Thanks. This workflow of 4 prompts, has the benefit of not using the mouse.
I have a friend who uses Photoshop to make posters for bands, the resulting images are better, faces of real people are put in the poster, but he does 1 million clicks every time. I use only Emacs to make the image, much faster workflow, more relaxing, i just edit text most of the time.
Gemini's image generation abilities, especially regarding typography, are in the same ballpark as Ideogram. Ideogram is a little bit better sometimes, vertical text for example trips up Gemini, but Gemini being native multimodal, works very well with text descriptions of images.
Ideogram has an upper limit to the total number of tokens it can accept as a text input. It is not native multimodal as far as i know.
emporas|9 months ago
I have a friend who uses Photoshop to make posters for bands, the resulting images are better, faces of real people are put in the poster, but he does 1 million clicks every time. I use only Emacs to make the image, much faster workflow, more relaxing, i just edit text most of the time.
Gemini's image generation abilities, especially regarding typography, are in the same ballpark as Ideogram. Ideogram is a little bit better sometimes, vertical text for example trips up Gemini, but Gemini being native multimodal, works very well with text descriptions of images.
Ideogram has an upper limit to the total number of tokens it can accept as a text input. It is not native multimodal as far as i know.