Starting off with a picture, the Hybrid Pixel project took on a challenge to experiment with the ‘flip-flop’ phenomena, to explore the pixel in its many forms, by bringing its creative process back and forth between the physical and digital worlds. I wanted to explore how pixels changed as they were stretched into different forms, and how I could use all of that to push boundaries in being more creative as a designer. Let me take you on this journey:
As a designer and photographer, I believe that photography is a great example on demonstrating how the relationship between physical and digital happens. It's pretty much instant and fascinating: You take a photo of something that is physical, and then you get a representation of whatever was in front of your lenses, digitally on the screen. Isn't that interesting? However, most of us stop there - we tend to keep our pictures inside the digital world, editing with some filters and posting on our social medias to share with our friends. But we don't have to stop there. In fact, what this project shows is that the more you continue to push the pixel between physical and digital, the more chances of getting incredible - unpredictable - outcomes.
I was like, what if I could represent the pixel - the small squares that light up and make up the content on your screen reading this - into something 'tangible', something that could bring awareness of this whole relationship between the physical and digital?
Looking at a very cute embroidery frame hanging on the wall, something sparkled in my head - the fabric sometimes looks like a grid - just like how pixels sit next to each other. What if I learned to stitch pixel-like patterns on top of the photography I make? Or even better! What if I could get a computerised embroidery machine to translate the pixel data into stitches on fabric? - I saw a very interesting potential there.
First of all - I am a photographer and designer. I had never had any experience with embroidery. But I felt like I should learn something new and try. I decided to learn the basics of embroidery and then learn a few different styles. It turned out to be quite a few hours of practice.
After failing and succeeding a few times, I was quite happy to start my first hybrid pixel prototype. The two embroidery styles I liked most were the needle painting, for the ability of gradually change colours; and weaving which reminded me very much of a standard pixel layout. This process involved these steps:
1. Taking the photo (physical)
2. Transferring it to a computer (digital)
3. Printing photo on transfer paper (physical)
4. Embroidering on top of printed photo with matching colours (physical)
5. Scanning final outcome (digital)
At this point I wasn't too worried about getting the pixel and colours perfectly right. The purpose of this prototype was to analyse each photo by its pixel/colour, bringing the digital to the physical by matching a few predominant colours with threads. The results are interesting and speaks about the limitations and advantages of physically threading a 'representation' of digital pixels. The style of embroidery, as well as the scale and amount of colours cannot truly imitate what a screen looks like, or how well a print imitates a screen. However, the physicality and tangibility of the thread emphasises colour and shape, almost resembling the form when we put a magnifying glass in front of a screen; which makes us suddenly more aware about the 'invisible' pixel and the digital process that created the physical picture.
After hand-sewing for quite a bit - really, it is not easy if you are not experienced - I began to take the idea of a hybrid pixel further - if hand-sewing cannot really copy how the pixels look in a screen, maybe a computerised embroidery machine can. My first attempt started differently - instead of taking a photo, I would scan an existing photo back to digital and see how its data/pixels were translated into the machine and then onto embroidery on linen. I was fascinated by the results:
Choosing just s small portion of the photo, the image went through an interesting process that distorted colours and reshaped lines to fit into the rules of computerised embroidery. With over 8000 stitches and just 10 colours, the machine shifted the pixels around to the best of its ability in the attempt to create a replica of the scanned picture. The result is unexpected, with 'strokes' and unusual angles that resembles of an impressionist painting.
The more I used the embroidery machine the more I realised its closeness to how a screen display its pixels. The outcomes of any embroidery, although not fully predictable, was still bond to the rules of a chosen setting:
If screens have their colour resolution and pixel definition, embroidery have number of colour of threads as resolution and run pitch/fine sewing as definition. With these similarities, I decided to take three photos of sceneries in nature around the lake Burley Griffin in Canberra, and by bring it to the digital, explore how I could stretch and tweak its setting in the attempt of getting a range of outcomes - from a closer representation of the original subject to the most abstract possible. This process involved these steps:
1. Taking a photo (physical)
2. Transferring to a computer (digital)
3. Printing onto linen (physical)
4. Transferring the photo to an embroidery software (digital)
5. Transforming the photo into an embroidery pattern (digital)
6. Selecting thread colours that the software has identified in the picture (physical)
7. Aligning and then sewing onto the picture on linen (physical)
8. Scanning all work (digital)
The first attempt was done in a smaller scale, using this picture I took of a female King-Parrot. The process involved matching measurement/scale in both digital and physical, to create the a perfect juxtapose effect. To choose the colour of threads, I was also limited to what I physically had - which in this case was an insane amount of colours and shades but still matching was up to my eyes and its ability to match colours - and to what the software was suggesting. As you can see in the colour chart picture above, each steps alter one colour pallet from the other, which makes the whole outcome even more uncertain and exciting.
The second attempt was on a larger scale. The image of the moving waters of Burley Griffin lake was professionally printed on linen in A3 size (297 x 420 mm) as three samples of the image were extracted to go through the "flip-flop" process. The first sample was made with a cross stitch embroidery style with 8 different colours. When the software deals with less than 10 colours, the tendency is to have a few colours replaced with a single one regardless if they match each other. It was the case of the predominant 'teal' colour that took over most of the space although in relation to the original picture, it had little to zero of that colour in its pixels.
The second sample was made with a punch embroidery style in 10 different colours. It is interesting to see how the digital embroidery path picked colours such as 'lavender' and a 'light pastel green' amongst the many blue shades the original picture has. The outcome took over 5 hours to be completed with over 47k stitches. This sample focused on covering the shapes of the ripples, waves, shadows and lights, while leaving some negative space between sections, allowing the contrasting with the colours of the picture printed on the background.
The third and last sample was an attempt to create a picture with just the RGB colour values of the pixel. The software was capable of analysing where red, green and blue was used most, and create paths that illustrate its use. It is no surprise that blue is the most distinctive and used colour - but green and red also make a few remarks.
The third attempt was also printed on linen in A3 size (297 x 420 mm). I took this picture of a beautiful blossoming tree thinking about the challenge in its colours, shades and details. I wanted to push myself to get the closest representation of the original subject as I could. Taking two samples of this photo, I wanted to contrast, first with a broader, less refined option and the other being the one with the most details and colours of the entire project.
The first sample is an interesting 4-colour embroidery style that used each colour to fill all the spaces, only using the order of colours and thickness of pitch for emphasis. Although the darker colours have much more contrast than the other two brighter colours, the beautiful precision of thickness and overlap of colours accentuates the shapes and spaces in the sample.
The second sample was the most challenging of them all. With over 30k stitches and 18 different colours, the embroidery machine took on the biggest challenge of placing each colour precisely in its place. Unfortunately, the very precise process of aligning the embroidery with the picture suffered minor errors that are unknown and as seen in the picture do not perfectly align to the background image - the measurements were all doubled check and the alignment before stitching was done exactly like the others. However, the result has beautiful definition of a higher amount colours and finer stitching showed a closer representation of the original subject that was not really seen on the other examples.
To me, the Hybrid Pixel project became an incredible catalyst to explore the incredible creative potential in putting a process through physical and digital. It is not something completely new, and we do it all the time - we are just not aware of it. Creating a tangible 'representation' of a pixel made me stop and think about the processes that were involved and how they were naturally changing by simply going through the physical and digital and vice versa. The idea is that you become more aware of the processes that you take in your own practice, and how you can continue to push boundaries to bring the 'unseen' into something creatively unexpected.