Change the style of an image in Gemini

This is a diaslide I took a long time ago during a vacation in the North of Italy. When I shot the photo I felt attracted by the Renaissance architecture. I kept the slide, had it digitalized and then it sat in the depths of my hard drive.

Knowing that you can have changed the style of an image in Gemini I uploaded the image to Gemini and asked it to transform the photo into a watercolor and ink image.

Here is the result. Gemini actually changed the style. I personally would prefer it to the original.

Using old Dia Slides

The origin of this image is an old color slide that was taken sometime in the 1970s, when slide photography was at the height of its popularity. Over the years, I accumulated a large collection of these slides and eventually selected a number of them to be digitized. Once the scanning was done, I stored the digital files on my hard drive, where they remained untouched for quite some time.This morning, I decided to experiment with them using AI image generators. I first uploaded the digitized slide to ImagePrompt.org, which generated a descriptive prompt based on the original photograph. I then copied that prompt and used it in NightCafe Studio and Deep Dream Generator to create new interpretations of the image.

Naturally, the AI-generated artwork is not an exact copy of the original slide—but that was never my intention. Instead, I wanted to see how AI would reinterpret an old photograph, and I really like the result. It has inspired me to revisit more of my vintage slides and explore how they can be transformed into entirely new works of art.

 

The dinosaur roars

I actually created my first AI video with Meta AI. It shows a roaring dinosaur. The video is based on a ballpoint sketch of a crocodile which AI interpreted as a dinosaur. Meta AI offers to animate your images, and so I accepted this offer and here it is:

 

The Man in the Fern

I like to create images with AI which tell a story. 
Looking at the image on the left with a man's face is embedded in a huge fern growing out of a bucket I wonder whether a human has been turned into the fern maybe because he participated in destroying nature and now nature is fighting back? 

Or we see an anthropomorphic fern crying out loud because of all the damage humans brought on nature?

You, the reader, are kindly invited to write your own story when looking at the image.

Skull and cherry blossoms

For a few days now I have been changing the way I create prompts for my AI images. I got a nice little notebook with colored pages from one of my friends, and I sit down and manually write my prompts in this notebook. I do this in the morning while sipping my first cups of tea. 

The image on the left is one of my first results I got while using this method. I wanted to combine an object with cherry blossoms twigs. While a skull stands for death or the end of life, those cherry blossoms represent spring. 

 

From a vector drawing to an AI 3D illustration

 Quite some time ago - actually in 2022 - I did some vector drawings using Inkscape. Here is an example: 

I must admit, it is not very good vector art. I should have spent way more time working on the image. But it is as it is.

Now I was wondering what I could do with it with an AI image Generator. 

The first step was to upload the file to imageprompt.org in order to get a reasonable prompt which I could copy and paste.

The second step was to go to Nightcafe Studio and create an image with it.

And here is the result. I had added "wearing a red scarf" in the prompt and changed the colors given in the prompt.

As Nightcafe Studio offers to remove the background of the created images for free, I could easily do that and use the owl as a design element in my Zazzle store. 

How to create AI versions of your Traditional Art

 Actually there are two ways of creating AI versions of your traditional art - with some AI Generators you can upload your scanned painting and then write a prompt like "create a digital painting using this image

There is however a second way to do it, by using imageprompt.org first. Go on this website and upload your painting, drawing, sketch or whatsoever. I uploaded the image on the left. It is a scan of a mixedmedia painting I did a long time ago. 

After having uploaded the image you can choose the AI Model you want to use the prompt with and then click on "Generate prompt".

The prompt I got was:

"An abstract acrylic painting of a stylized mask, centered in the frame. The mask features a symmetrical design with a cream-colored face, framed by vibrant red and pink elements.  Large, curved green horns adorned with yellow accents extend upwards from the top. Black almond-shaped eyes are set above a stylized nose. The lower face is green with a large red oval shape for the mouth.  Details include a small diamond shape on the forehead and the mouth, as well as red spots on the face.  The overall color palette is bold and contrasting, with greens, reds, yellows, and pinks dominating. The background is a mottled mix of dark brown and olive green, providing a textured backdrop. The painting style is flat and graphic, with thick brushstrokes, evoking a tribal or folk art aesthetic. The lighting is even, with no discernible shadows, highlighting the geometric forms."

