The flatland technique was inspired by photos from Aydın Büyüktaşb, an amazing drone photographer from Istanbul. The process is fairly simple but requires quite a bit of work in post production. I’ll cut to the chase.
You’ll be taking a series of photos with a drone. You’re going to need a ‘subject’ which is the main focus of the photo. As you take the sequential photos you’ll be moving the drone and changing the pitch of the camera. Let’s begin.
To start, turn on the rule of thirds bars in the DJI GO app. This will help you align all the photos. I always chose a point on the first photo (e.g. the horizontal part of the bottom of the roof the the main building) and remember where that is in relation of one of the bars on the screen. The first photo you take should be from a height of roughly three or four meters; wherever the drone hovers nicely and doesn’t sway too much. As low as possible is what you’re going for here. The camera will be pointing directly at the subject. You can see where this photo was used below. It’s the base and the main focus of the final photo.
For the next photo you higher the drone by 10/15 meters and move slightly forward. Align the point you chose in the first photo (the bottom of the roof) with the rule of thirds bars again by tilting the camera down. Take the photo. What we’re going for here is the drone to make a nice curve up and over the subject you’re taking a photo of. You can see what I mean in the photo below. Once you’ve reached the top of the curve and your maximum height you can now continue in the direction you are flying and take more photos pointing straight down. This will extend the photo up the frame (remember, it’s going to be a portrait oriented photo). You can keep shooting depending on how tall you want the photo to be. I would recommend shooting a few extra shots after you think you’ve got everything you needed, just to be sure!
Jump into Photoshop and open the first photo. Extend the canvas up so you have loads of extra room to work with. Now open the next photo you shot and drop it onto the same canvas as a new layer. Drop the opacity to 50% and align the point of reference with the photo on the layer below. Use the marquee selection tool to delete everything below the point of reference (leave a little room below the point of reference for blending). Keep an eye on buildings, roads and walls and use the transform tool and warp transform to align the little details. Once I have everything more or less aligned I bump the opacity back up to 100% and use a soft eraser to blend the two layers together at the edges.
Repeat the same process above for the next image, but now you change your point of reference to something slightly behind the first initial point of reference. This is what gives you the warped effect. I hope the photo below illustrates what I mean. Essentially, you care using a little horizontal slice from each sequential photo to build a portrait photo from the bottom up.
After you stitch all the photos together and crop the negative space out, the result might be perfect, or it might be a little funky. So I like to save it as a PSD at this point, and then save another copy as a different PSD where I merge all the layers together and then do a final warp transform on the entire photo to move the photo around a little to get it looking better. Once you have it at a stage where you’re happy you can save it as JPG/DNG and import it into Lightroom for some final color adjustments and editing.
Hit me up on Instagram @ThisIsColm and Facebook: Colm Moore Photography if you have any questions about this! Let me see your results for sure and send me some tips or techniques, I'm still learning and perfecting this too!