yidall

Rosa's Self Pleasure? ひとりえっち？メイ(16illustrations)

Added 2023-11-14 10:00:05 +0000 UTC

昨日のスバル、今日のメイとリクエストいただいていたものです。長い期間で少しずつ混ぜながら投稿していきます。スバルもメイもまだ作成経験が浅いので納得いくものがなかなかできませんが、失敗・経験あるのみなのでこういうリクエストをいい機会と捉えて、少しずつ改善できたらと思っています！

AIイラストなのに練習が必要ってなんだか一般的なイメージと違って面白いですよね。自分はあまり追加学習等に頼りたくないので、どうしてもトライアンドエラーが必要です（追加学習はそれはそれでたぶんいろいろ大変）。大枠で作って、過不足を特定して少しずつ近づけていくイメージです。詰めていく過程で大枠ごと別の発想で作り直さなければいけなくなったり、技術や特性的に限界があったり…。

自分は発想が写真制作ベースなので制作過程もそれに寄ってるのですが、人によっていろんなやり方があって面白そうです。

アンケートでお寄せいただいている質問に関しては今週末くらいにでもブログでまとめて回答させてください！

Yesterday's Subaru and today's Mei were based on requests I received. I plan to gradually incorporate such requests into my posts over a long period. Since my experience in creating both Subaru and Mei is still limited, it's been challenging to produce results that I'm fully satisfied with. But I believe in learning from failures and experiences, so I see these requests as great opportunities to gradually improve!

It's funny, isn't it? The need for practice in AI illustration seems contrary to the general perception. I prefer not to rely too much on additional learning, so inevitably, a lot of trial and error is involved (and additional learning has its own set of challenges, I'm sure). The process involves creating a rough outline, identifying what's missing or excessive, and then refining it bit by bit. Sometimes, I have to completely rethink and restart from a different concept, or I encounter technical limitations.

My approach is based on my background in photography, so my creation process leans in that direction. It's fascinating how different artists have their own unique methodologies.

Regarding the questions submitted through the survey, I'll compile and respond to them in a blog post around this weekend!