Now you can copy the prompt and head over to your favorite AI Art Generator. I am using various AI platforms and in this case selected piclumen.com. You can see the result on the left. 

Of course you can always change the prompt, for example change the colors or the painting style.

Image to prompt - another way to use your old photos

 

The image on the left is an old photo, taken many years ago during a vacation in the Provence in the South of France.

I decided to play with it and uploaded the photo to generateprompt.ai in order to get a prompt for it. I got the following prompt:

"A historic Gothic-style church with an ornate facade and tall, pointed windows, featuring a tall, spire-topped steeple with a weather vane at the top. The church is situated in a European city square, with surrounding old buildings and a clear blue sky overhead. In the foreground, there is a fountain with a rectangular basin, and a few children playing in the square, along with two women walking nearby. The scene is bright and sunny, highlighting the architectural details of the church and the lively atmosphere of the square."

I then went to piclumen.ai and insert it. Here is the result:

Of course, the prompt you get when you upload the image can be changed - you might prefer a painting in a certain style, or leave out the fountain, or change the history setting. 

I am a traditonal artist and I love AI Art!

For quite some time I have been creating traditional art (mostly art done with alcoholbased markers) and AI Art. There have been days when I focussed on AI Art, and days when I focussed on traditional art. Yes, traditional art takes longer, I usually need one week for preparing the image, selecting the colors, colorblocking and then shading with inktense pencils and adding the background with acrylic ink. 

The image on the left however is the result of directly combining a traditionally painted piece with using AI to edit it. 

First of all I was not so happy with the result. It was okay, but does not belong to the pieces I am particularly proud of. So I was curious whether AI would come up with something more interesting when I asked it to turn the style of the piece into Art Deco style. And here it is!

 

The Beauty of Winter

No, winter in general is not my favorite season. I especially hate November with its short and often dreary days. December is better, with all the lights of the Christmas season. 

However there is beauty in winter - like in the image on the left. The crispiness of the frozen leaves in contrast to the red of the berries - looking at such an image makes me happy. With all those AI generators on the web I can easily create my happy version of winter!

Using a one word prompt

 

It's amazing what AI Generators come up with when you insert just one word as a prompt.

I used Microsoft Copilot and only the word "trick-or-treat". The result was the image on the left. It even had the words spelt correctly - which is something AI Generators had a problem with for quite some time. 

I must admit that I like being surprised - as much as I like getting AI doing exactly what I want them to do...

What to do with your AI Images?

I have been using my AI images for card design, for designs on my Zazzle store and I have been putting the images into frames and put them on the walls of my flat. Regarding the latter there is one problem: Here in Germany you do not find frames for square (1:1) images easily. But I have a lot of very nice AI images which are in the 1:1 format, and it would be a shame not to display them. 

Yesterday I suddenly rememberd that I have quite a few hard foamed panels. So I fetched one of them, grabbed one of my 1:1 images and mounted it on the panel. 

Fixing a thread on the back of the panel allows you to hang the panel on your wall. 

One tip: You have to avoid to apply too much pressure while cutting the board. Rather be patient and move the cutter to and fro as long as it needs to cut the board...

The knight on the rock

This is the original image, created using Nightcafe Studio. A knight in a worn armour is standing on a rock at the sea, with waves crashing in and rain coming down from a dark sky.

Same scene, different style: The style is watercolor now, much lighter, sun coming in, the sea is much more peaceful. 

This image was created with the "Edit image" tool of Grok.

I haven't done much with this took yet, but everytime I use it I am amazed by its possibilities.

Creating a comic book Page

 I love to create images with AI Art Generators, and I also love to tell stories, even in English, my second language. For a long time it has been eigher doing the one or the other. But now I can see combining those two. Here's an example:

I first created the image of a rearing black stallion in the desert, and then I thought: Why would this horse rear? Why do stallions rear? They do it for example when they fight with other stallions in order to prevent their mares from being stolen. 

So I thought of a short sentence which could introduce the reader to this situation.

I uploaded the image to photopea, changed the color of the text to white and inserted the text.

Then I exported the image as jpg - and voilà! My first ever comic book panel was created!

