Sometimes I want to “dress-up” an otherwise decent photo to create something special. Of course its always better to have a strategy for this at the time of the shoot, but often one doesn’t have time, or resources, or one re-visits an image and just gets a better idea. Such was the case with this image, which was originally done for a video game package design that I was not in charge of. You can find a blog post about the original shoot here!
My new thought was to place the dancer into a fun high-enegry environment like a wild disco. I didn’t have any wild disco shots, however I did have some interesting images taken at the Neon Photo Walk in Las Vegas. I settled on a slow shutter Lensbaby shot taken on Fremont Street.
The next task was to mask the subject to isolate her from the original background. First I use the Magic Wand tool & click in the background to quickly select everything but the figure.
Now I choose Inverse from the Select Menu & then entered Quick Mask mode to fine tune the selection.
Now I like to change the Quick Mask color by selecting Quick Mask Options from the Channel Panel’s options fly-away menu at the upper right corner of the Channels Panel – the resulting dialog lets you pick a color and an opacity to represent masked areas. In this case, I picked a color similar to the dominant color of the new background image so I could more easily see how the edges blend in the mask.
There are almost always little defects in the selection that can be easily painted out – in this case small areas inside the figure are being masked out and simply need to be painted over with white in the Quick Mask. For this subject, I only concentrated on the hard edges and pretty much ignored the hair, because I will use a different method of blending to get hair wisp details.
Once I have the selection detailed, I duplicate the original to a new layer and click on the Refine Edge button to finish up the selection – you can see a more detailed step-by-step tutorial on this process in my YouTube video here.
I simply set a 10-pixel radius for edge detection in the resulting dialog and then clicked “OK” – again I was completely unconcerned with the hair. The result is shown below:
So, right now I have two duplicate layers of the dancer – the top one is fully masked, the bottom is the original unaltered layer – I’m going to use this one for the hair wisps. The original bottom image is shot against a plain gray background, I now need this gray to become as close as possible to a level 128 gray. So I make a curves adjustment layer and make sure its clipped to the underlying layer. I then adjust the curve so that the background tone right behind the hair reads 128 in all three channels.
Next I select all three layers…
and then select “New Group from Layers” from the Layer Options fly-away menu…
The Layers are now contained in a “Group” represented by a little folder icon in the Layers stack. This Group can be treated as if it were one layer!
Now simply drag the Group out of the figure document by click-dragging, from inside the window with the Move tool, up to the Tab for the new background & down into the new Background window.
Now I select the bottom figure layer (the one without a mask) and change the Layer Apply Mode to Hardlight, using the blend mode drop down menu.
The gray background of the bottom figure layer completely disappears where it is exactly 128 – there are a few spots at the edges where it is subtly visible but these can be masked – we only need to preserve the parts of this layer where the hair lives!
I now select the bottom figure layer, and make a layer mask where I can paint with black to mask off the edge and leave jsut the part with the hair.
Now I select the top figure layer, and because I don’t want to mess with the existing mask (because I want to protect it from mistakes), I select “New Group from Layers” from the Layer Options fly-away menus at the upper right of the Layers Panel. This puts the layer with its mask into a group that I can attach another mask to. Now I simply paint out the hair wisps with black in the group mask.
The result is a completely seamless blend of the very fine hair wisps with the new background. I show a very detailed step-by-step video tutorial on YouTube here.
Next I use the mask of the top figure layer (inside the group) to create a mask fro and adjustment that darkens the background in a way that suggests the type of shadow effect you would get from the drag shutter effect. When you use a slow shutter speed with flash front light against a continuous light background, you get a shadow where the moving subject is displaced during the exposure.
Next I add a motion blurred hair layer underneath everything as an additional “hair shadow” to help sell the effect. To help separate the dark hair from the background, I lightened the background & cooled it off a bit with a curve adjustment.
The Layer stack looks like this in the end.
Here is the completed Image!
I hope this gives you an idea for your own images. I now prefer to shoot on a medium gray background, whenever possible because it gives me so much flexibility with how I can treat the background. This image combines two scenes that were shot in completely different ways with radically different lenses but somehow manages to look fairly convincing. Although a real drag-shutter shot can be awesome, it is very difficult to control and the best shots are often happy accidents. Using this approach you can craft the image exactly the way you want it with complete control over all the elements.