A friend of mine and I had often joked about inspirational quotes and life coaching popping up on social networks; at the idea that it’s a magic bullet in and of itself. We’d tweet brief exchanges in the Drunk Hulk mould of caps lock, exuberance and typos, having a good chuckle about it each time we had an episode of what we termed LIFE COUCHING.
Back in January this year, it occurred to my good friend there was some mileage in developing this character as a parody Twitter account and thus @LIFEEXSPURT was born. Recently named as Davey Mellencamp, our “exspurt life couch” has made topical observations, offered crucial life coaching tips, and has promoted his range of services.
Although we hadn’t discussed it in much length, I offered to make an avatar and graphics pack for the parody account. I wasn’t sure how to approach the profile banner and background for the Twitter page, but I had an inkling of the avatar in my mind’s eye.
Facing the avatar
Davey follows a few choice Twitter accounts which, along with his wisdom, gave me the shape of his character and the people he finds interesting. I felt that a range of successful and famous people of the sort Davey would admire would provide a bed of references for how Davey himself would look.
I wasn’t initially sure how I would build the face—whether I would take the nose from one person and the eyes from another, for example—but I wondered what sort of result I could get from layering all the faces on top of each other. Could I make a whole new face from combining all these other faces?
I needed clean and ideally full-face head shots, and I wanted the photos to have a licence for remixing where possible. Wikipedia was my natural first port of call and thankfully I was quite lucky with what I found, both in terms of usability and Creative Commons licensing.
Many of the images were of suitable resolution and quality, to the point where the images of people I’d like to include which weren’t so suitable could still be used to add texture, and have an influence on the final result.
Making a face
I started with a 1024 by 1024 pixel canvas in Photoshop and imported each image, making sure they were all Smart Objects. Whilst Smart Objects ramp up the document’s file size you benefit from non-destructive scaling, referring back to the original image data if there’s a need to do more than one transform.
I marked out an oval shape layer in the document for the approximate size of the face I was looking for, and then positioned each layer so the faces fit within the oval. I dropped the opacity of each layer to around 8% and added a multiply layer blend in order to see all the faces merge together without the mid tones and shadows dominating the composition. From there I could mark out with guides where the eyes and mouth would land, and then drew a very rough outline of the main features of this new face. I would use this sketch to help align facial features to help bring the face to life, starting with the eyes.
Here’s the result after the eyes are aligned. It’s remarkable how a new face is already forming after only aligning one element. At this early stage the image is full of peripheral information and yet the brain can still make something new from it. I was energised by this, and it became clearer how I would continue. I drew another slightly more detailed outline of the face, picking out and deciding on the features. This would help me align the nose and mouth.
So now there’s a base layer group for the eyes and face as a whole, and layer groups for the nose and mouth respectively. From here, I copied and flattened those layer groups so I could do some fine tuning, largely involving tone blending and masking. The one thing interfering with the quality of the face at larger sizes were spectacle frames from a few of the source images. Using a combination of the clone tool and spot healing tool, I painted out signs of the frame from around the eyes, brow and nose. The details around the hairline and ears were left in as they didn’t break the spell enough and contributed to the textures outside of the face.
With the facial features done, I was impressed how effective the image was, particularly at the smaller sizes. A real but made-up face with a ghostly periphery. One thing wasn’t sitting right though: I needed to align the ears and include those too. By masking those into the final image the eyes were framed better and there was a stronger line around the jaw.
It’s a fascinating technique to work with and I’m very tempted to try it out again with a different set of photographs to see what comes together.
Prologue: Tongue in cheek
The humour behind the avatar is several layers deep (literally!) and not obvious without explaining, but I wanted the result to be straight and look genuine. It was another matter when developing the rest of the graphics for Davey’s Twitter profile.
The aim with the profile banner and background was to overload on the earnestness but to have a rough and home-baked look. As such, it would be more believable if the result wasn’t that of a professional. Going against my instincts of fit and finish was a touch uncomfortable but dammit I was going to have fun with it.
Davey’s profile banner is a representation of his ego, his third eye and his particular brand of blue sky thinking. The background follows a lead taken by companies, brands and social media experts by using it as another method of communicating as opposed to a pleasing image or texture. Davey’s approach is communicated through a selection of keywords such as dream, inspire and motivate, and he’s also made the wise choice of picking one of the friendliest typefaces available in Comic Sans.
Postscript: References
- Alec Baldwin
- Ben Stiller
- Billy Crystal
- Ethan Suplee
- George Clooney
- Jay Leno
- John Travolta
- Matt Damon
- Paul McKenna
- Ricky Gervais
- Seth MacFarlane
- Stephen Fry
- Steve Ballmer
- Steve Jobs
- Tom Cruise
- Tom Hanks
Note: Tim Robbins was also part of the reference haul and whilst his face was used to help form the nose and mouth, he didn’t form the initial build of the eyes and face.