Isn’t it funny how getting up at 5.30 in the morning to go shoot the sunrise seems like a great idea the week before, but increasingly terrible until the alarm finally chimes following a terrible night’s sleep at which point it is DEFINITELY the worst idea ever?
This Sunday morning at 5.30 when my alarm went off I have to admit I was very, very tempted to jab the “dismiss” button, roll over and go back to sleep. But I didn’t. I dragged myself out of bed, pulled on two sets of clothes (I was riding later in the day and it was cold!) and headed out into the chilly but supremely clear morning. After taking the Mi drone just to get the hang of it on a rather gloomy, windy afternoon last weekend I wanted to make the most of the good weather and test out the capabilities of the camera a little bit more.
With two fully charged batteries the plan was to shoot the Humber Bridge (it’s a cliche, but I’m learning a new toy and sometimes familiarity is good) from a couple of locations on the north bank; one to the west of the bridge with the sun rising behind it, and one from the east with the glorious golden light behind the camera. I parked up as close to my first location as I could manage, and walked a few hundred yards to my planned takeoff point. The tide was very high, which meant I had less space than I’d planned for, but there was still more than I needed.
First interesting thing to note – there was no faffing with the compass here whatsoever (on my first flight I was having to wave the drone in the air then take off really quickly due to compass error notices). I popped it down, turned it on and took off. It was a much clearer day (and I suspect lower pressure, though I didn’t check) but I was also taking off from a hard surface rather than grass.
I was mainly shooting video (see top of page) but I shot a few stills too to test out the camera in some reasonably challenging conditions, shooting straight into the sun just below and above the horizon. I was expecting it to struggle – the camera only has a tiny wee sensor (I believe a 12mp Sony CMOS, which I guess is similar to that used in mobile phones) but in all honesty I was pleasantly surprised. It’s never going to change a DSLR (or mirrorless, or even a decent compact) for image quality or dynamic range, but I managed to get some decent images out of it with no trouble. These are lightly processed as I shot in RAW but nothing excessive. I’m also impressed at how well it handled lens flare – even looking straight into the sun there was hardly any apart from on a couple of shots.
The camera struggled a little bit when tilting up from the water to the horizon but to be honest I don’t think any camera would have managed that so I can’t really moan too much. The gimbal mostly behaved too – I had a few wonky horizons but generally flying around for a bit would see it righted. I can see it getting annoying shooting anything critical, but as it’s just a toy for me it’s not too big of an issue.
I’d turned off beginner mode before taking off, and set my limits to 120m altitude and 500m distance, which seems like a decent amount, certainly for the photography I want to do. at 500m it was getting fairly hard to see – I’d push it a bit further if needed, but to be honest I’d rather move down the road so I can see what I’m doing properly. Apparently there’s an altitude exemption for FPV drones to allow you up to 1000ft (so about 300m) with a spotter, so I’ll probably give that a go at some point.
After flying non-stop for about 20 minutes (down to a battery of 37% – I’ll have to do some timed flights and see how long it actually lasts) I came back down in my little landing spot and headed back to the car to warm my frozen fingers for a bit before continuing. I took off from the opposite side of the bridge this time, to capture some of the glorious golden hour light.
Again I was really quite impressed with how well the camera rendered some fairly challenging lighting conditions, with pretty dark areas of shadow next to fairly bright areas in the golden light. If you’re expecting the images and video to be on a par with something like a Phantom 4 then I’m pretty sure you’re going to be disappointed but on paper and in practice I don’t think it’s far away from what you’ll get out of a Mavic, which has a similar size and quality sensor. If Xiaomi could beef up the camera options in the software (more manual control, burst still shooting, time lapse and the option to record video in AVCHD or similar would go a very long way) then the Mi drone starts looking like a pretty serious bit of kit. As it is it’s still really impressive but just that little bit lacking.
I brought it down a couple of times during this flight – once because the camera part of the app locked up and I wanted to reset it. This isn’t the end of the world as you can fly perfectly well without the camera (and could well be down to my flaky phone rather than the Xiaomi) but it’s worth mentioning, and definitely a plus point to staying legal and in LOS.
After packing the drone away and charging the battery while we were out riding I took a friend out for his first flying experience in the afternoon. The light was much less pleasant so we took the opportunity to spend some time flying around a fairly small area, getting a feel for the handling of the drone a bit closer to the ground rather than miles up in the air. It’s actually really responsive and quick (though nothing like a racing drone, the stabilisation makes it much more forgiving). Even in beginner mode (limited to 8m/s) it was really good fun to just fly around.
Next plan is to find some trees to fly it in, possibly in Opti mode (still uses the altitude technology but turns off the GPS) and see if I can not crash. I’m also really keen to try out some of the automated modes (waypoint, orbit etc) and see how well they work.