatable and adjustable. I have a plan that is related to my messing around with servos on my CNC machines.

Ok, I’ve been playing with an arduino and a sound sensor. I have captured the sound event as a trigger and taken an action based on it. Best laid plans – I used a buzzer as I forgot about the onboard led so you can guess what happened, very persistent buzzing. I also have a couple of beam break style leds arriving from Amazon today so some timing measurements are in order. Barrel to target using beam breaks each end and then compare that with the sound trigger. The fun bit might be breaking the beam at 30ft with a very small pellet. Some sneaky thinking needed me thinks. My cheap and cheerful eBay mirrors should be here next week some time too.

31/12/17: The servo I ordered from China has arrived and after a nervous false start is working from the pc via a serial link. Whilst tuning it I have realised that it won’t accelerate as quickly as I had rather naively hoped. This has meant a rethink whereby I will be spinning the mirror and timing the firing the gun with it. A challenge for another day. In the meantime I have been building an interface using a 4.3″ capacitive touch screen and an Arduino Uno. This is progressing nicely with the motor controlled via an interrupt and the interface is a couple of slider bars on the touchscreen.

I also have a hall effect sensor set up on it reporting rpm but this is not very accurate when checked against an optical rpm counter. The rpm from the maths and the counter is actually very good against the external check. At this point I am thinking an optical counter for the trigger will be needed. This will give me a consistent point in each rpm and I can then count pulses to determine when to fire the rifle.

This is the ‘workbench’ next to the keyboard. There isn’t much room

And the Arduino Uno

This is the 180W servo I am using along with the heath robinson contraption to hold the hall effect sensor

The ACS606 step direction controller

And finally the current prototype screen design on the 4D Systems development screen.

4/1/2018: I managed to find about 4 days over the Xmas / New Year period and have been busy learning the 4D Systems design/development environment and how it interacts with the Arduino Uno. It turns out that after a short while the screen was easy to use so over to coding. Screen design is almost complete and there are now four ‘pages’ devoted to different tasks. Page 1 is all about setting the rpm. Page 2 adjusts the triggering time in degrees, page 3 is a simple save choice and page 4 is setup options. There is a memory function too so one rpm speed can be used over and over. The appropriately named fire button will run the whole process from running the servo up to speed, triggering the gun at the correct time and running everything to a stop. There are a few more updates to make however the back of the coding is broken so it’s on to experimentation time.

My initial plans are to have the arduino turn on something and the mirror swing over it. The video should then show the led or whatever coming on in the middle of the swing if my timing is right. I haven’t thought this through yet so room for change and mess ups.

If anyone reading this wants to leave some thoughts or ideas as to where I can take this then pop them in below.

2/1/2018: Well, Santa took rather a long time crossing the pond but he has finally made it and delivered a Chronos 1.4 high speed camera. Coupled with the FPS1000 I already have this should enable me to achieve some of my plans in this field. I’ll have to do some experimenting first and get used to the kit and its limitations.

Thanks

More soon

Last updated 4/1/2018