When people see one of my abstract/surreal images, they sometimes ask ‘How did you do that?’

So, in this post I unpack some detail about the process of creating one of my Photoshop Creative readers’ competition entries. This is not a tutorial (I may produce some of these in the future), but rather I’ll show you some key points in the process of generating this artwork.

The challenge!

Each month Photoshop Creative magazine holds a competition for readers. Recently the challenge has been to take four random pictures that the magazine has chosen and turning those four images into a creative piece of artwork. You can use your own images if you like, and you don’t have to use all the images. However, I prefer if possible to set myself the challenge of trying to combine just all four images supplied and manipulate them in some way to make an image that makes sense.

The original images

Below are the four originals that were supplied – this shows you the entire image area that was available:

 

fouroriginals125

1. Brainstorming

I spent some time considering how best to use the images, their quality, size, subject matter so I could consider how I could combine them to make something that worked. Creating something which included use of the four randomly supplied images proved to be quite tricky, but eventually it made sense to feature the model – who, looking right at the camera, made me think of a selfie taken on an iPad, and build the other elements into the background, creating something a little quirky and ‘arty’ as if sketched on the iPad – hence ‘Painted Selfie’.

2. Where to start?

So as I often do, the first task was to isolate the model and manipulate the image to take it from a quite ‘flat’ photo and make it more dynamic, as well as enhancing it a little. To do this, I added quite a few subtle touches as well as more obvious manipulation using layers and blend modes which you can see in the following four images.

The first (top left) shows the original image, the second (top right) shows the first stages of giving the image more contrast and greater depth. The third image (bottom left) shows further blend modes added, and other manipulation include giving the hair more volume on the right-hand side of the face and extending the hair down.

You’ll see that using the blend modes has washed out the detail in face a little too much. Image four shows more detail filled back in on another layer, completion of the hair extensions and other small details such as lightening the pupils in the eyes and finally roughly cutting the model out. The cut-out didn’t need to be too accurate as I knew the ‘arty’ background would be rough so no need for an extremely accurate cut-out. Sometimes its worth planning ahead like this, because knowing what you want to end up with can save you time in the construction process.

 

girl-2-process-fin

3. Changing the perspective on the iPad image

Next I needed to ensure I could convincingly change the perspective on the iPad image to make it flat – or straight-on – to match the model as she is straight-on to the camera. It was relatively easy to do this using mainly the warp, perspective and scale tools. The following two images show the before and after images.

 

Ipadmanip

4. Creating the ‘arty’ bit!

Using the landscape shot added in the background and painting with several different kinds of Photoshop brushes and colours, blend modes and layers I created the ‘arty’ background. The following four images show the basic progression. I knew that I wanted quite a strong frame to surround the model as if painted in – so I created that. The final feel I wanted was also to be quite grungy and rough, but also with some recognisable images.

You can see that as I built the background I masked the area for some of the layers where the model would be placed. It took some time to make the multiple blend modes and layers work together in the way I felt happy with – you can see how the colours are changed as more layers and modes are added in the progression of pictures below. Some of the blend modes would also ultimately affect the image of the model.

 

brushesborderconstruct

5. Sometimes you have to experiment

I had fun spending time experiment with all the blend modes and how the altered the image. The pictures below show two alternatives I wasn’t happy with. This was the stage at which I also started to look at blending in the multi-coloured paint drips / splash image too. To help blend this image properly, I ended up ensuring that just a few of the paint drips interacted with the model. You can see this subtle touch on the final image.

 

couple-of-fails

6. The chosen ‘arty’ bit joined with the model

The image below shows the version of the background with the manipulated version of the model image dropped in. After this I also blended in the multi-coloured paint drips / splash image.

 

finalgirlvers

7. The final image…

In the final stage I added in the manipulated iPad photo and further lighting and shadow/fixes and enhancements to make the image complete. This I usually do by eye and keep going until I’m happy with the overall general effect of what I’m trying to achieve. Once all the manipulation was done, this how the final image turned out…
idea3FINALwinner

 

Phew! Hope you enjoyed travelling this journey of creation with me. Thanks for reading! Comments welcome below.



Other posts: