Search
(Supports wildcard *)Artist
- ? kinkcurator (manipper) 9
General
- ? aware 3185
- ? bottomless 40822
- ? breasts 103370
- ? expressionless 20374
- ? femdom 30122
- ? femsub 132175
- ? green hair 6121
- ? happy trance 46670
- ? nude 38573
- ? purple eyes 6694
- ? robot girl 1089
- ? tech control 23070
- ? topless 44523
- ? wholesome 586
- ? yuri 6333
Meta
- ? ai art 2176
- ? caption 8912
- ? manip 16629
- ? text 82431
Statistics
- Id: 193972
-
Posted: 2024-01-25 22:21:48
by kinkcurator - Size: 1920x1080
- Rating: Explicit
- Score: 70 (vote up)
This image has been resized. Click here to view the original image.
Always view original.
Don't show this message.
1
>> #533362
Score: 0 (vote Up)
And more importantly: it's cute. Cute, I say!
>> #533366
Score: 0 (vote Up)
Query; I am looking forward to trying to post my own ai art captions soon, with first image pending moderation as of the time of writing. Do you, newly experianced, have any tips?
>> #533373
Score: 0 (vote Up)
Pay attention to readability, keep your text large and your font legible, use borders/backgrounds to ensure contrast. Pay attention to layout, tweak text to avoid lines with just 1-2 words. Keep the writing short and punchy, I *always* have to trim way down from my first draft, just to fit everything.
As a more subjective opinion: have a strong, novel concept. Especially if you're using AI art - it should be something that would be hard to find a good pre-existing image for. I try to have some neat, innovative thing going on with every caption, whether that's a format trick, a trope subversion, a text effect, whatever.
For AI specifically, work on both the image and text in parallel - let details from one, feed into the other. If the AI adds a cool detail, integrate it into the text; if the text requires something, add it to the image. Also, your generation prompt is a good starting point for tagging your posts, a lot of the tags will be the same. Learn the advanced tools at your disposal: this one used a symmetry tool, to force the layout I wanted. Regional prompting can give you areas of low detail, to lay your text over without needing a full flat-color background - I've started doing this in some upcoming captions, can't remember if I've used it before. Grab a bunch of Loras, too, and not just hypno eyes ones.
Good luck, and hope this helped!
>> #533379
Score: 0 (vote Up)
>> #533442
Score: 1 (vote Up)
>> #533516
Score: 1 (vote Up)