



This is an embedding merge of the line: <'pixel art style'*0.2595 + 'tapestry 8bit'*0.0146 + '16bit'*0.0812 + 'dithering contrast'*0.0851 + 'pixel cross stitch'*0.0377 + 'bead embroidery mosaic'*0.1274 + 'pointillism'*0.1429 + 'nostalgic arcade game'*0.0436 + 'indie retro game'*0.0657 + 'spritesheet'*0.0319 + 'impressionist landscape render'*0.1104 + 'jpg compression artifact'*-0.0227 + 'highly detailed noise'*-0.1154 + 'sloppy messy blurry'*-0.0458 + 'modern television system'*-0.0633 + 'scifi techno display'*-0.0527>
Produced via hand pseudo-training. IE, automatically produce large numbers of variants, generate images, choose the images I like and use that as the seed for the next generation, repeat.
It combines a number of concepts related to images being produced via lots of semi-large spots, to hopefully produce images that look good when downscaled. Also designed with an eye towards posterizing (limiting the pallet to say, 16 colors)
A note on Gimp: when downscaling, don't use LoHalo, or anything, just use None. If you want to pixelize ahead of time, it's often a good idea to go with circle instead of square, then remove the alpha later using standard transfer alpha to layer mask, delete layer mask. A light unsharp mask can somtimes help after you've downscaled.
1. Quyền đối với các mô hình được đăng lại thuộc về người sáng tạo ban đầu.
2. Người sáng tạo gốc muốn xác nhận mô hình vui lòng liên hệ nhân viên SeaArt AI qua kênh chính thức. Nhấp để xác nhận
