This is a case study on a shoot I did earlier this year, focusing on the challenges I was faced with and how I overcame them. It was a night shoot, set in the depths of the Cotswolds, filming for an episode of Celebrity Haunted Mansion Live featuring Katie Price and the rest of the ghost hunting gang.
As per every night op, we do a site visit during daylight hours to recce the area and identify any possible hazards that may not be immediately apparent once the sun has gone down.
With the final glimmers of light still in the sky, we did some test flights and played with a few different flight paths and shot types to give the director some ideas and options. Once we had established the shots we wanted, we stood down until around 10 pm when it was fully dark and shooting would begin.
It was only once darkness had fallen and we were getting into position that I realised I had made a slight oversight.
We were after a shot flying low along the valley floor while the camera locks onto the mansion as we approach and fly past it.
Usually a pretty simple shot right?
From where we were stood, I would be flying the aircraft below the tree line (putting it against the darkness of the trees) and with the absence of light, depth perception is a lot harder to gauge, making it more challenging to judge the distance between the drone and the tree line. Usually, I would just change the position that we were operating from but that wasn’t possible in this case due to reasons beyond our control.
Thanks to the onboard GPS, you can usually view your aircraft’s location overlayed onto a google satellite view which would help me determine the drones distance from the treeline. However, in this case, we didn’t have cellular signal to download the satellite imagery and the location staff had gone home so we were unable to access their WiFi. So this solution was out the window.
Map without internet connection to load the satellite view.
Map with loaded satellite imagery
Flight Log Telemetry
I had flown the path earlier that evening in the daylight so I dove into the flight logs to read the recorded telemetry from the previous flights. Using this, I could tell exactly how many meters the drone was away from the take-off spot and also it’s altitude.
With that data to hand, I now had an accurate start, mid and end point to aim for by using the flight telemetry alone.
I was also able to use the VPS height (which uses sonar) to give me a more accurate reading of my height above ground level rather than my altitude relative to the take-off point.
Collision Avoidance Sensors
It’s important to note that although the DJI Inspire 2 does have front facing obstacle avoidance sensors, these would not be effective in this scenario as they are just stereoscopic sensors so solely rely on the obstacle being lit so that the sensors can see it. Thankfully the flight path was over an open field with just one tree to skirt around and I was stood in a position that allowed me to spot it well enough thanks to the ambient light given off from around the house.
Flying In GPS Mode
To avoid any unintentional drift, I kept the drone in GPS mode so that I’d have peace of mind that the drone would only move to where I wanted it to go, regardless of wind direction or the aircraft’s momentum.
What Have I Learnt?
Thankfully with my experience and knowledge of the drone, I was able to adapt and overcome in this scenario but no matter how good or experienced you are, there are always lessons to be learnt.
If in doubt, download the map of the area you’ll be flying as part of your pre-site checklist. This will save you time, agro and having to use the workarounds I deployed in this situation.
If you fail to do that, well then at least you’ve read this and can fall back on these solutions.