Two version with one prompt

Sometimes I like to work with several AI Art Generators and one prompt, and everytime I am surprised how different the results can be. 

The image on the right was created using Piclumen AI, which has a nice free version. You are getting reward points by uploading your images to social media (pinterest for example) and liking and commenting other people's creations.

The image on the left was created using Nightcafe Studio. Here you get reward points by showing up every day.

Marker Art and AI version

This is one of the two marker paintings I did in July. Only two? you might ask. Well, I focused on swatching all of my markers in July, and that took more time than I thought it would. 

But I now know that painting / drawing abstract art with alcoholbased markers plus shading with inktense pencils is the thing I really love to do, and so doing these swatches was an absolute must.

However I am still creating AI images, and lately I have been creating a lot of images with Grok.

Grok allows to upload one's own images and to create AI versions of them. 

As you can see on the right, the result is quite similar with a bit of change of color in some places. 

 

Creating Color palettes with AI images

 I am an AI artist and a traditional artist who focuses on abstract art. Colors and shapes are an important part of abstract art, and so I am always looking for color palettes. There are hundreds of color palettes on the web, but most of them are not the kind of what I am looking for. 

So I was very glad when I found coolors.co, a great online tool to create palettes from your own images (you can do way more with this tool, by the way) 

Here's an example:

The landscape image was created with an AI Image Generator. I uploaded it, and then I could select certain parts of the image which display the colors I would like to have in the palette.

This can be done easily with a free account; with a pro account you can get the hex numbers of each color, in case you need them for digital painting. 

Robin "The Cat" Hood

 

As you might know you can use Grok to help you with prompts. So I asked:

I am an AI artist and I want to create an image depicting a cat disguised as robin hood. Can you provide me with a prompt, please?

Grok answered with:

A whimsical, highly detailed digital painting of an anthropomorphic cat dressed as Robin Hood, standing in a lush Sherwood Forest. The cat has sleek tabby fur, a mischievous grin, and wears a green feathered cap, a leather quiver with arrows slung over its shoulder, and a flowing emerald cloak. It holds a finely crafted bow in one paw, poised confidently atop a tree branch, with dappled sunlight filtering through the leaves. The background features a medieval forest setting with soft mist, scattered gold coins, and a hint of a distant castle.

So I used this prompt and asked Grok to create an image. You can see the result on top of the page on the left hand side. 

Not bad, but I was curious and wanted to find out what Bing would do with the prompt. Here's one of the four images Bing created:

As you can see, there are similarities and differences between the two images. 

I just discovered Cogniwerk AI

 

There is a new AI Image creator online almost every day. I usually get myself a free account, experiment with the settings and then decide to either use it on a regular basis or only now and then.

Cogniwerk AI is one of my "newer" AI Image creators. To create one image costs one credit; every account has been given a certain number of credits, and when you have used up your credits, you can request more. 

Many AI Image Creators have difficulties with horses - they tend to create animals with five legs. I wanted to know what Cogniwerk does with a prompt like this:

A black mustang galloping over a yellow green meadow, vector style

Well, it actually added an additional leg where it does not belong. However they offer Inpainting, and so I managed to create the anatomically correct mustang in the image above. So up to now I am very pleased with Cogniwerk AI and will use it regularly.

How I connect AI Art and Traditional Art

The painting on the left was done with alcoholbased markers and inktense pencils. In order to get there, I combined AI Art and traditional art. In detail:

1. The first step was to create the image of a wooden mask as a reference image. I later on wished I had spent more time looking for an image which would serve as a good reference image (clear lines, good contrast, no shadowy parts and such). 
2. In a second step I downloaded the AI image and pasted it into my Open Office Writer.  What I did not think about was the size of the watercolor paper I wanted to use. I have A4 paper and I have A5 paper which is smaller. Open Office Writer is not very good in handling images (text processing software!), so I had to limit myself to A4 paper. The next time I will use Scribus, the free alternative to InDesign, in order to put the image to the proper size. 
3. I then had the file printed out as pdf and tried to trace the photo onto my watercolor paper. Not a very good idea. So I grabbed my tracing paper and used it for tracing the basic shapes with a fine liner. 
4. Now I could trace the result onto my watercolor paper and then continue with markers and inktense pencils. 

All in all it was a very educational and even joyful process.