Text to Image Prompting: How-to Create Eye-catching Generative Lions
I present another romp around the digital plains with my trusty steed Midjourney. This trip to the animal kingdom today is a MJ V4 wildlife Safari. We will be working with some basic prompts to create eye-catching lions. If you use different text-to-image generation software like Stable Diffusion or Canvas this prompt series can still be helpful for composition. I created this how-to guide for people who are interested in learning to use Midjourney.
My day started out with some development work on the upcoming Midjourney V5 platform. The MJ team have asked subscribers like myself to help train the next version of the text to image platform. The training is really an easy job, we just look at several images and select the best one based on our own ascetic preferences. While I was scanning hundreds of images I came across a pride of lions and they jumped right into my imagination. So without further ado lets go on safari to find the Lion King.
I wanted to start the series of images with some lion samples that are photo-realistic as well as easy to isolate, so I use the prompt: professional photograph of running lioness on white background. --ar 4:3 --q 2 --v 4 --uplight
I use the words 'professional photograph' because other variations such as 'realistic photograph', 'photo-like', or '8k' tend to get poor results in my experience. I'm using the adjective 'running' because it gives the subject more visual interest for the final image. I then go on to add 'on white background' to isolate the lion from the rest of the image, this is good step for graphics design reasons, as the lion can be easily cut and pasted into different images or used readily in advertising. Finally I've added '--ar 4:3', this sets the aspect ratio for the project.
The '--q 2' tag lets me know I am generating the highest image quality possible. This can be changed in your Midjourney settings and is not part of the actual prompt. Lastly to be sure it says '--v 4 --' this lets me know what version of Midjourney I am rendering the image with. You can change the version by typing '/settings' in the Discord channel your working with MJ in. Then you simply select the version that you prefer.
With version five of Midjourney being available now I would recommend using version 5, you will get better results with legs and paws or hands and legs on people. Thanks for dropping by the pie. If you like our articles plz share,