adding stylistic prompts actively changes the some of what I would consider to be content
Your examples here are not good since e.g. ”...painting by Alphonse Mucha” is not just a rewording of ”...in the style of Alphonse Mucha”: the former isn’t a purely stylistic prompt. For a [painting by x], x gets to decide what is in the painting—so it should be expected that this will change the content. Similarly for “screenshots from the miyazaki anime movie”.
Of course it’s still a limitation if you can only get really good style results by using such not-purely-stylistic prompts.
That’s a reasonable point. I have definitely found that saying “a painting by X” or a “a movie by X” gets results that a) I personally like much better, and b) are much more consistently and recognizably in the requested style!
I’m not sure whether “in the style of X” just ends up being less of a strong hint for DALL-E, or whether it’s pulling on a much bigger set of training data. Maybe there are all sorts of images online labeled as “in the style of Alphonse Mucha” by people who don’t actually know how to assess styles? Anyway, this is “A woman at a coffeeshop working on her laptop and wearing headphones, in the style of Alphonse Mucha” and it’s fine but it’s much less what I ordered!
This is great! Thanks.
A nitpick:
Your examples here are not good since e.g. ”...painting by Alphonse Mucha” is not just a rewording of ”...in the style of Alphonse Mucha”: the former isn’t a purely stylistic prompt. For a [painting by x], x gets to decide what is in the painting—so it should be expected that this will change the content.
Similarly for “screenshots from the miyazaki anime movie”.
Of course it’s still a limitation if you can only get really good style results by using such not-purely-stylistic prompts.
That’s a reasonable point. I have definitely found that saying “a painting by X” or a “a movie by X” gets results that a) I personally like much better, and b) are much more consistently and recognizably in the requested style!
I’m not sure whether “in the style of X” just ends up being less of a strong hint for DALL-E, or whether it’s pulling on a much bigger set of training data. Maybe there are all sorts of images online labeled as “in the style of Alphonse Mucha” by people who don’t actually know how to assess styles? Anyway, this is “A woman at a coffeeshop working on her laptop and wearing headphones, in the style of Alphonse Mucha” and it’s fine but it’s much less what I ordered!