There are few shots that elicit more of a “how did they do that?!” reaction from viewers than a great tilt-shift shot. This “miniature” technique has been used at monster truck rallies, construction sites, Disney World, Coachella, and even for Stephen Colbert’s Late Show credits, all to great effect.
How It Works or: The Scheimpflug Principle
The miniature effect is created by utilizing the “tilt” function in “tilt shift.” Tilting the lens changes the angle of the lens plane away from parallel (the Sheimpflug Principle) and allows you to change which parts of the shot are in focus. It creates an extremely shallow depth of field, which creates the perspective that it’s taking place inside of a diorama.
The “shift” function of the lens is great for shooting buildings, landscapes, tall landmarks, or objects that can get distorted by the wide-angle lens. Instead of pointing the camera up to capture a building with distortion, the lens is pointed straight ahead (or as close to level as possible) and then “shifted” with the knobs up or down to get the intended angle. (That’s about all I’ll say about that, since this post is more focused on the miniature effect.)
Tips for Shooting Successful Tilt-Shift Footage
The whole point of using a tilt-shift lens for the miniature effect is to make it seem like the whole scene is a scale model using a macro lens, so you need to consider a few things to replicate this look:
Be elevated: Looking down at the scene is crucial, because it makes objects appear smaller and gives you a bigger area to see in the shot.
Consider the scale: As you’re trying to replicate a scale model, you need to make sure your subject or objects are fairly small in the frame.
Zoom in (if possible): The longer your focal length, the more compressed everything will be in the shot, resulting in a better miniature effect.
Use Moving Objects: If there’s no change in the frame, the miniature effect may not translate as well. It’s why most tilt-shift shots are of people, cars, boats, etc.
Make it a timelapse: Most tilt-shift videos you see are timelapses. This is because our eyes think of tiny things as moving very quickly, and having tiny moving objects jump around in the frame adds to the diorama effect, making objects look like toys. Make sure to keep a faster shutter speed (faster than 1/60) to keep everything sharp. Blurred objects won’t look right, because, well, they’ll be blurry.
Saturation: This is more for post processing, but since toys are vibrant and very colorful, you want your video to also be highly saturated.
Tilt appropriately: You should line up your tilted focal plane so that it has the best effect on your footage. It can go in any direction (although I haven’t seen many focal planes that are perfectly vertical), so experiment with the angle and rotation of the lens or effect for the best look.
Replicating the Effect in Adobe After Effects
There’s less work in post-production when using a lens to capture tilt shift, but if the shot doesn’t come out the way you want, there’s not much you can do in post to fix it. The good thing is, you can actually replicate the look just by taking regular footage (that fits the criteria above) and adding some selective blurring. In fact, a lot of the most well-known viral tilt-shift videos have been done in post. Here’s how to get a basic tilt shift effect using After Effects.
Create a new project, import your footage, and make a new composition with it. If your clip is already a timelapse, pre-compose this layer (name it “footage”) and go to the next step about the blur map. If your clip is not a timelapse, but you want to speed it up while still making it look like a higher-shutter timelapse, right-click and select Time > Time Stretch. Put in the percentage that gives you the desired clip length. A lower percentage makes a faster, shorter clip.
Pre-compose this footage layer by right-clicking on it (control/command + shift + C), then check the box to move the attributes and name it “footage.” Next, apply the effect Time > Posterize Time to the pre-comp layer and select the frame rate you want to lock to the layer, making the footage look like a timelapse.
Play back the clip to make sure it has your preferred look.
Using the Blur Map
A blur map is essentially a layer with a gradient that tells the blur effect what shape and pattern to take. There are a couple of ways to make the gradient:
With a mask: First, you need to add an adjustment layer, so right-click in your composition and go choose New > Adjustment Layer, or go up to New > Adjustment Layer and place it on top of the pre-comped footage layer. Add a camera lens blur from your effects on this layer and hold off on doing anything with the controls until you’ve created the “blur map.”
Make a new white solid, pre-comp this layer, and open it, naming it “blur map.” Click the rectangle tool (shortcut: Q) and draw a long and skinny mask horizontally across the center of the solid that hangs over the edges of the canvas. Invert your mask so that the black is in the middle, then feather it 100 pixels to start. That’s all you need to do on this layer for now.
Go back to your comp with the adjustment layer, turn off the blur-map layer and go to the effect controls. Check the “repeat edge pixels” box first, then set a blur radius of 30. Change the shape to your desired “bokeh” look (decagon is the roundest, FYI), then boost the diffraction fringe to 50 as a starting point. Now, set the blur map to your “blur map” layer. You should see the effect applied to your footage using the blur map’s pattern. Proceed to the next step after the light-sweep effect section.
With a light-sweep effect: To use this method instead, make a new black solid and place it above the footage layer. Apply the light-sweep effect to the solid and adjust the angle to your preference. Expand the width over 100 for now and turn the edge intensity all the way down, then make sure the sweep intensity is at least 100 so there’s pure white in the middle, which will be the part of the clip that’s “in focus.” Now pre-compose these two layers together and name it “matte.”
Now you should have two layers: a footage later on top and a matte layer below. Apply the camera lens blur effect to the footage layer and repeat the steps in the paragraph above the light-sweep effect, setting the “blur map” to the “matte” layer.
From here, with either one of these setups, you’ll want to play back and adjust the mattes as necessary. Add some grain to your camera lens blur effect for a more realistic look, then boost the saturation with the hue/saturation effect to 20-30 to start and see how it looks. Boosting the saturation makes the colors more vibrant, which will add to the toy-like look of the footage.
*A masked blur map gives you a lot more freedom with making custom shapes, but it can also be more work. A light-sweep blur map takes takes less time (although, maybe only 5 minutes less), but you only get one shape to work with. It all depends on your preference and which results you like better. Here are the two different methods, before and after.
Regular footage before:
Masked method after:
Light-sweep method after:
Your goal is to have a very small depth of field, so make sure not to use too wide of a focal plane. Also, try to keep your shape at a skinny triangle or a rectangle, since this is how an actual tilt-shift lens works. You can, however, move the mask around objects that should be in or out of focus, so polygonal maps are okay too.
The last thing I’ll mention is that most of these effects and shots are done with a stationary or “locked down” camera, and, as I said earlier, they’re timelapses. However, as you saw in the Colbert Late Show intro, a shot can be moving, and it can also be real-time and still work just fine. It’s all about your preference and whichever settings work best for you.
If you have any questions about this post or any of the techniques discussed — or if you have your own method you’d like to share — let us know in the comments below!