For the Living Gallery website I wanted to have a small logo for a long time, something with a mountain or so. But since I’m neither good at drawing nor designing, all attempts remained fruitless on the hard disk. By the appearance of many artificial intelligences, which can also create images, I had the idea to try it with an AI.
The AI as a draft generator
After some research on what a good prompt for a logo generated by an image AI might look like, in the end only five words were enough to get acceptable results: “logo, icon, flat, mountain tours”.
I entered these words in the prompt of Stable Diffusion and with a little luck I got the following results, which excited me quite a bit:
Selecting the result and improving the image quality
I really liked the last of these results in particular. The mountain range is exactly as I imagined it and I also found the rest of the design very appealing. Who or what Flaan Tours is, however, only the AI knows. Even a Google search found nothing about it.
Probably it is because of the free version of the AI but the images have only a relatively low resolution. The advantage of a logo is its abstractness and the clear lines. So I used a tool to convert the bitmap into a SVG vector graphic and got a much better version. Here is the before / after comparison:
Further refinement in Photoshop
Next, I opened the image in Photoshop to refine it further. Using basic tools like the Selection tool, Pen, Brush, Lasso tool, Paths and with the help of transformations, I adjusted the parts that bothered me the most. The final result looked like this afterwards (the term Photo Tours is just an example here):
Conclusion
I think the result is quite impressive. I’m not sure yet if I’ll use this exact logo or parts of it for the website, but this little experiment showed me how quickly you can use AI to create a draft for further processing. It’s a very creative process and you get usable results incredibly quickly.
However, I wouldn’t expect more than a draft here either. For image processing, for example, I wish I could process my own photo in a certain direction determined by me with the help of AI, whereby the AI always gives me creative inspirations. So basically an idea generator, whereby I continue to determine or refine the creative process myself and thus reach my goal faster.
Originally, I also wanted to generate the text you are reading here with the help of an AI. But what the AI spit out sounded too much like canned and not individual. I gave up and wrote the text by hand. For the translation into English, however, I used the AI from DeepL 😊