Architectural AI image for magazine print
Stefano Casati is a London-based architect and artist. He AI-generated this Vegetable palladio villa for our customer magazine PAPER #6. Read about how he uses AI and and get some insights from production, prepress, and printing here below.
AI allows me to swiftly visualise complex ideas
I’m a London-based architect, passionate about the intersection of technology and design. At the end of last year, I began working with AI technologies. Personally, I find great joy in the ability to push the boundaries of creativity and idea generation using AI technologies. The limitless possibilities and the rapid visualisation of concepts with astonishing photorealism truly fascinate me. I view AI as my personalised Pinterest, seamlessly integrating it into my daily workflow. This not only enhances my efficiency but also enables me to explore uncharted avenues in tackling specific project requirements.
"I utilise AI to visualise and refine architectural designs, ensuring realism and practicality."
I also employ AI to generate intricate and artistic concepts, incorporating natural elements for seamless integration. My goal is to create visually appealing compositions that encourage viewers to reconsider established norms. AI has the advantage of facilitating the design process by introducing early-stage ideas for discussion. It also empowers individuals from all backgrounds to contribute their unique ideas effectively. Ultimately, the quality of the final image depends on the initial prompt or concept. AI allows us to enhance creativity and ensure that the resulting image reflects the original vision and potential.
Stefano Casati, @scasati
You can almost feel the texture
Stefano Casati's vegetable palladio villa contains a spectre of intricate details and shadowplay in green colour hues, just like many architectural and landscape images. On this close-up from the printed image, you can see that the details still reproduce well on the uncoated matt paper, as does the subtle change of colours in the sky.
How the architectural image was made and printed
Behind the image prompt
Writing "*/Imagine:" followed by a prompt initiates the image generation in the program used, and some of the AI prompts for making this image were:
- classical architecture
- architecture photography
- palladio villa
- vegetable structure
- structural kale
Format and sizes
The image was delivered in a JPEG. format, packed in a size of 64 MB, and opened up in 352 MB to a size of 740 x 870 mm in a 350 dpi resolution. It was offset printed in full page size 200 x 265 mm, on uncoated magazine paper Holmen TRND 80 gsm.
Preparing for print
To make this image look perfect in print on an uncoated paper, it's of course important to use the correct ICC profile. The colour profile adjusts the colours with respect to the paper's whiteness, brightness, and other properties. You will get the recommendation for which ICC profile to use from your printing company or your paper supplier.
But there are also some other details that are typical for images with landscapes and large buildings. Shadows and sidelight create some challenges, where prepress actions can improve the end result greatly, as you can see below.
The prepress perspective
How was this image treated in prepress? Did it come with any specific challenges? We asked pre-press technician Jens Jörgensen, OTW, to share his best tricks from his job behind the scenes with images for print.
Why do you need to work with the shadows?
This is a playful image, but its challenges are typical for many architectural and landscape images. Green colour hues are mixed with low-contrasting grey tones to define different sets of details. The building is positioned in sidelight against a blue-pink evening sky, causing both soft and dark shadows.
If the shadows become darker in print, details might go missing. If unprocessed, the image might give a slightly ominous impression in print compared to being seen on a screen where it's lit up from behind.
How do you handle details in shadows?
Contrary to what many believe, you should not enhance the contrast by making the shadows darker, but instead make the whole "problem area" of the image slightly lighter, so that contrast can be upheld, to create the best conditions for all details to show in print. Go a bit lighter, especially when printing on an uncoated matt paper because it consumes slightly more ink.
When finetuning areas in the image, be careful not to change too much. The main adjustments for the colour and contrast are simply managed by using the correct ICC profile for the paper and printing method.
What do you say about the print result?
This image is very interesting and I like how it turned out in print very much. Not a single pixel in this image is 100% white – and we have mentioned the low contrast and details as potential problems – but I would say they are part of the image's strength, and it all looks great. The tonal range provides a credible impression of soft sunrays near sunset in time. I was never particularly worried.
About the images
Learn more about how to prepare AI images in general to get the best possible reproduction in print.
See the images in print in our customer magazine!
Jens Jörgensen works as pre-press professional at content agency OTW in Stockholm, Sweden, with an avid interest in images and print for more than fifteen years. We asked him to share his methods for ensuring the best possible print result in three AI-generated images with different motives.