Each time I go out I will try to record lessons learned and the highlights of the mission.
My first flight was not quite what I expected, yet it was exactly what I had hoped for. I went out with a simple plan; become more familiar with my equipment and start scoping out good places to fly.
I started by using AirMap.com and Skyward.io to find an area without any airfields and few people. Living in western Washington state and it is hard to find an area without trees. There are often open areas along rivers and the Columbia river is close by. My research revealed an area about 45 minutes from my house and with destination in hand I was ready to start planning.
Checking ‘UAV forecast‘ I picked a time of day. I logged the flight on ‘AirMap’ so when I arrived it knew my plan. I hiked about 1/4 mile to scope out the area and to find an ‘off the path’ place I could launch from. The wind was steady and stronger than the UAV forecast predicted but other than that it was a beautiful day.
I like to pioneer new technologies which usually means figuring it out as I go. It is important to prepare by reading whatever posts are available and watching videos but there is no substitute for hands-on-experience.
When everything goes perfectly you have demonstrated competency in what you have learned. Needless to say my first flight didn’t go perfectly. Here are the lessons I learned.
I had read and seen YouTube videos that talked about this in detail, and yet it got me. When you first fire up the Mavic, before you take a picture or video, you must tap the screen to set a focus. It would be nice if the camera start focused at infinity, but it doesn’t.
I started recording and just let it run. I was working with a brand new SD. The recording seems to have just stopped and restarted on it’s own a couple of times. Hopefully shooting shorter scenes or formatting the SD in the Mavic will address this issue.
Processing the video and seeing the multiple files I realized starting and stopping the recording to capture a ‘scene’ is how you should shoot video. That way you can easily throw away unusable video. It takes time and resources to process video files. Shooting scenes eliminates processing time and provides a level of organization.
Flying where there are no airports could well be flying where there no cellular data capabilities. Programs that rely on the cloud will not work. I was planning to try DroneDeploy but it wouldn’t come up. This made me aware of the need to cache data and maps. I need to be able to fly regardless of the cellular service where I am flying. This will mean learning what applications allow caching and practicing in advance.
I thought I had this covered. However, the anti-glare screen I put on my iPad only makes it harder to see the display. I had to turn my back to the sun to shade my display. That meant when I was looking in the direction of the drone I couldn’t see the display. The anti-glare cover has to go and I purchased a sun visor for my iPad. Next time I should be able to see the display and drone at the same time.
A significant difference between a launch pad and a blanket is that the launch pad has a taut surface. A taut surfaced landing pad has significant value on uneven ground. The camera gimbals got caught in the blanket during calibration and during takeoff and landing the props were skimming the blanket. I have landing gear but didn’t want the extra wind resistance. I came to realize that a proper landing pad is a good idea.
The end product of most flights is video. Video processing requires a lot of CPU power, and disk space. Understanding a little about encoding is critical to success in editing, The ‘raw’ files captured on the SD card are in a format that isn’t good for editing. That is one reason you should start and stop the recording. You only have to encode and process the files you use!
The Mavic uses a codex that is best for recording. It is stored in a highly compressed format generated by the camera. The compression is focused on the size of the data and fast writes, not rendering. Changing the encoding from raw format on the SD card to something friendly to rendering allows an editor to use the CPU for editing, not rendering. Be aware encoding can expands the file by a factor of 10. A 900MB file becomes 9GB and it takes 10 to 20 minutes to encode.
Ah, but once you encode editors works well. Prior to encoding editing was painful and slow and the CPU fan ran at full speed. After encoded the file iMovie performed flawlessly and the CPU fan wasn’t noticeable. Video encoding is absolutely necessary!
From lesson 3: Load the applications with cached data and practice by disabling all network connections and walking through the process.
From lesson 2: Plan scenes. I want to work with the different modes and different software I have. Each task will be planned as a scene and I will create a script for walking though the software and flight. I will break the recordings between tasks and sub-tasks.
As a business I want to establish a set of ‘scripts’ that can be reuse or quickly customize to create new scenes. Capturing how to create different scenes will allow me to offer services that are well defined and consistent because they can be easily reproduced.
I feel the video from my first flight was mostly garbage because I couldn’t see the display very well. I was not having a good time and stopped at the end of my first battery. That was when I discovered Drone Deploy wouldn’t launch because I did not have a good enough data connection. I couldn’t look at tutorials for AutoPilot either.
The location had some of the features I was looking for but overall is not a great choice. Wind is something that is always going to be a factor along the Columbia river. Finding big open areas in Western Washington is not easy because of all the trees but I’m sure there are some out there and I will find them. I’ll also get good at flying around trees but that doesn’t seem like the best place to start.
I hope that in the sharing my experience you find something of value.Tags: Columbia River, drone, Hamilton Island, Lessons, Mavic, Mavic One, North Boneville, quadcopter, suas, uas, uav