Me and Umar’s end product is a race car driving through the mountains at sunset. We had lots of different iterations before we arrived here. We initially wanted to do a race car, but were thinking of having a race track in the foreground, and have a city landscape in the back, similar to an F1 race. However, this was not easily feasible. We drew inspiration from Cars, and tried using a picture from Radiator Springs. However, the nodes were not clearly showing a race track nor the iconic rock structure.
From there, we pivoted to real life. We tried having the background be the skyline of Tokyo. However, even with a fairly simple silhouette, Sangala wouldn’t trace it properly. At this point we pivoted to nature. We also moved the car to be the foreground, instead of a race track. We put mountains in the back, and put a skyline, having some clouds and the sun. We traced the mountain, which was easier than finding a simple and usable image that would work on Sangala. We also added a dust cloud to show how fast the car was going.
We have been finding images online. The car was a race car outline, and the sun and clouds were clip arts that we put into Sangala. We are now left with a race car driving down the mountains at sunset, which is far from where we initially planned. However, we are happy with our current results and look forward to seeing what it becomes!
This is so cool–You opened up a world of possibilities for me, thinking about how AI can disaggregate the images for us. I tried to use your promt to recreate my own image–and while I don’t like it as much as my own, I feel like I could have some fun with it, I am pasting my ChatGPT image so you can see the difference.
I wanted to share a bit of the process behind this piece! Willimiah and I worked on this one together through Sangala Studio, using AI prompting to generate and separate out the individual image elements (the panda, bamboo, blossoms, mountains) before combining them into the final composition.
We landed on a red panda as the centerpiece almost immediately (honestly, just because of how unbearably cute they are). From there, we started building out the world around it. A bamboo forest felt like the obvious setting, and we loved the idea of scattering cherry blossoms through the stalks to soften the scene and add a pop of color against all that green.
For the background, we wanted layered mountain scenery to sit behind everything, partly for depth, but mostly because we wanted to nod to the natural habitat red pandas actually come from in the wild. It felt important to ground the design in something a little more authentic rather than just a flat backdrop.
Really happy with how it turned out and please let us know what you think!
Hello again! Here is our finished collage. All went rather smoothly with die cutting and assembling (minus some smeared glue), and we pressed the glued collage between two thick pieces of rubber cutting boards to achieve a flat final product. The picture was taken using the photo booth in the classroom across the hall from ours, and allowed for a very clear, almost PDF-scanned looking image. A few notes from our experience that you may benefit from:
Only use full or nearly full sheets of paper in the die cutter, especially when using a worn rubber mat underneath, otherwise the paper will end up slipping and not cutting your desire shape.
For our more basic shapes, we opted to cut by hand rather than using the die cutter, as we found that to be more efficient and were able to make small changes without refiguring our designs in Sangala.
Try to not place images within a 1-inch(ish) boarder on your Sangala design sheet, as the die cutter may have trouble picking up your image.
Play around with colors! Obviously not all of the colors used in your digital collage will be available to you when assembling your physical iteration, so keep that in mind when creating your drafts and be open to different hues, shades, or entirely different colors in your physical collage.
Here is our finished collage! Overall, we are happy with how it came out, but have a few lessons learned to share:
If there are any layered background elements (we had several layers of hills/shrubs), give yourself some buffer so you don’t have to get the alignment perfect right! I would just add a rectangle to the bottom of the element in Sangala.
We printed out two of every element in our collage, figuring that we could either make two collages or have a backup in case we messed something up. Having two of each element solved our alignment issues mentioned earlier, as we could cover up gaps with the extra copy. That fix created a really cool layered effect in our final product.
We ran into some issues with shapes that went all the way across our page, it seems as if the die cutter is not able to cut past a certain point. For cuts that went all the way across our page, we oriented our parts vertically in Sangala.
I began by sourcing a reference photograph online of a buffalo standing beside a bird, which served as the visual foundation for the piece. To simplify the image into a shape I could work with for cutting, I imported the photo into Google Gemini and prompted it to generate a cartoonized, black-and-white version of the buffalo and bird. I refined the prompt several times, adjusting the wording based on each result, until I arrived at an output with enough detail (like the horns, ear, and eye) to make an interesting die-cut.
I then sourced additional reference images online for the rest of the scene: rolling hills, ground texture with pebbles and debris, tufts of grass in varying shapes and shades, a mountain silhouette, a rippling water pattern, and an acacia-style tree.
Once I had all my reference images gathered, I imported them into Sangala Studio, where I traced each subject individually using the software’s background removal tool to isolate clean outlines. Each traced shape was then sent to the die cutter and cut out of different colors of cardstock — purple for the trees, tan and brown for the hills and ground, green for the grass tufts, gray for the buffalo, teal for the water, and white for the egret.
Finally, I assembled the collage by layering and gluing the individual cardstock pieces onto a white cardstock background, building the scene from back to front — sky, trees, hills, grass, ground, and water — before placing the buffalo and bird in the foreground.
This is our printed out copy. We followed our draft pretty similarly. We changed some shades of the mountain based on the paper that we had. We also got rid of the browns in the skyline and replaced with reds and oranges to keep up with the sunset. We added snowcaps to the mountain peaks, and added some lines on the mountains to add depth and show the separate parts of the mountain, attempting to make the image more realistic. Overall, we were able to implement the plan very well, and are left with a finished product we are happy with.
We started the fabrication process with one of us starting from the topmost layer and the other starting from the bottom. After organizing and printing each layer, we worked together to accurately match the layers to the Sangala Studio rendition. Once we took a picture, we glued it together as closely as we could.
What worked well was splitting the workload to be as efficient as possible. We also both really enjoyed the design we were creating, even if it led to a bit more planning than we first thought. For example, we added the sun in the background to complement the dragon. Originally, we wanted lines surrounding the monster but ran into too many technical difficulties.
While gluing, we noticed that there were small pieces of white space due to inaccuracy. We thought that minuscule adjustments for the wave layers wouldn’t matter, but the whitespace was really noticeable. So we had to reglue a few times. Additionally, we purposefully cut a few pieces larger than the background paper. So our last step was cutting the edges of the paper to fit on the page.
