How to do a levitation photo

After posting the final images of our creative fashion photography shoot at Peckforton Castle I received a number of questions about how the final levitation images were taken so I thought it would be a good idea to post a tutorial on how to do a levitation photo.

To start with here are the two images that we did on the fashion shoot that sparked the “how did you do that” questions.

These two images were actually not planned as part of the shoot, we had got all the shots that had been planned (and a few more) and we decided that as we had some time left in the location, we’d experiment a little. Levitation images are not a particularly new idea and have been made very popular over the years by photographers such as Miss Aniela, Rosie Hardy and Brook Shaden but I fancied having a go at doing one myself in a fashion shoot environment.

Although the final images are impressive, the actual technique is fairly simple and could be done on any basic camera (or even camera phone if you really had to!). The basic concept of these kinds of shots is to take a number of images and composite them together using certain parts of the different shots to bring together the illusion of someone levitating.


  • Camera
  • Tripod (or other means to keep your camera in exactly the same place over a number of shots)
  • Computer with appropriate editing software (e.g. photoshop, photoshop elements, GIMP etc)
  • Remote – wired or wireless (this is optional but will ensure your camera is not accidentally moved by pressing the shutter button).


Once you have decided on the location and position of your subject set up the camera and have your subject stand where they will be ‘levitating’ and focus the camera and then switch it to manual focus so that it does not shift focus during the different shots you will be taking (this is not 100% essential, if your camera doesn’t have a manual focus you can still do this technique, it might just make it more difficult to composite the final image).

As I was shooting indoors at night for these particular images I also had to position my lights and get them set but I would recommend trying something with adequate natural light on your first attempt at this as lighting (and related shadows) adds further complexity.

Once the camera is all in place and set up correctly, you need to take your base image (this is the image that generally contributes most of the background to your final image. It is absolutely essential that the camera does not move after this image is taken. If it does it is easier to start again here (and reset etc) than strive on and try and ‘make do’ and fix it all in post (this can be done but it will take you three times as long!).

Once your base image has been taken you move your subject into place, in these particular images we asked the models to lean back (supported by the MUA) and I took a number of images with the models holding alternate legs up in the air. It is important to try and get your subject to keep as much of their upper body as still as possible while alternating legs for these shots so that the images match up better (which is easier to edit). As such I ended up with 3 final images for each model; one base image, one with the left foot in the air and one with the right foot in the air.

If you have a stool or other suitable platform you can get your subject to lie or sit and lean holding both feet in the air which will probably allow you to take less shots and make for easier compositing (just make sure that there is some clothing hanging over where their body is touching the stool or it will look unnatural when you remove it is post).


Once you have the images it’s time to load them onto your computer and into your editing software. As I use Photoshop for this kind of work I’ll be using that as a reference to describe the process but most editing software packages will allow you to do this in a similar way.

I loaded all 3 images (for one model) into a single workspace as separate layers with the base image at the bottom of the stack. I then decided which of the 2 model images I preferred for the body and face (in this example I chose the image with the model’s left leg in the air) and worked with a layer mask to hide the MUA from the image and the right leg (that was supporting the model).

Once this was completed I moved to the final layer and masked the bulk of the image except for the right leg in the air. If the subject has kept their upper body very still across the images you may be lucky enough that the resulting image matches up and looks natural.

As you can see in this example we were not so lucky (mainly due to the very high heels making it very difficult for the Maz to balance!) so I had to do some further tweaks using the move tool and a bit of the warp tool to line things up.

After I was happy with the parts of the image all looked in line and as realistic as possible I added some shadows and some overall styling to finish the image off. The other levitation image from this shoot was done in exactly the same way.

I hope this helps explain the process and gives you an idea of how to do a levitation photo. If you have any questions or want anything clarifying, please contact us.