'Tis the season
Has anybody tried https://torqsightlabs.com/ to do this kind of thing?
Happy Halloween
The King in Yellow
Midjourney v5.2
Lord of bats
Midjourney v5.2
Looking for meaning
Midjourney v6.1
That would make an awesome gameOriginally Posted by Arch Fenster
The King in Yellow
Midjourney v5.2
Ho ho ho LOL
Awesome stuff guys!
Because obviously there is nothing more artistically meaningful than Beethoven playing chess with Death in a stone corridor
Except maybe Shakespeare enjoying ice cream
Last edited by LordBooford; 14th Nov 2024 at 16:46.
Let thy tongue become one with ye ice cream, and lo! Ye sixth finger shall assuredly spring forthExcept maybe Shakespeare enjoying ice cream
You do realise that with each AI generated artwork, a small child dies from starvation?
We shall double our efforts!
I know who I'm betting on
A solution to overpopulation?
Perfect cozy spot, I can spend hours just staring at a nice fireWe shall double our efforts!
I've been returning to Tolkien-esque fantasy recently. Ahh fairest Galadriel, Queen of the Elves....
I've been returning to Tolkien-esque fantasy recently. Ahh fairest Galadriel, Queen of the Elves....
I'm new to AI art and find it really interesting to see how easy it is to create images. A lot of the time it's not really exactly what you were expecting but it can have a slot machine type compulsiveness. I visit Night Cafe mostly but it seems so geared towards really sickly sweet stuff like neon butterflies flying over a full moon or pastel laughing kittens. I tried a real fantasy cliche of a smiling hobbit stealing treasure in a cave with a very obvious huge scorpion lurking behind him but nobody seemed to get the message and all the comments were like "nice work" and "super cute portrait!" Disney had more of edge than most Night Cafe.
With Night Cafe I usually pick a random model. Then I choose either sinister or horror styles, runtime: long, single image, and 16:9 ratio. Then I keep clicking until I get something that resembles the text prompt, and I like. One thing I learned the other day is that most AI art generators put more emphasis on the first parts of text prompts and the least amount towards the end. So if your prompt is too long sometimes the generator will ignore the last bits of the prompt.
These two use all the same parameters except I changed the model. The results were very similar image but different styles. Night Cafe generators are getting better because in the past the images produced by changing only the model would be drastically different but the examples below you have nearly same subject position and layout between the two images. Text prompt was "Saint Nicholas and Garrett from Thief II sitting next to a fire drinking ale in an old english pub on Christmas eve."
That's a telltale sign that the comments are from bots lol
I like Flux, Ideogram and my company's internal generator (originally based on Dall E 3, but now is its own thing)
Flux is my favourite, almost always precise, but has a plasticky quality
Ideogram is great for fine details, but if you zoom in you sometimes see distortion or inaccuracies
My work one is photorealistic, but has all sorts of composition and perspective issues
Victorian watercolours
Blue dream
[IMG][/IMG]
Posting Permissions
Reply With Quote
