Derpibooru
Filters (Default) SettingsRegisterLogin
Subjects of Equestria! We are pleased to kick off the Nightmare Night Festivities and the 14th anniversary of the return of our beloved Princess Luna! Join us in celebration!
114 160 152 8 9
Subscribe
Galleries
Featured in
Sign in to create a gallery
Related
View VS Download DS
Uploaded by aiLex
 1152x1152 PNG 1.88 MB
Interested in advertising on Derpibooru? Click here for information!
Ministry of Image - Fanfiction Printing

Help fund the $15 daily operational cost of Derpibooru - support us financially!

Description

Two royal sisters after a long separation.

Tags
Edit History (10)
suggestive
196740 ai content
31345 ai generated
29840 generator:stable diffusion
17801 prompter:ailex
28 princess celestia
116519 princess luna
121181 anthro
379647 g4
2115595 armor
32817 blushing
292711 breasts
412460 bust
82730 duo
200331 duo female
40757 eyes closed
146685 female
1895392 hug
39402 incest
18871 jewelry
123565 kiss on the lips
6890 kissing
34111 lesbian
122369 looking at each other
37853 looking at someone
20429 moon
33386 night
40659 royal sisters
7404 ship:princest
2957 shipping
264909 siblings
24926 sisters
20110 window
16258
Source
Add
not provided yet
Report Share List favoriters

Comments

Edit Preview
Syntax quick reference: **bold** *italic* ||hide text|| `code` __underline__ ~~strike~~ ^sup^ %sub%

Detailed syntax guide

9 comments posted
aiLex

@koznyshev
Artists have a style and the artistic merit allows them to keep creating good art, but an AI can be fleeting. Like it’s been dreamed up.
Most professional prompters sooner or later settle on just one or two image styles. Just look at their galleries, for example: kalmar, FoylerTF. Isn’t it true that the style of the images is similar? I often find out who generated it simply by the style of the posted AI image.
I’m still trying out different styles, searching the one that’s most convenient for me, but in general, I can already distinguish two main styles of the images I generated (this style is not included in them, it’s just an experiment).
Rarely do I find a style from an AI that I really like that has more than one image
The same with living artists. Many times, having found a stunning image, I went to the artist’s gallery - and there was only a sequence of endless YCHs or adopts.
In general, there are still quite a few prompters, much less than artists, so you also meet those you like hundreds of times less often. In the next five years the situation will significantly improve.
Yeah but I wanted more of this style, haha.
Well, maybe later I will generate more images in the same style. Currently there are plans to create a two-panel comic about Ember in trouble (slightly similar to this generation with Gilda). Then I plan to create an anthro image of Queen Novo, because there are quite a few good images of her. Then, perhaps, I will return to this style.
Posted Report
koznyshev

@aiLex
Don’t like AI?
Actually I do, I meant it more that there isn’t more. Artists have a style and the artistic merit allows them to keep creating good art, but an AI can be fleeting. Like it’s been dreamed up. Rarely do I find a style from an AI that I really like that has more than one image, and even when it does, AI particularities usually end up meaning that there’s one good one I really like and then 3-4 slightly cursed images in the usual AI madness way.
Maybe it’s due to the work you’ve done cleaning it up, but I was really impressed by this one. Well done.
pictures of princesses created by AI in a variety of styles
Yeah but I wanted more of this style, haha.
Posted Report
aiLex

@koznyshev
I was really hoping this was an artist
Don’t like AI?
Well prompted, I guess?
Good prompting, then some inpainting with the same AI and about two hours of manual fixing of clearly visible AI artifacts in a graphic editor (Krita).
I wonder if the AI set could do full body shots of the princess sisters.
Why not? Here on Derpibooru you can find some full body pictures of princesses created by AI in a variety of styles.
Posted Report
koznyshev

Dang, I was really hoping this was an artist. Well prompted, I guess? What a style. I wonder if the AI set could do full body shots of the princess sisters.
Posted Report
aiLex

@Background Pony #105E
In this case you will have to wait at least a week. Because the model that generates in this style unfortunately does not know anything about Chrysalis.
I’m currently collecting a dataset for training changeling LoRA, but this will take a noticeable amount of time.
Posted Report