I submitted Assignment 4 a few days ago but haven’t yet gone into any detail on my production process. There is a reason for this! The whole construct of my images is based on embedding text (names) into otherwise normal scenes, and I wanted to find out if viewers (including my tutor) had worked out how it had been done:
- Was the text already there?
- Did I vandalise public places?!
- Or was it done in post-processing?
One of the responses I got in peer review was:
“I think it works very well. Am fascinated to know whether you inscribed all the names or whether they are what you saw when you went there?”
Also, even my tutor who looked at A3 prints thought that one of the images (Flowers) was real text that I had taken a photo of.
The actual answer is… all of the text in the names in the images is Photoshop work.
Before I go into that part, I’m going to rewind a little and explain the overall production process.
Research and planning
Once I’d settled on the concept the first practical piece of work was to compile the names of the 86 victims that I intended to honour in the images. A number of different newspaper sources had partial lists, some with first names only, some with initials, some with alternative surnames (a fairly common occurrence in France where women sometimes use married names interchangeably with maiden names for different purposes).
One thing that the newspaper reporting helped with was that they grouped together the victims that were connected either as family or groups of friends – something that I subsequently built into the final work (explained later).
The breakthrough came when I discovered that the mairie (mayor’s office) in Nice had put up banners to commemorate the victims. This became my official reference point to check the other lists against.
This featured the initial 84 victims but two further people died later from their injuries, one that was named and one I had to include as ‘Anonyme’ (anonymous).
Once I had a finalised and cross-checked the names (a surprisingly emotional experience to be honest – knowing that 86 people died is one thing, reading biographies of 86 individuals is another) I made a master list so I could work out which names needed to go into which photograph.
The family and friend groupings was the main method of allocating names to images – I wanted to present together the people that knew each other. It’s a small detail, and one not obvious to most, but I felt it to be more respectful than randomly scattering the names.
The secondary consideration in allocating names was that the images each had a slightly different number of text opportunities. Dividing 86 by 6 meant an average of 14 or 15 images per image. Some images had fewer or more spaces (Boats only accommodated 11 names, Postcards took the most with 17 names).
The final practical point was that some spaces allowed for long names and some only for shorter names, so this was a way in which I tweaked the naming allocation as needed.
I had various ideas for usable scenes before I narrowed it down to a shortlist. The criteria were:
- A touristic, representative scene of Nice that would not look out of place in a calendar
- Where I could embed names into the image in a subtle way
A few ideas met the first criterion but stretched the second as too contrived:
- Shutters with graffiti on the slats
- Beach umbrellas with names stitched into the fabric
- Shop/restaurant signs
- Wine bottle labels
I eventually settled on six ideas, three of which I had candidate images (or at least test images) for at the point of shortlisting and three that I needed to go back to Nice to shoot – more detail here:
- Beach chairs
- Bakery / cafe / restaurant
This last one, the pebbles, didn’t make it into the final cut – it was replaced by Palm Trees. What’s interesting about its exclusion is that it was the only execution I did where the text was real – I actually wrote names onto pebbles on the beach with a marker pen and photographed the result.
Ironically, the only image where I had done the names ‘in real life’ is the one that I felt looked least realistic! So I’m glad I took the palm trees shot as a fallback.
I am not generally a big user of Photoshop. For general processing I tend to use Lightroom as it has enough controls to tweak images in terms of optimising colours, fixing highlight/shadow issues and simple cloning out of unwanted elements. I only use Photoshop when I really need to – and I really needed to here!
For reference, here are the original six images before Photoshop work (click for larger):
For each of the selected images I followed the following workflow:
- General optimisation in Lightroom per my usual post-processing routine
- Exposure, curves, little bit of vibrance boost, white balance adjustment etc
- Until the image looks just how I want to it to as a ‘pure’ photograph
- Import into Photoshop
- Spot healing and cloning work to remove distractions
- Small (dust spots, cigarette butts)
- Large (bicycles, people)
- If replacing existing text – remove existing text
- Via a combination of spot healing and cloning
- Apply new text layer
- Matching font as appropriate to the context
- Various layer adjustments in order to make the text look as realistic as possible
- Decreasing opacity, applying patterns, adding drop shadow etc
Examples of the last three steps can be seen below.
I just had to do that 86 times…!
One more post coming on this assignment – a roundup of the visual and conceptual inspiration and research into related projects to place this work in a wider context.