In an attempt to avoid giving the impression that I shoot exclusively film, here’s a post dedicated to a digital post-processing technique that has become a bit of a cliché, but that I like to use occasionally – the fake “tilt-shift miniature”. Like other effects, such as tone-mapping, or “Orton” processing it can be easily overdone, or done poorly and so result in a proliferation of not great images when they become the latest hot trend.
So what is it? Well, basically, it’s a selective blurring, applied to an image to try and fool the viewers brain into mistaking the scale of the scene and creating the illusion that it’s a photograph of a very detailed model. By applying blur to only a small slice of the scene to show only a shallow plane of apparent focus, we create false visual cues to the distances in the scene. Normally we are used to seeing only very close objects within such a narrow focus, so when a blur is used to mimic a shallow depth of field on a full-size scene, the brain interprets it as a close-up of a miniature scene.
It seems commonly referred to as a fake “tilt-shift” because in the days before Photoshop, it was done using movements on view cameras, or tilt-shift lenses on smaller formats. “The Sandbox” video by Sam O’Hare was assembled from thousands of stills shot using tilt-shift lenses, and shows just how convincing the effect can be. Since you only need to use tilt movements to achieve the selective blur it might be better termed “fake tilt” but, much like the abuse of the term HDR, I guess we’ve missed the boat on that.
Now, I’m not going to go into great details and create a tutorial, since there are plenty of those already out there, but I want to write about what I think helps make an effective miniature.
I generally use the Lens blur filter in Photoshop, using a gradient mask to selectively apply the effect. This gives me a chance to apply small corrections to the masks – for example to exclude something tall in the foreground from the focus – but these can be difficult to get right so having the right image to begin with is key.
Firstly, viewpoint. If you’re looking down on the scene from a high vantage point it’s easier to produce a convincing fake miniature. I think this is because it replicates the usual angle for viewing real models, and also it reduces the chance of foreground objects intruding on your chosen zone of fake focus.
Secondly, subject. Personally I think these work best on strong subjects with a bit of space around them, but with plenty of background to blur. Urban scenes are excellent as vehicles, buildings, street furniture and people look great as little models, but landscapes can work too. The key for me is to shoot something fairly distant with a long lens, to flatten the perspective and make everything appear at roughly the same distance in the original image. When I have used wide lenses for these, I usually find I end up cropping the image down substantially to achieve a convincing miniature.
Now, I don’t ever go out specifically to shoot a fake miniature, it’s usually more of a spontaneous opportunity, such as finding the hotel room I’m staying in has a suitable view, or finding that just as I stop to look at a view the shooting range in the valley below is in use, as for the Toy Soldiers photo here. On other occasions I only realise the potential for a good miniature when I look back at the photos later.
One final thing to note is that although as a post-processing effect, it can just as easily be applied to film images. One of my favourites in fact is the photo at the top of this post which was shot on Kodachrome in 1970 by my grandfather. When I was scanning in his old slides I thought it would make a great miniature so spent some time trying out different versions, before I settled on that one. Got to love the old Routemaster.
See more on my “Miniatures” flickr set.